PROCEEDINGS OF

SARC
INTERNATIONAL CONFERENCE
Madurai, India

Organized by

Date of Event:

03rd November, 2024

In Association with

Corporate Address

INSTITUTE OF RESEARCH AND JOURNALS

Plot No- 30, Dharma Vihar, Khandagiri, Bhubaneswar, Odisha, India
Mail: info@theires.org, www.iraj.in
Publisher: Institute for Technology and Research (ITRESEARCH)

2024, SARC International Conference, Madurai, India

ISBN: 978-93-90150-25-0
Edn: 115

No part of this book can be reproduced in any form or by any means without prior written
permission of the publisher.

Disclaimer: Authors have ensured sincerely that all the information given in this book is
accurate, true, comprehensive, and correct right from the time it has been brought in writing.
However, the publishers, the editors, and the authors are not to be held responsible for any
kind of omission or error that might appear later on, or for any injury, damage, loss, or
financial concerns that might arise as consequences of using the book.

Type Set & Printed by:

Institute for Technology and Research (ITRESEARCH)
Khandagiri, Bhubaneswar
About Institute of Research and Journals:
 Institute of Research and Journals (IRAJ) is an advanced Non-profit technological
forum under Peoples Empowerment Trust, for the Researchers & Scholars "to promote the
progress of Science and Technology" by displaying their knowledge in the vicinity of Science
and Technology for the service of mankind and the advancement of general welfare.

Objective of IRAJ:

 To provide a world class platform to researchers to share the research findings by

organizing International/National Conferences.
 To use the research output of the conference in the class room for the benefits of the
students.
 To encourage researchers to identify significant research issues in identified areas, in
the field of Science, Engineering, Technology and Management.
 To help dissemination of their work through publications in a journal or in the form of
conference proceedings or books.
 To help them in getting feedback on their research work for improving the same and
making them more relevant and meaningful, through collective efforts.
 To encourage regional and international communication and collaboration; promote
professional interaction and lifelong learning; recognize outstanding contributions of
individuals and organizations; encourage scholar researchers to pursue studies and
careers in circuit branches and its applications.
 To set up, establish, maintain and manage centers of excellence for the study of /on
related subjects and discipline and also to run self supporting projects for the benefit
of needy persons, irrespective of their caste, creed or religion.

About SARC:

South Asian Research Center (SARC) is a nonprofit association is dedicated to exploring

and innovating new scientific ideas in the field of engineering and technology. The purpose
of the center is to facilitate collaboration among researchers from different scientific fields
who share a common interest in the study, design and creative uses of information
technologies in order to enhance the social and everyday lives.

The centre aims to address important societal challenges such as enhancing health care and
improving education through interdisciplinary research and innovation. The South Asian
Research Center (SARC) is a multi-disciplinary and cross-faculty initiative within the
Institute of Technology and Research, Bhubaneswar, India an ISO 9001:2008 Certified
Organization joining expertise from the Faculty of Science and the Faculty of Engineering.
Conference Committee

Program Chair:

Dr. P. Suresh
M.E, Ph.D.,
Professor, Karpagam College of Engineering,
Coimbatore, Tamil Nadu, India

Managing Director:
Mr. Bijan Kumar Barik

Conference Convener:

Mr. J.R. Pattanayak, SARC

Mob: +91-8280047487

Publication Head:
Mr. Manas Ranjan Prusty, IRAJ, India

INTERNATIONAL ADVISORY MEMBERS

Prof. Goodarz Ahmadi, Harun Bin Sarip

Professor, Mechanical and Aeronautical Engineering, Head of Research and InnovationDept, UniKL-MICET
Clarkson University, USA Doctorate: Université de La Rochelle, France, Member:
International Society of Pharmaceutical Engineer, Singapore
Dr Chi Hieu Le, Chapter
Senior Lecturer, University of Greenwich. Kent ME4 4TB.
United Kingdom Dr.Bilal Ali Yaseen Al-Nassar
The World Islamic Sciences and Education University (WISE)
PROF. (ER.) Anand Nayyar Faculty of Business and Finance
Department of Computer Applications & I.T.KCL Institute of Department of Management
Management and Technology, Jalandhar Information System (MIS), Amman- Jordan
G.T. Road, Jalandhar-144001,Punjab, India.
Dr. Md. Al-Amin Bhuiyan
Prof. R. M. Khaire, Associate Professor
Professor,Dept. Of Elex. and Telecommunication, Dept. of Computer Engineering
B, V University,India King Faisal University
Al Ahssa 31982, Saudi Arabia
Dr.P.Suresh,
Professor,Karpagam College of Prof. (Er.) Anand nayyar
Engineering,Coimbatore,Tamilnadu Department of Computer Applications & I.T.
KCL Institute of Management and Technology, Jalandhar
Mark Leeson G.T. Road, Jalandhar-144001
Associate Professor (Reader) Punjab, India
Area of Expertise:nanoscale communications,
evolutionary algorithms, network coding and communication Prof. Aleksandr Cariow
systems institution or Company: West Pomeranian University of
Technology, Szczecin
Dr. P. K. Agarwal
Professor,Deptt. of Civil Engineering, MANIT Bhopal ,Ph. D: Dr. P. K. Agarwal
IIT Kanpur Professor,Deptt. of Civil Engineering, MANIT Bhopal ,Ph. D:
M.E: Civil Engg.IIT Roorkee, Membership: Indian Road IIT Kanpur
Congress (IRC), Institute of Urban Transport (IUT) M.E: Civil Engg.IIT Roorkee, Membership: Indian Road
Congress (IRC), Institute of Urban Transport (IUT)
Shahriar Shahbazpanahi
Islamic Azad University, Department of Civil Engineering, Dr. VPS Naidu
Sanandaj, Kurdistan, Iran, Principal Scientist & Assoc. Prof., MSDF Lab, FMCD
PhD (Structural Engineering), CSIR - National Aerospace Laboratories, Bangalore, India
University Putra Malaysia, Malaysia,2009-Present
Mr. P. Sita Rama Reddy Dr. Chandra Mohan V.P.
Chief Scientist ,Mineral Processing Department, CSIR - Assistant Professor, Dept. of Mech. Engg., NIT Warangal,
Institute of Minerals & Materials Technology Warangal. Ph.D : Indian Institute of Technology(IIT),Delhi
Bhubaneswar,India, M.Tech. (Chem. Engg., IIT, KGP) M.B.A: Alagappa University

Dr.P.C.Srikanth, Prof. I.Suneetha,

Professor & Head, E&C Dept, Malnad College of Associate Professor,Dept. of ECE,AITS,Tirupati,India
Engineering,Karnataka
Senior Member IEEE, Secretary IEEE Photonics Society, Dr.s. Chandra Mohan Reddy,
M.Tech: IIT, Kanpur, Ph.D: In IISc Photonics lab Assistant Professor (SG) & Head,Dept. of Electronics &
Communication Engineering,JNTUA College of
Engineering,Pulivendula,Ph.D,J.N.T. University Anantapur,
Prof. Lalit Kumar Awasthi, Anantapuramu
Professor,Department of Computer Science & Engineering
National Institute of Technology(NIT-Hamirpur), Gurudatt Anil Kulkarni,
PhD, IIT, Roorkee, M. Tech, IIT, Delhi I/C HOD E&TC Department,MARATHWADA MITRA
MANDAL’S POLYTECHNIC


 TABLE OF CONTENTS
Sl. No. TITLES AND AUTHORS Page No.

01. Effective Landmark Regression Using Attention Based-HRNet for Satellite Pose 1-7
Estimation
 Abraham Paul, Anoop G L, Ganesh Kumar R, Aryan Singh, Sri Aditya Deevi,
Ravikumar Lagisetty

02. Development of Water Resources Management Scheme for Sedawgyi Dam Using 8-13
WEAP Model
 Thet Zin Htoo, Yin Yin Htwe, Nilar Aye

03. Analysis of Water Allocation and Demand Management in Katha Basin 14-20
 Win Lwin Tun, Nilar Aye, Cho Cho Thin Kyi

04. Analyzing Cognitive Distortions and Behavioral Patterns in Children With 21-23
Learning Disabilities
 Sushmita Singh

05. The Impact of Personality Types in Adolescents: Implications for Mental Health, 24-27
Development, and Lifelong Outcomes
 Shaily Gambhir

06. AI: Unethical Usage May Sabotage Democracy 28-33

 Joseph Philips, R. Subramani

07. Accurate Brain Tumor Classification by Using MobileNetV1 34-44

 Sumit Kumar Yadav, Asif Khan

08. A Comprehensive Review of Offshore Carbon Capture and Storage Technologies 45-54
Innovation Challenges and Future Horizon
 Dhivakar Poosapadi


 EDITORIAL

It is my proud privilege to welcome you all to the SARC International Conference at

Madurai, India. I am happy to see the papers from all part of the world and some of the best
paper published in this proceedings. This proceeding brings out the various Research papers
from diverse areas of Science, Engineering, Technology and Management. This platform is
intended to provide a platform for researchers, educators and professionals to present their
discoveries and innovative practice and to explore future trends and applications in the field
Science and Engineering. However, this conference will also provide a forum for
dissemination of knowledge on both theoretical and applied research on the above said area
with an ultimate aim to bridge the gap between these coherent disciplines of knowledge. Thus
the forum accelerates the trend of development of technology for next generation. Our goal
is to make the Conference proceedings useful and interesting to audiences involved in
research in these areas, as well as to those involved in design, implementation and operation,
to achieve the goal.

I once again give thanks to the Institute of Research and Journals, All Conference Series,
ACN & GSN organizing this event in Madurai, India. I am sure the contributions by the
authors shall add value to the research community. I also thank all the International Advisory
members and Reviewers for making this event a Successful one.

Editor-In-Chief

Dr. P. Suresh
M.E, Ph.D. Professor and Controller of Examinations,
Karpagam College of Engineering,
Coimbatore, India.


 EFFECTIVE LANDMARK REGRESSION USING ATTENTION
BASED-HRNET FOR SATELLITE POSE ESTIMATION
1
ABRAHAM PAUL, 2ANOOP G L, 3GANESH KUMAR R, 4ARYAN SINGH, 5SRI ADITYA DEEVI,
6
RAVIKUMAR LAGISETTY
1
M.Tech Student, Department of CSE, Christ University, Kengeri, Bangalore
2
Assistant Professor, Department of CSE, Christ University, Kengeri, Bangalore
3
Associate Professor, Department of CSE, Christ University, Kengeri, Bangalore
4
B.Tech CSE Student, SRM Institute of Science & Technology, Chennai
5
Scientist/Engineer-SC, Mission Simulation Group (MSG), URSC (ISRO), Old HAL Road, Bangalore
6
Group Director, Mission Simulation Group (MSG), URSC (ISRO), Old HAL Road, Bangalore
E-mail: 1s.abraham@mtech.christuniversity.in, 2anoop.gl@christuniversity.in, 3ganesh.kumar@christuniversity.in,
4
ar8930@srmist.edu.in, 5saditya@ursc.gov.in, 6rkkumarl@ursc.gov.in

Abstract - For many space missions, it is important to estimate the position and orientation of the satellite for operations
such as docking and debris removal. It involves the following stages of object detection, landmark regression and pose
estimation. For objection detection, we used Faster-RCNN with HRNet as the backbone, the landmark regression part is
done using AHRNet architecture, the pose estimation is implemented using PnP algorithm. Firstly, each image was labeled
for object detection with bounding boxes around the satellite images created in Blender which were then used to train for
satellite detection. An AHRNet was further trained for landmark regression using a 4fold cross-validation approach which
involved splitting the dataset into multiple training and validation sets to enhance the Intersection over Union (IoU) metric.
After the landmark regression provided a 2D projection of 3D ground truth points, the PnP algorithm was then used for pose
estimation. To improve pose estimation accuracy, we integrated solvePnP with an iterator argument utilizing the Levenberg-
Marquardt (LM) method to reduce noise and outliers. Our methodology significantly enhances the precision and efficiency
of the translation and rotation error of the satellite during the docking processes offering a viable solution for autonomous
space missions with potential for future improvements in domain adaptability through the development of unsupervised
domain adaptation models. The results of the landmark regression using AHRNet shows an improved reprojection error,
orientation and translation errors.

Keywords - Satellite Pose Estimation, AHRNet, Landmark Regression, Pose Estimation

I. INTRODUCTION A key element of our method is the application of 4-

foldcross-validation to enhance the model's
With the growing interest in space exploration and robustness and generalization. The dataset was
satellite technology, there is an increasing demand for strategically split into multiple training and validation
reliable and efficient methods for satellite servicing sets to prevent overfitting, ensuring each valid JSON
operations such as docking, debris removal, and file contained unique images to maintain diversity.
maintenance. A critical component of these The Perspective-n-Point (PnP) algorithm[3] was then
operations is the accurate pose estimation of non- employed to estimate the satellite's pose using weight
cooperative satellites which are often unresponsive files obtained from the landmark regression. To
due to a lack of communication or the absence of further improve accuracy, we integrated the solvePnP
predefined markers. Accurate pose estimation is iterator within the PnP solver (cv2 function) into our
essential for the success of docking and other pose estimation pipeline utilizing the non-linear
proximity maneuvers ensuring both the safety and Levenberg-Marquardt minimization method to
effectiveness of space missions [4], [5]. mitigate noise and remove outliers. This combination
significantly enhances the accuracy and reliability of
In this paper, we present an optimized and effective the pose estimation process. By utilizing attention
approach for landmark regression and pose estimation based high resolution representation network like
using advanced computer vision techniques. To AHRNet, the landmark regression is implemented to
begin, a comprehensive dataset of 18,000 satellite enhance the pose estimation of the satellite.
model images was generated using Blender software We conclude that the AHRNet implementation for
where world, camera and object (satellite) frames landmark regression resulted in a better translation
were defined. Various satellite models and camera and rotation error which in turn helped in estimation
positions were simulated to create a diverse set of of the pose of the satellite in an effective manner.
images. These images were processed using a
customized Faster R-CNN approach [1], [16] for II. RELATED WORK
object detection [9], with bounding boxes annotated
around the satellite in each image for training the The existing work focused mainly on monocular pose
model. The annotated dataset was subsequently used estimation. The following are some of the methods
to train AHRNet for landmark regression [10], [14]. adopted in monocular pose estimation:

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

1
 Effective Landmark Regression Using Attention Based-HRNet for Satellite Pose Estimation

A. Keypoint methods Indian Space Research Organization (ISRO). These

Traditional approaches for estimating position and satellite images were labeled for object detection
orientation often rely on manually crafted descriptors using a Python script. The Faster-RCNNalgorithm [1]
and keypoint detectors. The process typically begins was then applied to perform object detection with
by establishing correspondences between two- bounding boxes drawn around the satellite to pinpoint
dimensional or two-dimensional-to-three-dimensional its location.
landmarks. From there, non-linear optimization is
applied to this correspondence set to estimate the
object's posture. Keypoints areautomatically
recognized and then described using heuristic metrics
based on geometric and photometric invariance [9].
While keypoint-based methods can be fairly robust,
they tend to struggle in environments with significant
variations in pose or lighting conditions.
Alternatively, some methods integrate various steps
into a unified network, using convolutional neural
networks (CNNs) designed for 2D landmark
localization to detect the satellite’s bounding box
coordinates or predict landmark positions [13]. These
CNN-based techniques have become widely used in
satellite pose estimation tasks.

B. End to End learning methods

These methods aim to bypass the use of traditional
geometric solvers, instead employing end-to-end
approaches that directly map key feature
representations to a 6-D output space. Typically, Figure 1: Complete Architecture of the Model
these techniques leverage the architecture of
convolutional neural networks (CNNs) [15] to learn a For landmark regression, the 3D satellite model was
complex function that transforms an input dataset into reconstructed with 7 landmark coordinates placed on
the corresponding pose output. While end-to-end distinct features of the satellite. Cross-validation was
methods have shown considerable potential, they employed by dividing the dataset into several training
have not yet achieved the same level of accuracy as and validation sets, ensuring robustness and diversity
geometry-based approaches. Handcrafted algorithms, in the model. The AHRNet algorithm was used to
which match different parts of the satellite to train the model for landmark regression [1]. The
predefined templates are still able to deliver precise 6 output predictions are then used by PnP algorithm for
degrees of freedom (DoF) pose estimations. In estimating the pose of the satellite. To further
contrast, CNNs in learning-based approaches [6] are enhance the accuracy and reliability of the pose
used either to classify data into specific poses or to estimation, the solvePnP iterator argument was
train models to directly detect the 6 DoF pose of the integrated into OpenCV library to address noise and
object [10]. outlier removal. This comprehensive approach is
designed to significantly improve the precision and
C. Concentrated learning methods efficiency of satellite docking procedures which
Facebook AI Research’s innovative DensePose provides a robust solution for autonomous space
estimation method establishes detailed missions [17].
correspondences between a two-dimensional image
and the three-dimensional surface representation of 1. Dataset Creation
an object, mapping connections at both the pixel and
patch levels [14]. But the existing work do not For any deep CNN algorithm to be both accurate and
provide effective results for landmark detection. effective, the dataset plays a crucial role. Therefore,
while generating the dataset for the satellite model,
III. METHODOLOGY our primary focus was on creating a wide variety of
distinct and random satellite images. We considered
Figure 1 illustrates the complete architecture of our multiple factors such as varying lighting conditions,
model, starting with data generation. A 3D satellite the distance between the camera and the satellite, and
model was incorporated into Blender 3D software to the camera’s orientation from different angles, to
create a dataset of 18,000 training images and 500 ensure a diverse dataset and minimize the risk of
test images. For 4-fold crossing validation we are overfitting. Additionally, we considered camera
using around 14,000 train images and validation set resolution and focal length to obtain the camera's
of 4000 images recreated in the HILS Lab of the intrinsic matrix [19]. We generated the satellite
dataset in Blender 3D software where we set the
Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024
2
 Effective Landmark Regression Using Attention Based-HRNet for Satellite Pose Estimation

world, camera, and object frame coordinates. Using dataset, as it helps mitigate overfitting by allowing
this approach, we generated around 18,000 satellite the model to recognize patterns beyond a single data
images for model training, which were later tested on partition. Additionally, cross-validation improves the
test set of 500 satellite images from Blender. model’s ability to generalize to new and unseen
satellite imagery, which is common in the space
The dataset of GSAT-12R satellite model (inspired industry. For our landmark regression model, we
from [21])which is diverse was generated using implemented a 4-fold cross-validation approach. This
Blender tool by incorporating an element of strategy helps reduce overfitting and enhances
systematic randomness to ensure proper training of accuracy during the final stages of the process.
the model:
4. Landmark Regression
 The range between the camera and the satellite is
between 3 meters and 19 meters. Each training image is accompanied by a set of
 The camera positions were restricted to only a ground truth 2D landmarks and a bounding box.
part of the sphere where 00 ≤ ɸ ≤ 800 These labels are used to train a landmark regression
 Thompson sampling technique was used to model, which predicts the two-dimensional keypoints
ensure camera positions have good sphere in test images. A cross-validation strategy is
coverage employed to minimize the risk of overfitting.
 The camera positions were perturbed to ensure Specifically, we used the pose-hrnetw48 model,
the object is not always at center of image which which, as its name suggests, features 48 channels in
can cause redundancy in the dataset the highest-resolution feature maps [16], to predict
 The camera was also rotated along boresight axis 2D keypoint positions as outlined in [5]. The model
to ensure the dataset is diverse and all the images generates a tensor containing seven keypoints or
where captured with different poses landmarks each representing a 3D landmark as shown
in Figure 3. This type of mapping allows the model to
2. Object Detection focus purely on learning the position of each 3D
landmarkwithout needing to infer correlations
The architecture of our model begins with labeling between heatmaps and 3D landmarks.
the satellite images required for object detection. This
labeling process is handled by a Python script created
in Blender, specifically for the training images. We
then train the satellite image dataset using the Faster
R-CNN architecture [4] with HRNet as the backbone
[1]. To run the object detection pipeline, we utilize
the mmcv platform [9]. Figure 2 shows a visual
example of bounding boxes applied to the satellite
images for object detection.

Figure 3: The keypoints of the Satellite Model for Landmark

Regression

AHRNet is particularly effective at generating high-

resolution heatmaps with exceptional accuracy as it
maintains a high-resolution representation while
capturing features across different resolutions to
ensure that no details are missed [1]. The model was
trained using the Adam optimizer for 10 epochs,
achieving superior accuracy compared to other
Figure 2: Visual example of bounding boxes applied to satellite similar models.We implement the landmark
images for object detection. regression using the attention based high resolution
network (AHRNet)which helps in successfully
3. Cross Validation predicting the landmarks with a higher accuracy and
lower error rate.
We opted to use the cross-validation method in our
machine learning pipeline for several key reasons. It The following diagram illustrates the architecture of
is particularly useful when working with a limited the AHRNet which is used for landmark regression:
Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024
3
 Effective Landmark Regression Using Attention Based-HRNet for Satellite Pose Estimation

The figure 5 shows the final estimated pose of the

satellite for the GSAT-12R satellite model generated
using Blender.

Figure 4: Architecture ofAHRNet (adopted from [2])

Figure 4 shows the overall architecture of the

AHRNet model where attention blocks are used Figure 5: Final Estimated Pose of the Satellite Model
which was inspired from [22]. AHRNet builds upon
the standard HRNet architecture by incorporating IV. RESULTS
attention mechanisms to further enhance its ability to
capture fine-grained spatial details while maintaining We evaluated our proposed model using standard
high-resolution representations. metrics to quantitatively assess and compare its
accuracy with other leading models, as shown in
The attention modules allow the network to focus on Table I.
the most relevant parts of the image, improving the
accuracy of landmark detection and regression, For object detection, we used the Intersection over
especially in complex or noisy environments. Union (IoU) metric for each satellite image in the
dataset. This metric calculates the ratio of the area of
In the context of landmark regression, AHRNet not intersection to the area of union, resulting in a value
only preserves the spatial information from the high- between 0 and 1.
resolution feature maps but also uses attention to
selectively emphasize key areas, ensuring more Our dataset consists of satellite images with only a
precise predictions of 2D keypoints. 3D face rather than a complete 3D model and it
includes images captured from distances from
We trained our AHRNet landmark regression model 3meters to as far as 19 meters from the camera,
for 10 epochs on a NVIDIA A4000 GPU with a which adds complexity to detection and reduces
VRAM of 16GB in an Ubuntu 22.0.4 workstation. accuracy. Nonetheless, our model is effective under
these more challenging conditions.
5. Pose Estimation
We assess the predicted pose of each image using two
The last and final step in the process is determining error metrics: rotation error (ER) and translation error
the position and orientation of the satellite in the test (ET).
images using the 2D-3D correspondences. We begin
by combining all the predicted 2D landmark The rotation or euler ground truth for an image is
coordinates obtained from the landmark regression denoted as eul∗ with its corresponding estimate as eul
after applying 4-fold cross-validation. (in degrees).Similarly, the ground truth translation
vector is tr∗ and its estimate is tr (in meters).
Next, we apply the perspective-n-point algorithm to
estimate the satellite's three-dimensional position and We define ER and ET as follows:
orientation [5], [12].
ER=∣eul* - eul∣, where
While the initial results were promising, we aimed to ER is the rotation error,
refine the model further by addressing potential noise eul*is the ground truthrotationvalue,
or outliers. To achieve this, we employed the non- eulis the estimated rotation value
linear Levenberg-Marquardt method [3], which
adaptively combines the Gauss-Newton and gradient ET=∣∣tr∗- tr∣∣2, where
descent methods by adjusting the damping coefficient ET is the translation error,
based on various parameters and applications. Using tr* is theground truthtranslationvalue,
PnP algorithm, we were able to estimate the pose of tr is theestimated translation value
the satellite in an effective manner.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

4
 Effective Landmark Regression Using Attention Based-HRNet for Satellite Pose Estimation

These are the test results:

Sl Error Metric Value

Mean Translational
1 0.016617009651114525
Error (ET) [in m]

Mean Rotation Error

2 0.27031458666224356
(ER) [in deg]

Mean Absolute 0.00988876 (X)

3 Translational Error 0.00972417 (Y)
(|tgt - tpred|) [in m] 0.00770925 (Z)
Figure 9: Orientation Error vs Radial Distance
Mean Absolute 0.37910969 (X)
4 Orientation Error 0.38968987 (Y) Figure 6 and 7 show the histograms for the mean
(|eulgt - eulpred|) [in deg] 0.0421442 (Z) translation and mean orientation errors respectively
based on its frequency for the effective pose
The following are the graphs of the results: estimation of the satellite.

The translation and orientation error for all the

keypoints of the satellite shows an improved
landmark detection based on the radial distances from
3 meters to 19 meters as shown in Figure 8 and 9
respectively. The usage of attention
basedHRNetmodel has significantly improved the
accuracy ofthe landmark regression part of satellite
pose estimation.

The following table shows the results of reprojection

error for the keypoints E1, E4, W1, W4, ST, NT,
NFAOut on our satellite model after performing
Figure 6: Translation Error vs Frequency landmark regression using AHRNet architecture:

Sl Reprojection Error Value (in pixels)

Reprojection Error for

1 1.0339158276185025
E1

Reprojection Error for

2 0.9305897301039263
E4

Reprojection Error for

3 1.075804264630154
W1
Figure 7: Orientation Error vs Frequency
Reprojection Error for
4 0.5296732370876553
W4

Reprojection Error for

5 0.775747395160931
ST

Reprojection Error for

6 0.8886139873071606
NT

Reprojection Error
7 0.6476353685377223
forNFAOut

Overall Reprojection
0.4743316143098673
Error
Figure 8: Translation Error vs Radial Distance

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

5
 Effective Landmark Regression Using Attention Based-HRNet for Satellite Pose Estimation

V. CONCLUSION ACKNOWLEDGMENT

In a nutshell, we have successfully implemented The research work is jointly supported by the Indian
satellite pose estimation using a new methodology Space Research Organization (ISRO) and Christ
which resulted in less error and improved accuracy University, Bangalore.
during pose estimation.
REFERENCES
Using AHRNet, we were able to perform landmark
regression and detect the key points in the satellite [1] J. Wang, K. Sun, T. Cheng, B. Jiang, C. Deng, ”Deep High-
Resolution Representation Learning for Visual Recognition,”
with a higher accuracy rate. Our methodology arXiv preprint arXiv:1908.07919, 2020.
resulted in better accuracy with a low error rate after [2] B. Chen, J. Cao, A. Parra and T. -J. Chin, ”Satellite Pose
training the model for just 10 epochs. Estimation with Deep Landmark Regression and Nonlinear
Pose Refinement” 2019 IEEE/CVF International Conference
on Computer Vision Workshop (ICCVW), Seoul, Korea
We proposed a framework for estimating the position (South), 2019, pp. 2816-2824, doi: 10.1109/IC-
and orientation of a single satellite in space. Our CVW.2019.00343. technologies,” Proc. 8th Int. Symp. Artif.
approach leverages the strengths of both geometric Intell., Robot. Autom. Space, 2005.
optimization for robust fitting and deep neural [3] H. Zhou, T. Zhang, and J. Jayender, ”Re-weighting and 1-
Point RANSAC-Based PnP Solution to Handle Outliers”
networks for feature extraction. Specifically, arXiv preprint arXiv:2007.08577, 2020.
AHRNet ensures accurate 2D landmark predictions [4] L. P. Cassinis, R. Fonod, and E. Gill, ”Review of the
by preserving high-resolution image representations, robustness and applicability of monocular pose estimation
while noise and outlier reduction, along with pose systems for relative navigation with an uncooperative
spacecraft” Prog. Aerosp. Sci., 2019.
refinement are achieved using the solvePnP iterator. [5] S. D’Amico, M. Benn, and J. L. Jørgensen, ”Pose estimation
This simplicity contributes to its superior overall of an uncooperative spacecraft from actual space imagery”
performance, combining both translational and Int. J. Space Sci. Eng., vol. 2, no. 2, pp. 171-189, 2014.
rotational accuracy. [6] S. Sharma and S. D’Amico, ”Pose estimation for non-
cooperative rendezvous using neural networks” in
AAS/AIAA Astrodynamics Specialist Conference, 2019.
However, relying solely on synthetic datasets to train [7] S. Sharma, J. Ventura, and S. D’Amico, ”Robust model-
the model presents challenges in domain adaptationas based monocular pose initialization for noncooperative
the model might struggle with real-world data. To spacecraft rendezvous” J. Spacecr. Rockets, vol. 55, no. 6,
pp. 1414-1429, 2018.
address this, experimental methods are being [8] A. Kendall and R. Cipolla, ”Modelling uncertainty in deep
developed to generate realistic satellite images. learning for camera relocalization” in Proc. IEEE Int. Conf.
Robot. Autom. (ICRA), 2016.
One method being explored involves importing 3D [9] A. Kendall and R. Cipolla, ”Geometric loss functions for
camera pose regression with deep learning” in Proc. IEEE
satellite models into Unreal Engine where a realistic Conf. Comput. Vis. Pattern Recog. (CVPR), 2017.
environment of Earth, the Sun, and the Moon is [10] A.Kendall, M. Grimes, and R. Cipolla, ”Posenet: A
modeled, complete with shadowing effects. The convolutional network for real-time 6-dof camera
software’s advanced rendering capabilities allow for relocalization” in Proc. IEEE Int. Conf. Comput. Vis.
(ICCV), 2015.
the creation of highly realistic images for space- [11] V. Lepetit, F. Moreno-Noguer, and P. Fua, ”EPnP: An
related applications like satellite pose accurate o(n) solution to the PnP problem” Int. J. Comput.
estimationwhich will help in debris removal and Vis., vol. 81, no. 2, pp. 155-166, 2009.
docking operations. [12] I. Melekhov, J. Ylioinas, J. Kannala, and E. Rahtu, ”Image-
based localization using hourglass networks” in Proc. IEEE
Int. Conf. Comput. Vis. (ICCV), 2017.
FUTURE WORK [13] M. Ozuysal, M. Calonder, V. Lepetit, and P. Fua,
”Fastkeypoint recognition using random ferns” IEEE Trans.
In the future, we plan to implement satellite pose Pattern Anal. Mach. Intell., vol. 32, no. 3, pp. 448-461, 2009.
[14] S. Peng, Y. Liu, Q. Huang, X. Zhou, and H. Bao, ”Pvnet:
tracking for tracking the satellite for each and every Pixel-wise voting network for 6DOF pose estimation” in
frame. While this concept is well-established for Proc. IEEE Conf. Comput. Vis. Pattern Recog. (CVPR),
objects on Earth, its application in space is still in its 2019.
early development stages and offers significant [15] T. Sattler, Q. Zhou, M. Pollefeys, and L. Leal-Taixe,
”Understanding the limitations of CNN-based absolute
opportunities for advancement. We also plan to use camera pose regression” in Proc. IEEE Conf. Comput. Vis.
dense correspondence methods which will further Pattern Recog. (CVPR), 2019.
improve the reprojection error, the translation and [16] K. Sun, B. Xiao, D. Liu, and J. Wang,” Deep high-resolution
rotation errors for estimation of the pose of the representation learning for human pose estimation”, 2019.
[17] S. Sharma, J. Ventura, and S. D’Amico,” Robust model-
satellite. Focusing on enhancing the model’s domain based monocular pose initialization for noncooperative
adaptability to handle various real-world conditions, spacecraft rendezvous” J. Spacecr. Rockets, vol. 55, no. 6,
such as unpredictable sunlight, reflections, and pp. 1414-1429, 2018.
diverse surface textures is also our future work. [18] K. Sun, B. Xiao, D. Liu, and J. Wang, ”Deep high-resolution
representation learning for human pose estimation” in Proc.
Improving adaptability in these environments would IEEE Conf. Comput. Vis. Pattern Recog. (CVPR), 2019, pp.
result in more precise pose estimation in actual space 5693-5703.
missions.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

6
 Effective Landmark Regression Using Attention Based-HRNet for Satellite Pose Estimation
[19] T. H. Park, M. Ma¨rttens, G. Le´cuyer, D. Izzo, and S. [21] S. Sharma, C. Beierle and S. D'Amico, "Pose estimation for
D’Amico, ”SPEED+: Next Generation Dataset for Spacecraft non-cooperative spacecraft rendezvous using convolutional
Pose Estimation across Domain Gap”, arXiv preprint neural networks," 2018 IEEE Aerospace Conference, Big
arXiv:2110.03101, 2021. Sky, MT, USA, 2018, pp. 1-12, doi:
[20] E. Shreyas, M. H. Sheth, and Mohana, ”3D Object Detection 10.1109/AERO.2018.8396425.
and Tracking Methods using Deep Learning for Computer [22] [22]Deevi, Sri Aditya, and Deepak Mishra. "Expeditious
Vision Applications” in Proc. IEEE RTEICT, 2021, doi: object pose estimation for autonomous robotic
10.1109/RTEICT52294.2021.9573964. grasping." International Conference on Computer Vision and
Image Processing. Cham: Springer Nature Switzerland, 2022.



Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

7
 DEVELOPMENT OF WATER RESOURCES MANAGEMENT
SCHEME FOR SEDAWGYI DAM USING WEAP MODEL
1
THET ZIN HTOO, 2YIN YIN HTWE, 3NILAR AYE
1,2,3
Department of Civil Engineering, Mandalay Technological University, Mandalay, Myanmar
E-mail: 1info.thetzinhtoo@gmail.com, 2yinyinhtwe.civil@gmail.com, 3dnilaraye@gmail.com

Abstract - This study aims to improve water resource management at the Sedawgyi Dam using an Integrated Water
Resources Management (IWRM) framework. The objectives are to promote efficient water use in hydropower, irrigation,
and domestic water supply, and to develop a comprehensive water management plan. Located near the Chaungmagyi stream,
the dam supports Mandalay City and its surrounding townships through a complex water distribution network.The Water
Evaluation and Planning (WEAP) model was used to identify significant gaps between water supply and demand. Scenario
analyses indicate that future water demand will increase substantially, highlighting the need for proactive measures such as
infrastructure upgrades, advanced irrigation techniques, and better groundwater management. Scenario 2 was found to be the
most effective for optimizing the water supply system, offering benefits like reducing groundwater extraction, expanding
domestic water supply, and sustaining downstream flow without impacting hydropower generation.The proposed water
management plan focuses on scenario-based planning and balancing economic development, social equity, and
environmental sustainability.

Keywords - IWRM, Scenario, Sedawgyi Dam, Water Demand, WEAP

I. INTRODUCTION methodology for accurately estimating water

requirements to enhance management practices and
Effective water management is crucial for boost productivity in Myanmar's central dry zone.
maintaining agricultural systems, particularly in areas The Sedawgyi Dam in Mandalay, Myanmar, is vital
with limited water supplies or increasing competition for meeting domestic, agricultural, and power
from other sectors. Integrated Water Resources generation water needs. It regulates water storage and
Management (IWRM) is essential for sustainable distribution for irrigation, flood control, and
development, as it balances economic efficiency, hydropower generation. Despite its critical role, the
social equity, and environmental dam faces challenges such as insufficient water
sustainability.[4]IWRM integrates various aspects of supply for domestic use and inefficient allocation
water management, including allocation, quality between sectors. Increasing water demand, seasonal
preservation, environmental conservation, and variability, and climate change necessitate a
stakeholder participation, ensuring the long-term comprehensive water resource management
availability of water resources through coordinated approach.[2] This study addresses these issues at
efforts across sectors. Sedawgyi Dam by utilizing WEAP to analyze current
WEAP (Water Evaluation And Planning), developed water demand patterns, assess supply capabilities, and
by the Stockholm Environment Institute (SEI), is a explore future scenarios, providing insights for
key tool for integrated water resources planning.[8] It improving water distribution and management
offers a robust framework for evaluating the complex strategies.
interactions between water supply, demand, and
environmental impacts within a river basin or II. OBJECTIVES
geographic region. WEAP is invaluable for decision-
makers, water managers, and researchers in analyzing The main objectives of the paper are as follows:
current water usage, simulating future scenarios, and  To encourage the effective use of waterfor
developing sustainable water management strategies. hydropower, irrigation and water supply sectors
By incorporating hydrological, economic, and from Sedawgyi Dam
environmental data, WEAP facilitates informed  To develop an effective water management plan
decision-making aimed at efficient water allocation, by IWRM approach
addressing climate change and population growth
challenges, and promoting water security. III. MATERIAL AND METHODS
In Myanmar's central dry zone, precise water
management is critical for economic stability and A. Study Area
food security, given agriculture's significant The Sedawgyi Dam, situated near the Chaungmagyi
contribution to GDP and employment.[3] However, stream in the northern region of the project area,
the lack of reliable water requirement data hampers serves as a pivotal structure. It forms the Mandalay
efficient resource allocation, leading to water channel to the north, dividing the Chaungmagyi into
wastage, decreased crop yields, and economic three distinct sections: the Chaungmagyi stream,
inefficiencies. This study aims to develop a robust Yenatha Canal, and Mandalay Main canal. Mandalay

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

8
 Development of Water Resources Management Scheme for Sedawgyi Dam Using WEAP Model

City area is composed with Aungmyaethazan Amarapura (AMA) townships are deliverd with
(AMTZ), Chanayethazan (CATZ), Mahaaungmyae channel network system.In order to build up a water
(MHA), Chanmyathazi (CMTZ) and Pyigyitagon resources model for the Sedawgyi Dam study region,
(PGTG) townsips. In the irrigation section, Mattayar Figure 1 demonstrates defining the study area.
(MTY), Patheingyi (PTG), Mandalay (MDY) and

Figure 1: Study Area [author creation]

B. Meteorological Data agriculture, industry, domestic use, and

A crucial phase in the WEAP modeling method is environmental needs, providing insights into how
data collecting. Data about the Sedawgyi Dam and its different allocation strategies impact water
environs were acquired as follows: availability and quality over time.
 HydrologicalData:The Department of Irrigation
and Water Resource Development (Mandalay D. Inputting demand and supply Data
Region) compiles historical data on reservoir Entering the collected data into WEAP, specifying
levels, precipitation, evaporation, and river flows parameters such as catchment characteristics, water
in the Chaungmagyi stram. demand profiles, and operational rules for the dam.
 Demand Data: The nodes (township base for
residential and area based for agricultural E. Reference Scenario
purposes) are designated to collect the current This includes establishing the current based year as
and projected water demand for household and 2020 and reference scenario timeline from 2020 to
agricultural purposes. 2050 with monthly time steps. Agricultural water
 Infrastructure Data:The Mandalay City demand for irrigated area was developed by net
Development Committee provides information irrigation requirement (NIR) with the subtracting of
on the dam's capacity, irrigation canals, water effective rainfall from the crop evapotranspiration[1].
treatment facilities, and distribution networks. NIR represents the amount of water that needs to be
applied through irrigation to meet the crop's
C. WEAP evapotranspiration (ET) needs, considering
WEAP model is an essential tool for assessing and contributions from effective rainfall and existing soil
optimizing water allocation in integrated water moisture. Also, domestic demand for all townships
resources management. It allows users to simulate are collected based on the 2014 census. All of the
scenarios that consider varying demands from demand notes and supply conditions do not change in

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

9
 Development of Water Resources Management Scheme for Sedawgyi Dam Using WEAP Model

current operations. These data prefer to use as the c) Scenario 3

reference. Based on reference scenario, three adaption In scenario 3, analysis will be carried out under this
scenarios with volume of water for hydropower options. (i)The cropping pattern for Amarapura
generation and environmental flow are fixed, are township are altered to meet the water demand,(ii)
considered inthis study. environmental flow of Chaungmagyi Stream is set
minimum of 20 m3/s in order to get continuous
a) Scenario 1 downstream flow is needed to maintain for the
To encourage the effective use of water for aquatic ecosystem of the river.
hydropower, irrigation and water supply sectors from
Sedawgyi Dam, , the performance analysis of the F. Water Resource Management
current operation supply movement will be taken out. PlanDevelopment
Analysis will be carried out under this options. (i) Based on the several scenarios, adapting the plan that
The surplus water of Sedawgyi Dam is fully will avoid and pre limitations for challenges of water
connected to supply Mandalay City and all the shortage for the future. In addition, consideration of
groundwater sources that are used for Mandalay City new finding for another phrase of supply source
are fully closed. (ii) Future water demand due to which can help for emergency supply and subsidies.
population growth is taking the average growth rate
of 2.03%. (iii) Based on landuse/landcover change IV. Results and Discussion
analysis, decrease of agricultural land used
percentages is considered as urbanization is growth. A. Modeling in WEAP
After establishing the denoted area of demand sites,
b) Scenario 2 along with their correspondance demands, supply
In scenario 2, analysis will be carried out under this nodes are also assigned for the reservoir, groundwater
options. (i) The irrigation network is extend to reach sources, and river lines. Transmission links are used
Mattayar for domestic purpose, and population to connect these nodes; their size and priority should
growth of Mattayar is considered. (ii) Due to climate be determined as relevance of each demand
change effect, rapid change of groundwater level location.Figure 2 domostratesschematic diagram of
decline rate is considered as 30% of decrease over the water distributed areaand its components in WEAP
four decade (>0.5 m/year), thus half of groundwater such as demand sites, supply sites, transmittion links
pumping is reduced i.e, GW 1, 4, 7, 8, 10, 12, 13 and and return flow.
14 are closed.

Figure 2: Schematic Diagram of WaterDistributed Area [author creation]

B. Reference Scenario township requires around 27 Mcm during the dry

Analysis using the WEAP application revealed that season and less than 10 Mcm in the monsoon season,
the highest domestic water demand in Mandalay City with Patheingyi (PTG) also having notable but lower
is in the Chanayethazan (CMTZ) and needs. The agricultural water demands are fully met,
Aungmyaethazan (AMTZ) townships, with suggesting a surplus or adequate provision, while
Pyigyitagon (PGTG) and Maharaungmyae (MHA) domestic water supply remains critically inadequate,
showing similar patterns, totaling approximately 1.6 with coverage consistently between 10% and 35%
million cubic meters (Mcm). However, the existing across various townships.Hydropower demand is The
water supply system meets only 10% of this demand. water demand for all notes and supply requirement of
For agricultural water demand, Mattayar (MTY) reference scenario are shown in Figure 3.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

10
 Development of Water Resources Management Scheme for Sedawgyi Dam Using WEAP Model

(c) unmet demand

(a) demand

(d) supply coverage

Figure 4: Comparison of Water Demand and Supply of
(b) unmet demand Difference Scenarios

In Scenerio 1, domestic and argriculture sectors faced

ummet demand as all tube well are closed. The total
water requirement is 140Mcm while the reference
scenario projected a demand below 80 Mcm,
considering population growth and resource
constraints.
In scenario 2, domestic demand for Mandalay City,
almost 2.3 Mcm of water is needed to deliver for the
future period. Agricultural water demand, in fact,
(c) supply coverage remain the same with scenario 1. This scenario
Figure 3: Water Demand and Supply of Reference Scenario drametically met the supply water demand in
domestic sector. As the domestic water supply is
C. Scenario Analysis applied perior node, the unmet demand for
After analysis of reference scenarios, three adaption agricultural demand occurred suddenly change to
of scenerios are analysis and the results are evaluated 0.25 Mcm in Amarapura township (December and
in Figure 4. Jannuary).
Domestic demand for Scenario 3, alternating
cropping pattern change in Amarapura, will sharply
increase to 2.0 Mcm in 2050. In agricultural demand
supply, unmet demand of Amarapura can probability
decrease due to the alternative crop pattern reach to
the value of 0.6 Mcm. However, scenario 3 forecasts
a dramatic increase to 140 Mcm in 2050. Conversely,
agricultural water demand shows a slight increase
across all scenarios, reflecting a consistent pattern of
irrigation requirements. Despite persistent unmet
(a) domestic water demand domestic demand, it is clear that the supply capacity
of the Sedawgyi Dam remains generally stable
against future demand challenges, with exceptions
noted in January and December.

D. Develop Water Resource Management Plan

By considering the various scenarios for the dam
operation, it should develop the water management
plan for reservoir operation in order to balance for all
sectors such as hydropower generation, irrigation,
domestic supply and e-flow.
(b) agricultural water demand

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

11
 Development of Water Resources Management Scheme for Sedawgyi Dam Using WEAP Model

If the operation of the dam regulate in accordance 1. Scenario 3 is basically depend on the scenario 2
with scenario 1, the following facts should be in domestic supply and environmental flow
implemented. consideration. Thus, domestic water supply
1. Intake infrastructures that link with MDY canal should make the same with scenario 2
to BPS4 should extend to get adequate amount of suggestions.
water for domestic supply. 2. For the agricultural section, the requirement
2. The linkages between BPS 4 and Moat, and BPS should fulfill with the direct connection from the
4 to BPS 1 should construct for the sufficient Ayeyarwaddy river.
delivering supply system. Because BPS 4 is the 3. And then, the new channel should develop from
most essential station in Mandalay water supply Dohtawaddy river in order to supply the area
system, there are four stations that directly link where lack of water from the Sedawgyi source
with BPS 1. and domestic supply usages.
3. In order to get PGTG and MHA townships, there
should be developed new supply system to BPS V. CONCLUSION
4.
4. It is needed to deliver water from MDY canal at In conclusion, effective water resource management
least 11.3 Mcm per month which is the for the Sedawgyi Dam is crucial in the face of
maximum supply from Sedawgyi Dam. increasing demands and climate uncertainties. The
5. To meet the required volume of water, it should application of the Water Evaluation and Planning
prepare provisional measures like river pumping (WEAP) model has highlighted critical gaps between
and rainwater harvesting in AMTZ and MOAT water supply and demand. Scenario analyses project a
which need monthly 0.1 Mcm and 0.73 Mcm significant rise in future water demand, emphasizing
only in December and January. the need for proactive strategies such as infrastructure
6. In the agricultural sector only 0.29 Mcm and upgrades, advanced irrigation techniques, and
0.23 Mcm will be taking from other sources of sustainable aquifer management. Adopting Integrated
water such as Sedawlay weir and Dohtawaddy Water Resources Management (IWRM) principles is
river with river pumping system (only in essential to balance economic development, social
December). equity, and environmental sustainability.By
incorporating stakeholder input and scenario-based
If the scenario 2 has been adopted for water supply planning, the proposed management plan seeks to
system, the followings should be considered and optimize water allocation, improve water use
planned: efficiency, and ensure equitable distribution across
1. The irrigated water from the Sedawgyi Dam sectors. Among the scenario results, Scenario 2 has
must supply about 55 Mcm per month through been identified as the optimal operating procedure for
with MDY canal. the Sedawgyi water supply system. This scenario is
2. And then discharged from turbines should be beneficial in minimizing groundwater withdrawals,
delivered directly to the Chaungmagyi stream. expanding the coverage of domestic water supply,
The amount discharge should maintain 55 Mcm and maintaining sufficient downstream flow in the
which is adequate for the downstream of Chaungmagyi stream, all without affecting the
Chaungmagyi stream ecosystem. hydropower generation schedule.
3. Domestic water supply should make the network
system as same as scenario 1. But the amount ACKNOWLEDGMENTS
delivering water can change into 54 Mcm per
month, including surplus from the agricultural The author would like to thank Dr. Cho Cho Thin
and diverted from the Yenathar weir. Kyi, the former professor at Mandalay Technological
4. As the purpose of water planning is to deliver the University's Department of Civil Engineering, for her
adequate water to the system, the required important supervision, editing, and assistance. The
volume of water (20 Mcm) for domestic supply Stockholm Environment Institute's free license for the
should take from BPS 1, 3, and 5. These stations WEAP program, which was essential for this
must operate fully in December and January. investigation, is also appreciated by the author. Extra
5. In agriculatral sector, Amarapuara area can take gratitude is extended to the officers from Mandalay
water from the Sedawlay weir, and some area City Development Committee for their indispensable
should be supplied from water collecting ponds. information and assistance. Lastly, a sincere thank
6. It is suggested that water collecting ponds should you to everyone who supported and encouraged me in
create in the Amarapuara area one pond per this endeavor.
hectar because this township needs only 0.25
Mcm in December and 0.08 Mcm in January. REFERENCES

If the scenario 3 must be adopted for this system, it [1] Thet Zin Htoo & Yin Yin Htwe ―Assessment on Total Crop
Water Requiremnt for Sedawgyi Irrrigated Area‖,
could follow up with the followings: International Conference on Civil and Environmental,

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

12
 Development of Water Resources Management Scheme for Sedawgyi Dam Using WEAP Model
Thailand, (2024). [9] Sieber, J., & Purkey, D. WEAP (Water Evaluation and
https://drive.google.com/file/d/1_R0Ax2YJT5QE5utO6Gn4o Planning System) User Guide. Stockholm Environment
1dRChm1zlfD/view?usp=sharing Institute. Available at: [WEAP User
[2] Thet Zin Htoo, Yin Yin Htwe & Cho Cho Thin Kyi, Guide]https://www.weap21.org/downloads/WEAP_
―Assessment on Water Resource Management for Sedawgyi User_Guide.pdf. (2011).
Dam: A WEAP Analysis Approach‖, Vol. 13 No.5 (2024): [10] Joyce, B., Mehta, V. K., Purkey, D., Dale, L., Hanemann, M.,
The Indonesian Journal of Computer & Vicuna, S. ―Modifying Agricultural Water Management to
Science.http://ijcs.net/ijcs/index.php/ijcs/article/view/ 4272 Adapt to Climate Change in California's Central Valley.‖
[3] Loucks, D. P., & van Beek, E. ―Water Resource Systems Climatic Change, 109(1), 299-316. (2011).
Planning and Management: An Introduction to Methods, [11] Rani, D., & Moreira, M. M. ―Simulation–Optimization
Models, and Applications‖. Springer, (2017). Modeling: A Survey and Potential Application in Reservoir
[4] Kaung Myat, Nilar Aye: Assessment of Current and Future Systems Operation.‖ Water Resources Management, 24(6),
Domestic Water Allocation for Mandalay City, The 1107-1138. (2010).
proceedings of the International Conference on Disaster [12] Liu, Y., Gupta, H., Springer, E., & Wagener, T. ―Linking
Management, Civil and Architectural Engineering (DMCAE Science with Environmental Decision Making: Experiences
- 16), Pattaya, Thailand, (2016). from an Integrated Modeling Approach to Supporting
[5] United Nations. Sustainable Development Goal 6: Ensure Sustainable Water Resources Management.‖ Environmental
availability and sustainable management of water and Modelling & Software, 23(7), 846-858. (2008).
sanitation for all. Retrieved from [13] Gassman, P. W., Reyes, M. R., Green, C. H., & Arnold, J. G.
https://sdgs.un.org/goals/goal6. (2015). ―The Soil and Water Assessment Tool: Historical
[6] IPCC. (2014). Climate Change: Impacts, Adaptation, and Development, Applications, and Future Research Directions.‖
Vulnerability. Part A: Global and Sectoral Aspects. Transactions of the ASABE, 50(4), 1211-1250. (2007).
https://www.ipcc.ch/report/ar5/wg2/ (2014). [14] Yates, D., Sieber, J., Purkey, D., & Huber-Lee, A.
[7] Srinivasan, V., Seto, K. C., Emerson, R., & Gorelick, S. M. WEAP21—A Demand-, Priority-, and Preference-Driven
―The Impact of Urbanization on Water Vulnerability: A Water Planning Model: Part 1: Model Characteristics. Water
Coupled Human–Environment System Approach for International, 30(4), 487-500. (2005).
Chennai, India.‖ Global Environmental Change, 23(1), 229- [15] Kashyap, A. ―Water Governance: Learning by Developing
239. (2013). Adaptive Capacity to Incorporate Climate Variability and
[8] Gohari, A., Eslamian, S., Abedi-Koupai, J., Massah Bavani, Change.‖ Water Science and Technology, 49(7), 141-146.
A. R., Wang, D., & Madani, K. ―Climate Change Impacts on (2004).
Crop Production in Iran's Zayandeh-Rud River Basin.‖
Science of The Total Environment, 442, 405-419. (2013).



Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

13
 ANALYSIS OF WATER ALLOCATION AND DEMAND
MANAGEMENT IN KATHA BASIN
1
WIN LWIN TUN, 2NILAR AYE, 3CHO CHO THIN KYI
1,2,3
Department of Civil Engineering, Mandalay Technological University, Myanmar
E-mail: 1kolwinwlh1994@gmail.com, 2dnilaraye@gmail.com, 3ccthinkyi@gmail.com

Abstract - The Katha Basin, part of the Upper Ayeyarwaddy River Basin in Myanmar, faces significant water resource
challenges due to rising water demand, population growth, and climate variability. This study assesses and simulates the
spatial and temporal dynamics of water demand in the Katha Basin, utilizing the Water Evaluation and Planning (WEAP)
model to evaluate unmet water needs across domestic and agricultural sectors under various scenarios. The findings reveal
consistent unmet water demands, particularly in Bhamo, Myitkyinā, and Puta-O, indicating a need for improved water
management strategies. Scenario analyses, including enhanced Demand Management Strategies (DMS) and Optimal
Cropping Patterns (OCP), demonstrate potential reductions in unmet demand by implementing water-saving techniques and
adapting crop patterns to mitigate climate impacts. The study underscores the importance of equitable water resource
allocation, efficient management practices, and stakeholder engagement in addressing the basin's growing water scarcity,
ensuring sustainable water availability for both human and environmental needs by 2100.

Keywords - Katha Basin, Water Demand, WEAP, Demand Management Strategies, Optimal Cropping Patterns

I. INTRODUCTION system.The increase in water demand and the lack of

demand management measures will lead to a real
The fresh water resource is finite in time and location. clean water scarcity in the basin.
The resource has been facing challenges as a The primary objective of this paper is to assess and
consequence of unprecedented change in settlement, simulate water demands in the Katha Basin,
water supply and utilization. In the face of growing providing a comprehensive understanding of the
water demand, natural river flow is being challenged spatial and temporal dynamics of water demand. This
by land ownership, economic growth and advances in knowledge is crucial for the effective implementation
technology. Watershed scale hydrological processes of water allocation policies by the Water Resources
are also being affected by climate and land use Management Authority and other governmental
variability. The major challenge of climate change is bodies. The scenarios will serve as a foundation for
its impact on regional and local freshwater constructive dialogue among stakeholders, facilitating
availability and distribution thereby disrupting discussions on various water resource allocation
livelihood and ecosystems. options and the potential trade-offs involved.
Water demand is not only the projected water
requirement of individual sectors but also the change II. METHODOLOGY
in behavior of consumption while adapting to
scarcity. Fair and efficient distribution thrives to use A. Study Area
robust techniques to estimate the water availability The Katha Basin is situated within the upper
and demand, setup evaluation tools and feedback Ayeyarwaddy (Irrawaddy) River basin in Myanmar.
mechanism. In addition to the quantity and This area lies approximately between 24° to 25°
distribution of available water, allocation procedures North latitude and 96° to 97° East longitude, it is
need to attempt to address the relationship between surrounded by the western foothills of the Shan
availablewater and biological indicators of an Plateau in the east and the central plains in the west.
adequate environmental flow. The Katha basin area is 132662 km2.The Katha basin
These challenges are felt at varying levels by heavily relies on the Southwest Monsoon, which
different watersheds. The Katha Basin (Upper brings the majority of annual rainfall between June
Ayeyarwady River Basin) is one of the troubled and October. The basin experiences 155 rainy days
basins that has experienced relatively minor and 128.8 mm of precipitation annually. Elevation of
hydrological changes from water resource the basin varies from 100 metres above sea level
development and extraction. The high-water use for (masl) in the south to 5,775 masl in the mountainous
industrial, domestic and agricultural sectors in river region in the north. The middle of the basin is
basin due to the lack of hydrological knowledge, comprised of plateaus (∼500 masl) and floodplains.
unimplemented water rights and ignorance of the
environmental water demands has decreased water B. Modelling Using WEAP
quantity. Moreover, the need to satisfy water WEAP is a user-friendly software tool that is
demands for both economic and social development developed by the Stockholm Environmental institute
in the basin has created conflicts due to the lack of (SEI) to assist decision makers in managing water
equity in the allocation method and the permit demand, water availability, waste generation and

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

14
 Analysis of Water Allocation and Demand Management in Katha Basin

water costs and to evaluate water development and WEAP has an integrated approach to simulate both
management options [1]. natural and engineering components such as
reservoirs, groundwater discharge and water demand
WEAP incorporates water supply in the context of and supply, which can give water planner a more
demand-side management, and water quality and comprehensive view of the broad range of factors that
ecosystem preservation and protection into a practical must be considered in managing water resources for
tool for water resources planning and policy analysis present and future uses [2]. It can analyse a diverse
[2]. The model places demand-side issues such as range of issues such as climate variability, watershed
water use patterns, equipment efficiencies, reuse conditions, anticipated demands, ecosystem needs,
strategies, costs, and water allocation schemes on an available infrastructures and operational objectives in
equal footing with supply-side themes such as stream a transparent manner [3].Figure 1 illustrates the
flow, groundwater resources, reservoirs and water schematic of the WEAP model set for this study.
transfers [3].

Figure 1. The schematic of the WEAP model for Katha Basin

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

15
 Analysis of Water Allocation and Demand Management in Katha Basin

C. Water Demands in Katha Basin 2014 cencus. Population estimates are derived from
the 2014 country census report and are projected to
(1) Domestic Demand the year 2100 by incorporating the population growth
Ayeyarwady river provides the life to all activities trend observed during the baseline period from 2000
and livelihoods within the basin. The population of to 2014.Figure 2. shows historical data and projected
the basin has grown from less than 738,600 people in population growth trends.
2000 to almost 1,419,000 people, according to the

Key Assumptions (monthly)

2,200,000 D1
2,000,000 D2
1,800,000 D3
1,600,000 D4
1,400,000
cap

1,200,000
1,000,000
800,000
600,000
400,000
200,000
0
Jan Oct Dec Jan Feb Apr May Jun Jul Sep Oct Nov Jan Feb Mar Apr Jun Jul Aug Oct Nov Dec Jan Mar Apr
2000 2003 2007 2012 2016 2020 2024 2028 2032 2036 2040 2044 2049 2053 2057 2061 2065 2069 2073 2077 2081 2085 2090 2094 2098

Figure 2. Observed (2000-2014) and Projected (2015-2100) Population

The population within the Katha Basin shows an increasing trend more specifically within the periods 2014 and
2100. Table 1. presents the monthly unmet domestic demand (million cubic meter-Mcm) for SSP585 and
SSP245 scenarios, from 2024 to 2100.

Branch Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Katha
0 0 0 0 0 0 0 0 0 0 0 0
(D1)
Bhamo
4.01 3.65 4.01 3.88 4.01 3.88 4.01 4.01 3.88 4.01 3.88 4.01
(D2)
Myitkyinā (D3) 0.72 0.66 0.72 0.7 0.72 0.7 0.72 0.72 0.7 0.72 0.7 0.72
Puta-O
1.18 1.08 1.18 1.14 1.18 1.14 1.18 1.18 1.14 1.18 1.14 1.18
(D4)
Sum 5.91 5.38 5.91 5.72 5.91 5.72 5.91 5.91 5.72 5.91 5.72 5.91
Table 1. Unmet Monthly Water Demand (Mcm)

(2)Agricultural Demand
Irrigation is practiced in various parts of the basin at both small and large scales. Land used for agriculture are
displayed in Table 2.

Catchment (ha) Rain fed Cropland (ha) Irrigation Cropland (ha)

Katha (7722929) 402590 (5.2%) 75472 (1%)
Bhamo (707729) 70048 (9.9%) 22054 (3.1%)
Myitkyinā (2504166) 22871 (0.9%) 8716 (0.3%)
Puta-O (2331386) 14797 (0.6%) 2891 (0.1%)
Table 2. Land used information in Katha Basin

Branch Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Katha 0 0 0 0 0 0 0 0 0 0 0 0
Bhamo 0.5 1.5 6.97 4.79 1.18 0 0 0 0 0 0 0
Myitkyinā 6.9 13 21 22 15.3 3.3 8.8 0.4 0.2 0 0.4 2.3
Puta-O 1.6 4.02 8.06 9.20 6.36 1.2 0.3 0.2 0.2 0.2 0.2 0.5
Table 3. Unmet Demand for Agricultural (Mcm)

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

16
 Analysis of Water Allocation and Demand Management in Katha Basin

The computation of the unmet water demand are based on subbasin, Table 3. presents water requirements for
agriculturalin SSP45 scnario while Table 4 presents water demands at SSP585 scnario.

Branch Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Katha 0 0 0 0 0 0 0 0 0 0 0 0
Bhamo 1.26 3.53 6.49 5.23 0.32 0.62 0 0 0 0 0.07 0.04
Myitkyinā 14.3 19.6 23.1 25.1 16.8 4.7 13.0 1.61 0.72 1.09 3.18 8
Puta-O 3.73 6.35 8.71 9.6 6.4 1.36 0.23 0.23 0.23 0.23 0.35 1.53
Table 4. Unmet Demand for Agricultural (Mcm)

Recent quantitative studies showed that the available supply is sufficient to meet the demand in this region.
water cannot meet the competing sectoral demands Bhamo consistently experiences the highest unmet
due to the population growth and the crop farming water demand among the four branches, fluctuating
system [4]. To achieve a balance between human between 3.65 Mcm and 4.01 Mcm each month. The
needs and environmental sustainability, it is essential demand remains relatively stable throughout the year,
to develop mechanisms that ensure the equitable with slight variations. Myitkyinā has an unmet water
distribution of resources. This can only be demand ranging from 0.66 Mcm to 0.72 Mcm
accomplished if people recognize the intrinsic value monthly. Puta-O experiences unmet water demand
of these resources and acknowledge the trade-offs between 1.08 Mcm and 1.18 Mcm each month. The
among different users. Such initiatives are crucial for unmet demand in Puta-O is moderate compared to
promoting equitable resource management, Bhamo and Myitkyinā.
prioritizing sustainable ecological and social benefits.
By enhancing water use efficiency, improving When examining the total unmet water demand
economic returns, and mitigating hydrological across all branches, it ranges from 5.38 Mcm to 5.91
fluctuations, these efforts contribute to long-term Mcm throughout the year. The data shows that the
sustainability. total unmet demand remains fairly stable across the
months, with slight increases and decreases, which
D. The Development of the Scenarios may correlate with seasonal patterns or regional water
Scenario can be defined as a set of assumptions or usage practices. The consistent nature of these
alternative mechanisms (policies, pricing and demand shortfalls highlights the need for targeted
management strategies) that form the basis for the interventions to improve water resourcemanagement
projection. Scenarios are self-consistent story-lines of and supply across the affected regions, particularly in
how a future system might evolve over time in a Bhamo, Myitkyinā, and Puta-O.
specific socio-economic condition and under a
specific set of policy and technology conditions B. Unmet Agricultural Demand
[2].Scenario analysis is an effective approach for Across all branches, water requirements are generally
answering "what if" questions, allowing for the higher in the SSP585 scenario compared to SSP245.
investigation and testing of many options within In SSP245, water demand is more concentrated in the
predetermined parameters. The reference scenario is early part of the year (January to May), while in
based on the features of the current situation and SSP585, the demand is more evenly distributed
serves as a benchmark for understanding existing across the year, with increased requirements
trends. Alternative scenarios are created as variants extending into late winter and early spring. Myitkyinā
on this baseline, each aiming to achieve the primary consistently has the highest water demand in both
goal. This method improves the ability to predict and scenarios, reflecting the potential for intensive
adjust to possible outcomes. agricultural activities in this region. Bhamo and Puta-
Based on the reference scenario, two scenarios were O also show increased demand under SSP585, though
analysed to project different demand management not as pronounced as in Myitkyinā.
strategies within the basin: 1) Enhanced the optimal
cropping pattern and 2) increased DMS by utilization The comparison between SSP245 and SSP585
of measures such as; tiered water pricing, water highlights that agricultural water requirements are
efficient appliances, and monitoring increase. expected to rise under SSP585, with more consistent
and higher demands throughout the year. This
III. RESULTS AND DISCUSSION indicates a need for improved water management
strategies to accommodate these potential increases,
A. Unmet Domestic Demand particularly in regions like Myitkyinā and Puta-O,
Katha exhibits a consistent unmet water demand of 0 where the demand is significantly higher.
Mcm throughout the year, indicating that the water

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

17
 Analysis of Water Allocation and Demand Management in Katha Basin

C. Scenarios February and April experienced a lower demand

shortfall of 1.64 Mcm and 1.75 Mcm respectively
(1) Demand Management Strategies DMS within the Bhamo Catchment while in mostother
Scenario months, the unmet demand is stable at 1.8
This scenario simulated the impact of enhancing the Mcm.Similar to Bhamo, Myitkyināexperiences
Demand Management Strategies on the reference slightly lower unmet demand in February (0.36 Mcm)
scenario.The following are the key assumptions for and April (0.38 Mcm), but otherwise, the unmet
scenario: demand remains steady at 0.4 Mcm for most of the
1. Increased penetration of water availability within year. Demand Site Puta-O experiences consistent
the basin unmet demand similar to Myitkyinā and February has
2. 45% of the population have alternative water the lowest unmet demand (0.59 Mcm), while the rest
supply sources such as harvested rain water and of the year, the unmet demand hovers around 0.63 to
springs and wells. 0.65 Mcm. Figure 3. shows unmet domestic
3. Water pricing (based on a block tiered format). demandunder DMS scenario.
On the other hand, the annual unmet demand analysis
Therefore, in this scenario only one item was between the reference scenario and the increased
simulated; the effect of water saving techniques in the DSM scenario illustrates a net reduction of the total
Katha Basin. unmet demand during the simulation periods. The
Water demand shows a drastic decrease compared to unmet demand decreases from 127 Mcm to 61 Mcm
the reference scenario. On a monthly average, in 2100.

Unmet Demand
Scenario: Scenario-2, Monthly Average

1.80 D1
D2
1.70
D3
1.60 D4

1.50

1.40

1.30

1.20
Million Cubic Meter

1.10

1.00

0.90

0.80

0.70

0.60

0.50

0.40

0.30

0.20

0.10

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 3. Unmet Domestic Demand under DMS Scenario

(2)Optimal Cropping PatternScenario climate change scenarios[5].Factors considered to

Sustainable production of current croppingsystems determine cropping pattern are;
during the current climate change scenario is under
threat due to both climatic variability and shortage of a) Climate: Prevailing climatic conditions,
resources for agricultural crops. There is an urgent including temperature, rainfall, and humidity,
need to improve crop production and mitigate the guide the selection of crops suitable for a region.
negative impacts of climate variability on current b) Water Availability: The presence of water
cropping systems, in order to fulfil the increasing resources, whether through irrigation facilities or
demands of ever-increasing world population. It is the rainfall patterns, influences crop choices and
need of the hour to adopt climate-resilient water requirements.
technologies and strategies for sustainable production c) Crop Suitability and Yield Potential: The
of cropping systems under changing environmental adaptability of crops to local conditions and their
conditions. These adaptation technologies and potential yield in a region are crucial factors in
strategies can increase the production of current selecting cropping patterns.
cropping systems and also mitigate the negative
impacts of climate change scenarios on the soil, The crops substitute in irrigation croplands are
water, and environmental quality.This studyfocuses sugarcane, maize, banana, groundnut. Figure 4.and
on the adaptation technologies and strategies that can Figure 5. presentthe water requirements for
reducethe crop water requirement under projected agriculturalin projected climate change scenarios.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

18
 Analysis of Water Allocation and Demand Management in Katha Basin
Unmet Demand
Scenario: Scenario-2, Monthly Average

12.0 Bhamo
Katha
11.5
Myitkyina
11.0 PutaO
10.5
10.0
9.5
9.0
8.5
8.0
7.5

Million Cubic Meter

7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 4. Unmet Agricultural Demand under SSP245 Scenario

Unmet Demand
Scenario: Scenario-2, Monthly Average
13.0 Bhamo
12.5 Katha
12.0 Myitkyina
11.5 PutaO
11.0
10.5
10.0
9.5
9.0
8.5
8.0
Million Cubic Meter

7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 5. Unmet Agricultural Demand under SSP585 Scenario

Across both scenarios, unmet demand peaks in the Myitkyinā, and Puta-O, which have significant unmet
spring (March-May), which is likely due to higher demand under the reference scenario, see a reduction
agricultural water requirements during this period. in demand shortfall under the DMS scenario,
This trend is especially evident in Myitkyinā and particularly in the months of February and April. This
Puta-O.In the reference scenario, in the same period, scenario also suggests a reduction in total unmet
96 Mcm is utilized for irrigation in SSP245. demand from 127 Mcm to 61 Mcm by 2100,
However, 35 Mcm is utilized in the OCP scenario, highlighting the effectiveness of water-saving
which presents a reduction of 61 Mcm of water that is interventions.
saved through the enhanced crop pattern. In SSP585, The Optimal Cropping Pattern (OCP) scenario
the irrigation water requirement declines from 102 to demonstrates that adjusting cropping patterns can
32 Mcm and detains 70 Mcm water consumption. significantly decrease water consumption. By
The consequence of enhanced crop rotation implies optimizing crop choices based on climate, water
an overall decrease in water consumption of 101Mcm availability, and crop suitability, the study projects a
(SSP245) and 248Mcm (SSP585) in 2100. reduction in irrigation water requirements, saving up
to 101 Mcm (SSP245) and 248 Mcm (SSP585) by
IV. CONCLUSION 2100. These findings underscore the potential of
climate-resilient agricultural practices in mitigating
The analysis of unmet domestic and agricultural water scarcity under changing environmental
water demand across different scenarios reveals conditions.
critical insights into the water resource challenges The results of this study indicate that without
and potential management strategies within the intervention, regions such as Bhamo, Myitkyinā, and
studied regions. The total unmet domestic demand Puta-O will continue to face significant water
across all branches is relatively stable, fluctuating shortages, particularly for agricultural purposes.
between 5.38 Mcm and 5.91 Mcm monthly, However, the adoption of effective demand
highlighting a consistent water deficit that requires management strategies and optimal cropping patterns
intervention, particularly in Bhamo, Myitkyinā, and can mitigate these challenges. These strategies not
Puta-O. Agricultural water demand is significantly only reduce unmet water demand but also promote
higher under the SSP585 scenario compared to sustainable water resource management, crucial for
SSP245, with the demand in SSP585 being more meeting the future water needs of both domestic and
evenly distributed throughout the year. agricultural sectors. The findings emphasize the
The Demand Management Strategies (DMS) scenario importance of proactive planning and implementation
shows that enhanced water-saving techniques can of water-saving technologies to ensure water security
substantially reduce unmet water demand. Bhamo,

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

19
 Analysis of Water Allocation and Demand Management in Katha Basin

in the face of growing demand and climate Water Demand in the Katha Basin Using Water Evaluation
and Planning (WEAP) Model”, Vol. 13 No.5 (2024): The
variability. Indonesian Journal of Computer
Science.http://ijcs.net/ijcs/index.php/ijcs/article/view/4272
REFERENCES [5] Abdul Ghaffar, Muhammad Habib Ur Rahman, “Adaptations
in Cropping System and Pattern for Sustainable Crops
[1] Van Loon, A.; Droogers, P. Water Evaluation and Planning Production under Climate Change Scenarios”, Improvement
System Kitui – Kenya, 2006. of Plant Production in the Era of Climate Change, 1st Edition,
[2] Sieber, J. M Purkey, D. Water Evaluation And Planning 2022, CRC Press, ISBN 9781003286417.
System (WEAP), User Guide. Stockholm Environment [6] Ghimire, U.; Babel, M.S.; Shrestha, S.; Srinivasan, G. A
Institute, U. S. Center 11 Curtis Avenue 2011. multi-temporal analysis of streamflow using multiple CMIP5
[3] Yates, D.; Sieber, J.; Purkey, D. WEAP21 – A Demand-, GCMs in the Upper Ayerawaddy Basin, Myanmar. Clim.
Priority-, and Preference-Driven Water Planning Model. Chang. 2019, 155, 59–79.
International Water Resources Association. International [7] Mounir, Z.M.; Ma, C.M.; Amadou, I. Application of Water
Water Resources Association Water International, 2005, 5 Evaluation and Planning (WEAP): A Model to Assess Future
(4), 487–500. Water Demands in the Niger River (In Niger Republic).
[4] Win LwinTun, Cho Cho Thin Kyi & Yin YinHtwe, Math. Model. Methods Appl. Sci. 2011, 5, 38–49.
“Assessment of Climate Change Impacts on Stream Flow and



Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

20
 ANALYZING COGNITIVE DISTORTIONS AND BEHAVIORAL
PATTERNS IN CHILDREN WITH LEARNING DISABILITIES
SUSHMITA SINGH

PhD Scholar, Department of Psychology, RTMNU University, India, MA Psychology, B.Ed

E-mail: ssushmita098@gmail.com

Abstract - The sample is based on a comparative investigation of male and female children's behavioural features and
cognitive distortions. This study shows how the viewpoints of children without learning difficulties and those with specific
learning disabilities differ. Their cognitive processes and behavioural features when compared to average children. A sample
of 100 students was collected, of which 50 were female and 50 were male. Several tests were used, with a two-way ANOVA
serving as the research design. Anova was employed as the research approach. The findings revealed that certain learning-
disabled children had behavioural problems, and that female children, particularly learning-disabled children, had greater
behavioural problems than male children. Here are some ideas for coping with and motivating pupils with learning
disabilities.

I. INTRODUCTION Individuals with dyslexia are assumed to have this

learning style from birth. Impairment does not
Distortion Cognitive functions, such as thinking, indicate a faulty brain. Teaching reading and spelling
reasoning, and memory, enable individuals to requires a distinct strategy.
comprehend and perceive information. It is a Everyone is affected with dyslexia differently.
procedure for information processing. Learning
disabilities can manifest itself in a variety of ways, The behavioral profile of individuals with learning
affecting a learner in many facets of their life. disabilities often includes indicators such as:
Concern is most in areas involving fundamental  School avoidance or reluctance to attend classes.
academic ability. Four stages of cognition and  Self-deprecating statements reflecting a lack of
cognitive processes are discussed: 1. Input.2. confidence in abilities (e.g., "I'm not smart
Integration.3.Production.4. enough," "I can’t do this").
 Evasion of homework tasks or assignments.
Remarks Learning/cognitive difficulties Cognitive  Perception of assignments as excessively
ability involves deciphering visual and aural stimuli, challenging or unachievable.
as well as interpreting nonverbal cues and body  Attribution of poor academic performance to
language. external factors, such as teacher bias.
 Reluctance to share completed homework or
Individuals suffering with these difficulties may find assignments with parents.
it difficult to learn new skills. They may draw  Resistance to in-class tasks or refusal to complete
conclusions from different situations and
assigned work.
communicate them verbally or in writing.
 Physical manifestations of distress (e.g.,
headaches, stomachaches, heightened anxiety, or
 Specific Learning Disability refers to a distortion
depressive symptoms).
in underlying psychological processes related to
 Noncompliance with classroom rules, often as a
written or spoken language comprehension.
means of escaping academic tasks or
 Dyslexia is a reading and language impairment.
responsibilities.
 Dysgraphia: difficulty writing or forming letters.  Avoidance of verbal engagement to sidestep
 Dyscalculia: arithmetic problem potential conflict (e.g., reluctance to discuss
 .Dyspraxia: motor skill difficulty. areas of difficulty like math tests).
 Class absenteeism.
II. ANALYSING THE ESSENTIALS OF  Engaging in aggressive behaviors, including peer
DYSLEXIA bullying.
Here are the essential points of dyslexia: People with
These behaviors can be early indicators of learning
dyslexia are assumed to have this learning style from
difficulties and may require targeted support to
birth. Dyslexia does not indicate a faulty brain.
address underlying challenges and improve adaptive
Teaching reading and spelling requires a distinct
coping strategies.
strategy. Everyone is affected with dyslexia differ The study aims to achieve the following objectives:
Here are the essential points of dyslexia:

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

21
 Analyzing Cognitive Distortions and Behavioral Patterns in Children With Learning Disabilities

 To measure cognitive abnormalities and adolescents with specific learning disabilities and
symptomatology among male and female their non-disabled peers.
adolescents.  To examine differences in competency levels
 To differentiate between adolescents with between adolescents with specific learning
specific learning disabilities and those without, disabilities and typically developing adolescents.
focusing on tendencies toward risk-  To identify distinctions on syndrome scales
preoccupation, self-criticism, feelings of between adolescents with specific learning
helplessness, and hopelessness. disabilities and their non-disabled counterparts.
 To assess variations in internalizing and These objectives aim to provide insights into the
externalizing behavior problems between cognitive and behavioral dimensions of specific
learning disabilities relative to typical developmental
patterns.

III. RESEARCH METHOD

IV. RESEARCH DESIGN helplessness, hopelessness, and preoccupation with

danger—and to assess distinctions within eight
This study employs a non-experimental, ex post facto behavioral competency areas: somatic complaints,
design. A 2x2 factorial design was utilized to social problems, rule-breaking behavior, attention
investigate the independent variables across their issues, anxiety, depression, and withdrawal.
levels, enabling a structured analysis of the variables' Additionally, this study evaluated the influence of
interactions and effects. gender and disability status (Normal/LD) on
cognitive distortions among participants.
V. RESULTS AND ANALYSIS
The research sought to examine the effects of gender
A sample of 100 students was collected, evenly and disability types on behavior competency factors
divided into 50 female and 50 male participants. and cognitive distortions, distinguishing between
Several psychological and behavioral assessments normal and learning-disabled participants based on
were conducted, utilizing a two-way ANOVA as the both cognitive distortion domains and behavioral
research design and primary analytical approach. The competencies. Participants were organized into four
aim of this study was to conduct a comparative subgroups—Normal Male, Normal Female, LD Male,
analysis of male and female participants with and and LD Female—to comprehensively analyze
without learning disabilities across five domains of variations in cognitive and behavioral characteristics
cognitive distortions—self-criticism, self-blame, across groups.
Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024
22
 Analyzing Cognitive Distortions and Behavioral Patterns in Children With Learning Disabilities

The findings indicate significant disparities, with behaviors, nor was there a significant
specific learning-disabled groups exhibiting more relationship between withdrawn/depressed
pronounced behavioral challenges. Notably, female behavior and gender.
participants with learning disabilities demonstrated  Somatic complaints (such as headaches or
greater behavioral difficulties compared to their male stomachaches) were more prevalent in female
counterparts, particularly in areas related to anxiety, participants with learning disabilities, with a
attention issues, and rule-breaking behaviors. The significant correlation between somatic
two-way ANOVA analysis revealed the distinct complaints, gender, and participant type.
influence of both gender and disability status on  Social problems were more prominent among
cognitive distortions and behavioral competencies. male participants with learning disabilities than
female participants.
The study’s findings led to the following conclusions  Thought-related problems tended to be higher
based on a sample of 100 students, divided equally among participants with learning disabilities,
between 50 male and 50 female participants: though no significant gender difference was
 Self-criticism was notably higher among female observed; however, there was a significant
participants with learning disabilities, with a interaction between thought problems and the
significant correlation between gender and type combined variables of gender and participant
of participant (learning disabled vs. typical) in type.
relation to self-criticism.  Attention-related issues were more prevalent
 Female participants with learning disabilities among female participants with learning
showed a stronger tendency toward self-blame, disabilities than among typical male participants,
with a notable connection between gender and though no significant interaction was found
participant type (learning disabled vs. typical). between attention issues, gender, and participant
 Feelings of helplessness were more commonly type.
reported by female participants with learning  Rule-breaking behaviors were observed more
disabilities, with a significant correlation frequently among male participants than females,
between helplessness, gender, and type of with participants with learning disabilities
participant. showing a higher tendency for rule-breaking
 Female participants with learning disabilities compared to typical participants.
showed a higher prevalence of pessimism, with a
significant association between gender and These conclusions highlight distinct cognitive and
participant type in terms of pessimistic outlook. behavioral characteristics across gender and learning
 Compared to male participants, those with disability status, providing valuable insights into the
learning disabilities, particularly females, unique challenges faced by students with learning
demonstrated a stronger tendency toward risk- disabilities.
related preoccupations. However, no significant
interaction was found between the variables of REFERENCES
risk obsession and participant type.
 No gender differences were observed on the [1] Prifitera, Aurelio, Saklofske, H. Donald, and Weiss, G.
Lawrence (1998). WISC-IV Clinical Assessment &
anxious/depressed scale, though participants with Intervention (2nd ed.).
learning disabilities reported higher levels of [2] Kapur, Malavika (1997). Counseling Children with
anxiety and depression overall. No significant Psychological Problems.
correlation was found between participant type [3] Thackeray, E.J., & Richdale, A.L. The Behavioral Treatment
of Sleep Difficulties in Children with Intellectual Disabilities.
and gender regarding anxiety or depressive Behavioral Interventions.
behaviors.
 Learning disabled participants showed higher
levels of withdrawn and depressive behaviors,
with no observable gender differences in these



Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

23
 THE IMPACT OF PERSONALITY TYPES IN ADOLESCENTS:
IMPLICATIONS FOR MENTAL HEALTH, DEVELOPMENT, AND
LIFELONG OUTCOMES
SHAILY GAMBHIR

PhD Scholar, Department of Psychology, RTMNU University, India, MA Psychology, B.Ed,

E-mail: shailygambhir@hotmail.com

Abstract - Adolescence represents a transformative period, bridging childhood and adulthood, marked by profound physical,
cognitive, and emotional development. During this phase, the foundations of identity and personality traits begin to solidify,
directly influencing an individual’s mental health, social integration, and adaptive capacities. This research paper delves into
the well-established Type A (competitive, driven) and Type B (relaxed, patient) personality frameworks, examining the
unique developmental trajectories of each type, their respective impacts on mental health, academic performance, and social
relationships, and their predictive value for adult life outcomes. The role of both biological and environmental factors in
personality formation is explored, underscoring the interplay between genetic predispositions and life experiences.
Recognizing the long-term influence of adolescent personality on psychosocial adjustment, this paper emphasizes the
importance of early interventions and tailored coping strategies to promote resilience, aiming to lay the groundwork for
healthier, more adaptable adult personalities.

Keywords - Adolescence, Mental Health, Personality, Type A, Type B

I. INTRODUCTION also in brain structure and function. Key among these

changes is the maturation of the prefrontal cortex,
Adolescence is a pivotal period of physiological, which oversees critical functions such as decision-
emotional, and cognitive transformation that making, planning, and impulse control. This region
significantly influences the formation of personality. continues to develop throughout adolescence, shaping
During this phase, young individuals actively explore personality traits related to conscientiousness and
their identities, navigate complex social dynamics, emotional self-regulation.
and develop autonomy.
Hormonal fluctuations accompanying puberty further
Personality, defined as the enduring patterns of contribute to shifts in mood and behaviour, adding
thoughts, emotions, and behaviours, becomes complexity to the adolescent personality.
increasingly stable, with traits established during
adolescence playing a critical role in shaping long- 2. Social and Environmental Factors:
term behaviour, mental health, and life outcomes. Adolescent personality is significantly moulded by
family dynamics, peer influence, and broader societal
This research paper examines the development of two expectations. During this period, adolescents navigate
primary personality types—Type A, marked by a complex balance between striving for independence
competitiveness and urgency, and Type B, and relying on parental guidance. Peer relationships
characterized by relaxation and patience—during play an especially central role, with adolescents
adolescence. increasingly turning to their social networks for
validation, identity exploration, and behaviour
Analyzing how these traits emerge and affect modelling. These social environments help frame
adolescents provides crucial insights for identifying their values, behaviours, and emerging sense of self.
those at risk for mental health challenges and for
designing targeted interventions that foster resilience, 3. Cultural Influences :
healthy coping mechanisms, and positive Cultural context profoundly shapes personality
psychosocial outcomes. Understanding these development during adolescence. In collectivist
developmental pathways is essential for promoting cultures, emphasis on group harmony and social
well-being and adaptive behaviour that will persist responsibilities often fosters traits such as
into adulthood. agreeableness and conscientiousness.

II. IMPACT OF VARIOUS FACTORS ON In contrast, individualistic cultures tend to encourage

ADOLESCENT PERSONALITY qualities like extraversion, assertiveness, and
DEVELOPMENT autonomy, aligning personality development with
values of independence and self-expression. This
1. Biological Factors : cultural framework influences adolescents' goals,
The onset of puberty initiates profound self-concept, and interpersonal relationships, adding a
transformations, not only in physical development but layer of diversity to personality trajectories.
Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024
24
 The Impact of Personality Types in Adolescents: Implications for Mental Health, Development, and Lifelong Outcomes

III. BACKGROUND OF PERSONALITY TYPES Impact on Mental Health

The heightened stress levels associated with Type A
Origins of Type A and Type B Personality adolescents can significantly impact their mental
Concepts health. Research indicates that these adolescents are
The Type A and Type B personality framework was at greater risk for anxiety, depression, and other
pioneered by cardiologists Meyer Friedman and Ray stress-related disorders, especially when they feel
Rosenman in the 1950s, who observed a link between unable to meet their own high standards.
personality traits and cardiovascular health. They Perfectionism, commonly linked to Type A traits,
theorized that individuals with Type A personalities, often exacerbates these mental health challenges,
marked by competitiveness, urgency, and a high- leading to emotional distress. Studies have found a
stress response, had an increased risk of heart disease. strong connection between Type A personalities and
In contrast, Type B individuals, characterized by a burnout, particularly among high-achieving
calm, relaxed demeanour, appeared less vulnerable to adolescents who experience persistent pressure to
stress-related illnesses. Although initially developed perform.
within the context of cardiology, this framework has
since broadened to encompass diverse domains such Academic Performance
as mental health, stress resilience, and interpersonal Type A adolescents are often high achievers, driven
behaviour across different age groups, including both by their ambition and desire for success. This drive
adults and adolescents. can lead to outstanding academic performance, as
they push themselves to excel. However, the intense
Personality Development in Adolescence pressure they place on themselves to meet their goals
Adolescence represents a critical period of can result in negative consequences, such as stress-
neurobiological, social, and psychological related issues, burnout, test anxiety, and physical
transformation. During this phase, individuals engage symptoms like headaches and fatigue. Research
in identity exploration and consolidation—a central suggests that while Type A students often thrive in
tenet of Erik Erikson's psychosocial development competitive, performance-oriented environments,
model. The maturation of the prefrontal cortex they may face difficulties in settings that prioritize
facilitates abstract thinking and improved emotional creativity, collaboration, and less structured
regulation, providing a foundation for the approaches to learning.
solidification of personality traits.
While personality traits demonstrate relative stability V. CHARACTERISTICS OF TYPE B
over time, adolescence remains a dynamic period PERSONALITY IN ADOLESCENTS :
where environmental influences, peer relationships,
family interactions, and academic pressures Behavioural Traits
collectively shape personality type. Childhood In contrast, adolescents with Type B personalities
temperament may evolve, and distinctions between display a more relaxed, easygoing approach to life.
Type A and Type B traits often become more evident They exhibit patience, lower levels of
as adolescents navigate the challenges and competitiveness, and a diminished focus on deadlines
complexities of this developmental stage. This and time constraints. These individuals tend to
intersection of biological maturation and social approach tasks with a calm, methodical mindset,
context underscores adolescence as a key period for finding enjoyment in the process itself rather than
understanding the emergence of Type A and Type B fixating solely on outcomes. Additionally, Type B
personality distinctions. adolescents are generally more adaptable to change,
less likely to engage in social conflicts, and often
IV. CHARACTERISTICS OF TYPE A demonstrate higher emotional stability and resilience.
PERSONALITY IN ADOLESCENTS
Impact on Mental Health
Behavioural Traits Adolescents with Type B personality traits,
Adolescents displaying Type A personality traits are characterized by a more relaxed and easy-going
typically characterized by high levels of nature, generally experience lower levels of stress and
competitiveness, a strong drive for achievement, and anxiety compared to their Type A peers. This relaxed
an ambition to succeed. They possess a pronounced disposition allows them to adopt more effective
desire for control over their surroundings and exhibit coping mechanisms, making them less vulnerable to
a sense of urgency, often engaging in multitasking. stress-related mental health issues. However, in
These individuals tend to internalize stress and highly competitive or achievement-focused
frustration, particularly in the face of academic or environments, Type B adolescents may encounter
social setbacks. Impatience and hostility may emerge challenges. They might struggle with feelings of
in their interactions with peers and authority figures, inadequacy or underperformance, as these settings
leading to potential conflicts and strained tend to reward the drive and ambition typically
relationships. associated with Type A personalities.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

25
 The Impact of Personality Types in Adolescents: Implications for Mental Health, Development, and Lifelong Outcomes

Academic Performance stress, cardiovascular health issues, and burnout. In

Type B adolescents may not exhibit the intense contrast, those with Type B traits may enjoy more
academic drive characteristic of Type A students; balanced mental health but could encounter
however, they tend to excel in environments that challenges in achieving success in highly competitive
prioritize collaborative learning, creativity, and fields.
emotional intelligence. Their balanced approach to
academics often contributes to better mental health, Early Interventions and Adaptive Coping
even in high-pressure situations. Nevertheless, in Strategies
highly competitive settings, they might face Given the potential negative outcomes associated
challenges due to their lower sense of urgency and with both personality types, early intervention is
preference for steady rather than high-stimulus or essential.For Type A adolescents, stress management
high-challenge tasks. methods—such as mindfulness, cognitive-
behavioural therapy (CBT), and relaxation
Biological Foundations of Personality Types: techniques—can help reduce stress and improve well-
Genetics and Temperament being. For Type B adolescents, promoting goal-
Research indicates that temperament, an early setting skills and encouraging healthy competition
determinant of personality, carries a genetic basis that can support them in achieving greater academic and
shapes traits associated with Type A and Type B personal success while preserving mental health.
personalities. Type A tendencies, including high
arousal, impulsivity, and environmental sensitivity, VII. CONCLUSION
are often linked to genetic factors, as are Type B
traits like emotional stability and lower reactivity to The emergence of Type A and Type B personality
stress. Neurobiological components, particularly the traits in adolescence reflects a dynamic interplay of
function of the hypothalamic-pituitary-adrenal (HPA) genetic, environmental, and social influences. Type A
axis and serotonin levels, may further influence these adolescents, characterized by high levels of ambition
personality predispositions. and competitiveness, are at a heightened risk for
stress-related challenges such as anxiety and burnout.
Environmental Shaping of Personality: Conversely, Type B adolescents, known for their
Socialization and Contextual Influences relaxed and easy going nature, may encounter
While genetics lay the groundwork, environmental difficulties in high-pressure or competitive
factors heavily impact whether adolescents exhibit environments. Early identification of these
Type A or Type B characteristics. Influences such as personality types is essential for developing
family interactions, parenting approaches, peer interventions that nurture mental resilience and
dynamics, and academic expectations play pivotal adaptive coping strategies. By recognizing and
roles. For instance, adolescents in competitive, high- addressing personality diversity during this formative
pressure environments are more likely to develop stage, educators, parents, and mental health
Type A traits, whereas those in nurturing, supportive professionals can offer more personalized support,
settings are more inclined to show Type B traits. helping adolescents build emotional strength and self-
awareness. This proactive approach not only supports
Cultural and Gender Influences on Personality adolescents’ immediate well-being but also
Development contributes totheir development into balanced, well-
The process of socialization, encompassing gender adjusted adults, underscoring the importance of
norms and cultural expectations, also affects focusing on personality development throughout
personality outcomes. In some cultural contexts, Type adolescence.
A traits like assertiveness and competitiveness may
be actively encouraged, especially among male REFERENCES
adolescents. Meanwhile, Type B traits, such as
emotional regulation and patience, may be more [1] Carver, C., & Connor-Smith, J. (2010). Personality and
Coping. Annual Review of Psychology. Palo Alto, Vol. 61
valued in female adolescents, contributing to the pg. 679.
formation of personality types within specific cultural [2] Chinaveh, M. (2014). A comparison of Type-A and Type-B
frameworks. Learners in the perception of stress level and use of coping
responses in the Campus.Procedia - Social and Behavioural
Sciences 143: 384 – 388 .
VI. LONG-TERM EFFECTS OF TYPE A AND B [3] Erikson, E. H. (1959). Identity and the Life Cycle.
PERSONALITY TRAITS IN ADOLESCENTS International Universities Press.
[4] Ferguson, E. (2001). Personality and coping traits: A Joint
Impacts on Adult Life Outcomes factor analysis. British Journal of Health psychology; 6: 311-
325.
The personality traits formed in adolescence often [5] Friedman, M., &Rosenman, R. H. (1974). Type A Behaviour
shape adult life outcomes. Adolescents with Type A and Your Heart. Knopf.
personality traits may be driven to succeed [6] Lateef , A., Dahar , A. M. , Yousuf, M. I. , (Spring 2019) ,
professionally but face increased risks of chronic Global Social Sciences Review (GSSR), Vol. IV, No. II |
Page: 80 – 8

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

26
 The Impact of Personality Types in Adolescents: Implications for Mental Health, Development, and Lifelong Outcomes
[7] Matthews, K. A., & Haynes, S. G. (1986). Type A behaviour Organizational Psychology Scope. Psychology Research,
pattern and coronary disease risk: Update and critical Vol.12,No.12,951-956
evaluation. American Journal of Epidemiology, 123(6), 923- [10] Steinberg, L. (2014). Age of Opportunity: Lessons from the
960. New Science of Adolescence. Houghton Mifflin Harcourt.
[8] McCrae, R. R., & Costa, P. T. (1997). Personality trait [11] Thomas, A., & Chess, S. (1977). Temperament and
structure as a human universal. American Psychologist, Development. Brunner/Mazel.
52(5), 509-516. [12] https://www.medicinenet.com/type_a/article.htm
[9] Samaras, A. &Galanakis, M, (2022).Type A and Type B [13] https://www.simplypsychology.org/personality-a.html
Personality Types: An Evaluation in the Modern



Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

27
 AI: UNETHICAL USAGE MAY SABOTAGE DEMOCRACY
1
JOSEPH PHILIPS, 2R. SUBRAMANI
1
PhD Scholar, Department of Journalism and Mass Communication, Periyar University, Salem - 636 011, Tamil Nadu, India.
2
Associate Professor, Department of Journalism and Mass Communication, Periyar University, Salem - 636 011,
Tamil Nadu, India.
E-mail: 1james2000leslie@gmail.com, 2erasubramani@gmail.com

Abstract - AI is all pervasive technology which has brought in transformative changes in several industries leading to
enhancement of technological capacities. It facilitates facial identification, object detection and quality control in
manufacturing industries. Automotive industry using it to read sensor in a self-driven car. It can effectively thwart attacks in
cyberworld; be it weather forecast, healthcare, financial services, or education, AI has capacity to galvanize these sectors. At
the same time, AI poses potential risks like job displacement, security risks, financial crisis, health hazards, and peril of
Unarmed Aeriel Systems are noteworthy. More importantly, with the advent of High-Level Machine Intelligence (HLMI),
identifying trends in political sector has become an easy task. And of late, many breakthroughs in HLMI are heavily used to
devise election campaigns based on the trends deftly exposed by an amalgamation of Machine Learning, Deep Learning, Big
data analytics, and Artificial Intelligent system. This way, AI techs lay bare social, psychological, political, financial,
spiritual preferences of people, thus, making them vulnerable at the hands of political parties which exploit their data that
reveal penchants or mindset of electorates. This modus operandi of political parties shows their vicious design in obtaining
data of innocent people, who do not know when it is poached and used to gain political milage. This very tactical approach
shows the shrewd system at work is unethical because it is stark manipulation of gullible minds. It is outrageous when you
exploit ‘decision-making’ of huge mass of people to usurp political power and be the ruler. Cannot it be termed as sheer
robbing of people’s discretion? It is undemocratic and against the spirit of Constitution. Discretion is something sacrosanct
where consent is paramount, it’s not commodity that can be poached, but in today’s scenario above method has
commoditized ‘discretion.’ Such political tendencies are threatening democracy and AI is the main culprit. Therefore, it is
collective imperative of a civil society to formulate regulatory system to reign in AI, that cannot de-destabilize democracy.

Keyword - Machine Learning, Natural Language Learning, Deep Learning, Etc.

I. INTRODUCTION touching smart devices like mobile, laptop, desktop,

tablet, AI begins to operate in the background. And
Artificial Intelligence has become the most pervasive when one visits website and logs in, a set of
technologies in the world which has firmly seeped algorithms will set in to analyse and synthesis data
into and encompassed almost every civilian sphere of pertaining to an individual behaviour and
life and now it is transforming military sector as well. instantaneously it identifies as what service and types
With AI’s introduction in the field of education, of products entice you. And forthwith specific
healthcare, finance, business, fast moving good and advertisements are produced which often catch your
manufacturing sectors, crucial functions like imagination. In this cut-throat competitive world,
productivity and service delivery per se are China seems to have outwitted US in earmarking
effectively optimized and intelligently galvanized in a billions of dollars to ramp up multiple technologies
way that it has completely metamorphosed the way and to make leeway, whileUS firms are investing
human thinks and fetches unexpected dividend of their two-third profits in AI-applications. AI’s pivot is
their hard-work. This sophisticated technology which computational processing and data serve as the new
has become the part and parcel of human lives may oil. Incredible advances in robotics and unimaginable
not be thoroughly anthropomorphised but to the progress facilitated by machines are so remarkable
larger extent, it perceives and efficiently responds to that what was prophesized before are pales in
the emerging needs and potentially building the comparison. As of now all cutting-edge technologies
technical capacity that enables sector to leapfrog are helmed by AI, wherein machine learning is
traditional barriers which otherwise could have kept advanced and self-trained to use statistical methods to
us grappling with impediments which we often learn and process data. Google, a globally known
surmise to be unsurmountable. And, this very smart search engine is AI-driven was built by two Standford
amalgamation of three sophisticated technologies i.e. University students, Larry Page, and Sergey Brin in
machine learning, deep learning, and natural language 1997 and such know-hows are involved in
processing (NLP) has created a new epoch that has sophisticated search engine, grammar correction and
diametrically changed the world dynamics. AI is chatbots, etc. Deep Learning is strong form which
called artificial for it relies on computational emulates neural pathways in the brain and get
programs rather than biological, since it can sense, strengthened every time they are used. With this Deep
reason, learn and act that is why it is intelligent in its Learning locates and understands web of amorphous
operation. Its overarching and amaranthine presence data and replicates semblance of knowledge that
has enveloped life itself. The moment one start resembles reasoning and these attributes have made

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

28
 AI: Unethical Usage May Sabotage Democracy

AI powerful in this world. As of now, it has whether machine can mimic human-like intelligence.
transformed almost overall aspects of human lives AI has its origin in ancient Greece, where various
across the globe but it seems it has not caught the lores of the yore including telltale colossal bronze
imagination of Indian authorities because only private statue of ‘Talos’ created by Hephaestus (Greek god)
sectors have applied it in business sectors and making supposedly to guard an island of Crete. Second, an
profits. Potential sectors such as finance, healthcare, order by a mythical god of Greek pantheon, Zeus, to
law enforcement, transportation, education, form an imaginary woman to rebuke those seeking to
agriculture, environment protection, etc. need it discover fire or golden autonomous handmaid
which if applied will transform the scenario, yet one endowed with divine power are regarded as fabled
must be circumspect about its darker sides. Areas of forms of AI. These mythological characters represent
life such as face recognition, search human beings’ innate desires to have someone who
recommendations, chatbots, social media suggestions, can be subservient in lessoning their burden and in
advertisements, self-driven vehicles, image this globalized world, researchers often thought of
recognition and social network moderations where AI crafting, or creating a machine to carry out tasks
got its foothold and deep learning is the force behind human does i.e. robots or AI-driven car are recent
these area applications. examples. Kaplan & Haenlein, 2020, is of the opinion
that it is difficult to clearly adumbrateas to what
II. HISTORICAL ORIGINS precisely intelligence can be called. McCarthy tried to
portray using computational part of the ability to
The terminologies which are now common like achieve goals in the world; and the researchers often
cybernetics, information processing and automate have pondered that ones the machines carry out
theory were often debated by the erudite scholars and certain tasks, such ability of execution cannot assume
research community in 1950s; then, the central idea as hallmark of intelligence. Cellan-Jones, 2014;
of their discussion was to find whether machines can Müller, 2020, while corelating the cynicism
replicate what human minds do, but conclave failed surrounding tapping of AI potential, used two
to make headway. However, Mr John McCarthy, an analogies beginning from intuitive fear expressed by
Assistant Professor of Mathematics at Dartmouth the renowned physicist, Dr Stephen Hawkings, who
College in UK, continued to peruse available hinting at a concept of singularity had prophesised
literature and deep engagement with scientific about AI transforming into superintelligent system
community. In 1955, Mr McCarthy hosted a scientific which may become an uncontrollable giant. Secondly,
symposium wherein researchers from diverse fields a group including Elon Musk, Steve Wozniak, and
assembled to think, discuss, develop, and crystallize Yuval N Harari have expressed their fear in an open
ideas about cerebral machine. He being the pivot, letter to the Future of Life Institute saying AI labs is
eschewed the then prevailing ideas i.e. automate aiming at developing powerful digital minds which
theory and cybernetics which revolved around analog may go out of its creators’ control (Future of Life
system and finally discussion was synergized and Institute, 2023).
researchers homed in on the term ‘artificial
intelligence.’ Promptly, McCarthy made proposal III. AI-DRIVEN NEW WORLD ORDER
entitled, ‘Dartmouth Summer Research Project’ and
approached Rockefeller Foundation for funding. In this present world order, there are three dominant
Mervin Minsky, Nathaniel Rochester, and Claude political forces namely US, China and Russia and
Shannon also collaborated with him and on 2nd they can be classified into three regimes like China as
September, 1955, the term ‘Artificial Intelligence’ authoritarian, US, liberal democratic, and hybrid
was introduced to the world. And since then, AI is regimes such as Russia. And all these three regimes
reigning, yes, it has not completely caught the have evolved into digital power hub, wherein we can
imagination of the world but in 2020s, it has begun to call China, digital authoritarianism, Russia, digital
envelop areas and encompassed human lives in a very hybrid regime and US digital liberal democracies.
subtle way that we just cannot pinpoint areas The interest in AI deepened immediately after the
untouched by it. This phenomenon is now so first industrial revolution and it is said, the
pervasive that from GPS navigation system to Antikythera mechanism, an analogue computer,
personalized music recommendation systems dating back to 87 BC, was found sunken off the coast
including chatbots, you name it, it is there. A British of Greece. Leonard Da Vinci designed a mechanical
inventor of computer, Alan Turing, who could crack knight in 1495. Herein two advances are notable, at
Enigma code in World War-II has had lions share in the outset engineering advancement at the beginning
AI and its far-reaching implications. His incessant of 18th and early 19th centuries where steam engines
attempts in decoding the code catapulted him toward had made epoch-making invention and second could
the idea that machine can think. In 1950s, in his be Charles Darwin’s Theory of Evolution which
research article, ‘Computing machinery and emphasised an advent and corresponding
intelligence, 1950,’ he talked about ‘Turing Test’ development of biological species.
(imitation game), which was aimed at knowing

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

29
 AI: Unethical Usage May Sabotage Democracy

IV. INDIAN SCENARIO development in science, technology, and innovation

as it has become part of Thailand’s 12th national
India stands perplexed when it looks at AI-driven economic and social development plan. Compared to
global economy, its poor standing is a matter of great above mentioned four South-Asian nations, Vietnam,
concern, unless it rewires and reframe education has started piloting AI solutions in the year 2021, and
policies or create an alternative model that would it will take years to scale it up and reap benefits.
address contemporary and emerging national and
global challenges, India may find itself way behind. VI. REVIEW OF LITERATURE
As of now under the aegis of Narendera Modi, India
on one hand accentuates digital India but initiatives Mohammad Saifullah in his research disclosed the
such as ‘Make in India’ and ‘Skill India’which might deception and highhandedness of a British political
make jobs obsolescent and we may not dare to consulting firm, Cambridge Analytica, which
achieve expected goal and this fear will further get procured Facebook data of tens of thousands of
compounded if we consider data colonization issue Americans without their consent and ultimately
which in the absence of globally accepted regulations enticed them in electing Donald Trump by flooding
causing anxiety. In thisAI-driven era, India should US voters with tailor-made messages. This above-
also think critically and formulate suitable policies to mentioned political firm had obtained the data from
absorb AI into public sectors to optimize efficiencies the third-party quiz app on the pretext that this
in vital sectors such as healthcare, economy, collected data were to be utilized for academic
agriculture, manufacturing, technology, etc. that will purposes but it was used to build voters’ detailed
usher in a new paradigm wherein effectiveness and profiles for Ted Cruz and Donald Trump’s
accuracy may work as an alchemy of total presidential campaign. That is how when harvested or
transformation. At this juncture, where every nation poached data further get processed using AI
sees AI as key to success, India also must develop AI technologies, poll drives get more specifically
infrastructure keeping in view all ethical parameters sophisticated, and we have seen as to how it worked
to make itself relevant. in 2016 US election, where entire poll campaigns got
provocative and which tried to discredit Donald
V. SOUTH-ASIAN ASPIRATION Trump’s opponent, Hillary Clinton.

Apart from these three super-giants; five major AI- ulletResearchers Ben Hawes, Wendy Hall, Matt Ryan
propelled economies in the South Asia that call the perusing the darker side of AI applications, state that
shot include Indonesia, Singapore, Thailand, such applications can be exploited to craft purpose-
Malaysia, and Vietnam. Indonesian President Jokowi driven videos, news articles, social media posts,
thinks at this juncture whosoever adopt AI will photos and even sound-bites to woo and influence
control the world and this kind of thinking has voters and even doctor their decision-making. And
spurred its digital transformation. Indonesia has had herein, we have wide spectrum of ways and means
the advantage over other four of its contiguous using fake news, misinformation, disinformation,
neighbours as its geographic location is strategic, it misattributions, misinterpretations; doctored videos,
has youthful population raring to go and ever- morphed photos, etc. that finally lead to orchestrate a
expanding market. With full-scale application in narrative which they want to proliferate and push
every sphere of its citizen’s life, it is aspiring to down the gullets of those who may not like it. Such
become front-runner in AI development. Across this drives primarily intend to exploit and manipulate the
archipelago, it looks, it is the digital way of thinking beliefs and prejudices of mass voters, groups or
that has turbocharged its economy which has particular community. Such tailored content
avowedly adopted AI technologies. And they accompanied with visuals has an objective to
intrinsically agree that AI solutions can rejuvenate influence, impress, create a smoke screen before
administrative set-up and make it effective, and genuine voters so that they cannot see the harsh
subsequently, automation can lead to enhanced reality. Most of the time, these set of individuals
productivity. On other hand, Maysia is operating on engaged in flooding social media platforms with such
its National AI Roadmap where ethical approach is provocative content do poison those vulnerable voters
embedded. And it is moving with the logic that AI against other sections to win their favour.
will be omnipresent in its development trajectory.
Singapore is firmly moving with the idea to become The researcher, Sarah M L Bender, in her submission
AI hub by 2030 by using and applying impactful and reasoned that too much dependence on AI cannot be
scalable AI technologies in every aspect of life; with termed sensible because faulty use or
this blueprint in mind, this island is expecting to misinterpretation of genuine algorithmic know-hows
reinvigorate its economy from 3.2 to 5.4% and often begets wrong results, and such phenomenon is
productivity by 41% by 2025. With its foray into AI called ‘automation bias’. The author further argues
in 1970, Thailand has completely encompassed it into that in general, humans very often avoid putting in
its every sector and AI has become crucial for rigorous cognitive effort to solve technically pesky

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

30
 AI: Unethical Usage May Sabotage Democracy

problems, and mostly rely on machines or automated will lead to the exacerbation in labour market
system because they believe machines to be more disparities and secondly, adoption of newtechnologies
effective and accurate than themselves and such will create urban-rural divide. In such circumstances,
fatalistic attitude of self-abasement give an the role of government becomes more important for it
emblematic credibility to AI technologies. On the must reshape capital and labour relations and
contrary, the reality is such advanced machines are reconfigure working conditions. In India, where caste
riddled with prejudices and imprecisions. system still determines the social hierarchy, adoption
Consequently, such humanistic tendencies or of AI technologies may as we know data-driven
automation bias inevitably get users into algorithms pick up biases will stoke social
overgeneralization of smartly-designed AI know- discrimination including gender inequalities.
hows and at the same time get them needlessly over- Therefore, it is imperative for all stakeholders to
depending upon outcomes, this intrinsically may create a framework for the ethical usage of AI. And
prove to be counter-productive, the researcher education being the dominant sector, should
maintained. accommodate ethics as the important component of
syllabus. In contemporary context, having ‘ethical
IIT Patna’s researcher duo, Nihal Raj and Manoj AI’ sounds weird but in can be made possible with
Tiwary, in their paper revealing the facts about the introduction of ethics in curriculum and orienting
India’s political scenario, state the entire campaign developers and users to be the responsible and ethical
trails of India’s 2014 general election helmed by users of AI.
Bharatiya Janta Party under the leadership of
Narendra Modi was thoroughly digitally-driven, it VIII. MERITS & DEMERITS
was complete departure from the age-old traditional
election methods. This paradigm shift by the saffron AI is endowed with incredible potential which offer
brigade in India, has unexpectedly brought them us unfathomable slew of assured advantages but it has
massive political dividend and totally metamorphosed disadvantages as well that must be taken into stride.
the idea of poll planning as they inveterately used AI’s main advantages are: It has capacity to explore
data for accurate decisions, formulated real-time unexplored territories, it executes any given tasks
engagement, and personalized messaging, this event with amazing promptitude. It can perform multiple
embodied the emergence of digital electioneering in functions and accomplish any complex task with
India. This 360-degree turn has rendered the previous ease, its success ratio is high and it is less susceptible
methods like door-to-door contact, husting, rallies, to errors. It is more successful in less time and can
and use of audio-visual media either as obsolete or resolve any problems with greater ease and more
enervatedand strongly brought AI-harnessed importantly, it has the capacity to perform multiple
technologies such as machine learning, natural jobs simultaneously. The main demerits of Artificial
language processing and predictive analytics to the Intelligence-driven technologies are: It just cannot
fore and this innovative approach has galvanized their perform the way humans do because even specific
modus operandi in overpowering political opponent commands get dysfunctional. Its indiscriminate
hands down. They had shrewdly exploited AI application led to job retrenchment and misuse may
algorithms to know voters’ behaviour and preferences entail large-scale destruction. AI techs are time-
using data, and then they went on flooding that consuming, resource-centric and technology is its
demographic groups with tailor-made messages. lifeline. AI just cannot absolutely imitate human-like
Sentiment analysis of ever-expanding social media cognitive understanding and reflexive abilities.
posts enabled them to devise smart roadmap to reach Therefore, it fails to generalize learning from one to
out to swing voters and earmark resources in a right another environment and invariably lacks
manner. commonsense to perceive and instantaneously react
the way humans do in a perspective which may or
VII. DISCUSSION may not be appropriate in each context; that’s why
human-machine and semi-automated systemic
Till date, we know developed and developing world combination which is run or operated by human being
is frenetically endeavouring to build AI infrastructure seems to be the way out. The obvious and overt
which they think will not only revitalize their repercussions have been manifested particularly on
potential sectors but also place them in the driver’s the skill-biased technological change. It was noticed
seat in the new economic world order. In such during the global financial meltdown, many globally-
scenario, India finds itself giving piecemeal attention renowned publications had predicted loss of 47% jobs
to AI, thus, has not yet fully contextualized this in the space of 10 to 20 years in US market, followed
sublime technology in job and skill development by European Union with 54% and World Bank had
sectors. Many studies have brought it to the fore that also predicted 66% employment losses in the
application of AI in broad areas in India’s informal developing nations. The measurable fall out of the
economy tend to disrupt job market wherein semi- introduction of robotics, resulted into reduction of
skilled work-force bears the brunt. Consequently, this employment in US by 0.37% and 0.16 to 0.20% in

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

31
 AI: Unethical Usage May Sabotage Democracy

Europe. And the period between 2005 and 2014 with and how. Specially, when psychographic profiles of
24% of robots pressed into services had caused 1.3% an individual, decision-making, identity, contact info,
jobs displacement in the world and the brunt was etc., get shared, it poses serious threat to one’s
born by the emerging economies as they saw privacy. And unfortunately, still there is appropriate
corresponding 24% decline in jobs. safeguards in place when it comes to online sharing
of data. There is no well-defined law and regulations
IX. DARKER-SIDE that will clearly restricts indiscriminate usage of data,
we also do not have crystal clear boundaries and
Artificial Intelligence has multiple advantages and ethical parameters focusing toward equality and non-
array of benefits which co-exist with some underlying discrimination. Now it is high time that we must
concerns and probable dangers and if we juxtapose establish a well-thought-out equilibrium between
them, we will find why the authority and even firms state interference and private affairs of human being.
cringe to press AI in every aspect of business or Once privacy get infringed, then we are in for online
different services. Bias is one such concern, as we thefts, threats, cyberattacks, trolling, etc. AI has great
know algorithm is the pivot around which AI and immense impact on democracy, which can be
functions, and its developers invariably infuse it with utilized either way, therefore, if people are educated
bias for the reason best known to them. The full-scale and possesses right knowledge, the chances of
adoption of AI strengthens automation in job industry manipulation and rigging get diminished.
but on other hands it amplifies the risk of job
disruption thus causes socio-economic disparities in X. CONCLUSION
the society. The excessive application of AI raises the
most-feared issue of privacy infringement which In today’s machine-centric era, machines are trusted
creates antipathy in the minds of many. In recent more than humans, but it is undeniable fact still there
times, its exploitation in the field of politics and is no alternative for human mind. Even the most
social arena have created a greater discord among the advanced and sophisticated technologies need
public as many cash-rich political parties or humans to keep it going. The reality unbeknown to us
candidates often use AI to settle their score, discredit that machines have limitations as one wrong
their opponents and even launches disparaging command exposes all underlying inaccuracies.
campaigns. The potential threats are perceived in the However, as technologies get streamlined, optimized,
defence sector as we know the world is grappling upgraded, and enhanced, it also makes human to rely
with myriad security challenges in that wake, The on it, that is what happening with AI. With social
United Nations Security Council’s Counter Terrorism media boom and expansion of internet network
Committee while issuing a warning has identified across, AI has tightened its firm grip upon humanity.
Unarmed Aeriel Systems (UAS) as primary terrorist There is no gainsaying the fact about its measurable
threats as in every possibility they maliciously exploit benefits in certain sectors, yet its darker side is
AI which offers them offensive capabilities to engage threatening which predominantly include ethical
with the defence forces. UAS’ utilization as tools of aspect. This technology-driven life begets millions of
terror poses potential dangers to the global security. data every minute and that is new oil for ever-
And there is every possibility that militants can use compassing technology called Artificial Intelligence.
AI-driven autonomous weapon to hit the target In India, it has streamlined healthcare and business
causing disproportionate casualty and massive system, and if applied judiciously can transform
destruction in the warfare. Democratisation of new sectors like finance, judiciary, and military. AI has
technologies has also reduced the barriers for the caught the imagination of developed world but
entry of dangerous actors who always have developing nations such as India is yet to exploit its
malevolent intent and spawning cybercrimes are its full potential. Using AI technologies, BJP seems to be
result. On another hand, over-dependence on AI will in driver’s seat as they had used data analytics to
make society susceptible to mindboggling cyber- understand group communication behaviour and then
attacks that throws life out of joint. In recent times, developed appropriate technology to communicate
the proliferation of chatbot – Chat-GPTs which is AI- with those people. Its IT cell has pre-empted the use
driven computer software simulates human of 3D holographic campaign in 2014 Lok Sabha
conversation to assist in communication operates with election, this critical component of Modi’s
voice commands and text-based messages. When it electioneering had enabled him to address 700 virtual
comes to legal aspects, AI seems to be fraught with rallies which cost him Rs 60 crore. And such
various pitfalls that entails infringement of various holographic speeches were prepared at his residence
legal and human rights. Its algorithmic transparency and simultaneously broadcast from 53 locations
is always shrouded in doubts where names will be across 25 cities during Gujarat Assembly election.
misplaced or may not be there and the victims can not AL technologies used in 2014 election by BJP had
avail benefits or services offered by the government. brought them the windfall, and since then this saffron
In most of the cases people do not know as to which party has been strongly trying to digitize everything
data is collected, who does it, why it is being taken under the sun and has even floated their flagship

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

32
 AI: Unethical Usage May Sabotage Democracy

programs - Digital India and Skill India. When it [4] Hammer, A., & Karmakar, S. (2021). Automation, AI and the
future of work in India. Employee Relations: The
comes to electioneering, AI technologies have International Journal, 43(6), 1327-1341.
intrinsic ability to synthesis mass opinions and [5] Hawes, B., Hall, D. W., & Ryan, M. (2023). Can artificial
insinuate at likely emerging political trends giving a intelligence be used to undermine elections?
chance and an opportunity for parties to respond with [6] India, C. (2016). India and the Artificial Intelligence
Revolution. Carnegie India.
their tailor-made programs and cash in on that [7] Iqra Summan, (2023). Major Boons and Threats of Artificial
behavioural pattern of people. Psephologically, AI Intelligence. Bournemouth University, UK.
offers multiple benefits, to begin with it facilitates [8] Kalyanakrishnan, S., Panicker, R. A., Natarajan, S., & Rao, S.
instant engagement with voters, allow you to (2018, December). Opportunities and challenges for artificial
intelligence in India. In Proceedings of the 2018 AAAI/ACM
formulate crisp and effective personal messaging, and conference on AI, Ethics, and Society (pp. 164-170).
devise accurate poll campaigns. But subsequently, it [9] Kalyanakrishnan, S., Panicker, R. A., Natarajan, S., & Rao, S.
also poses bizarre challenges like manipulation of (2018, December). Opportunities and challenges for artificial
voters’ sentiment, propagation of false narrative to intelligence in India. In Proceedings of the 2018 AAAI/ACM
conference on AI, Ethics, and Society (pp. 164-170).
engineer polarization. India being frenziedly poll- [10] Manheim, K., & Kaplan, L. (2019). Artificial intelligence:
bound nation, all parties will heavily rely on AI tools Risks to privacy and democracy. Yale JL & Tech., 21, 106.
to obviate real-time campaign where deepfakes and [11] Naudé, W. (2021). Artificial intelligence: neither Utopian nor
bots will do the job. This will give them edge over apocalyptic impacts soon. Economics of Innovation and new
technology, 30(1), 1-23.
political opponents but at the cost of democracy! [12] Safiullah, M., & Parveen, N. (2022). Big Data, Artificial
Hence, there is an urgent need to create a regulatory Intelligence and Machine Learning: A Paradigm Shift in
framework and balance between digital ecosystem to Election Campaigns. The New Advanced Society: Artificial
safeguard and preserve democratic ethos and Intelligence and Industrial Internet of Things Paradigm, 247-
261.
immediate need to restore transparency. [13] Şenocak, D., Koçdar, S., & Bozkurt, A. (2023). Historical,
philosophical, and ethical roots of artificial
REFERENCES intelligence. PJE, 40(1).
[14] Sudmann, A. (2019). The democratization of artificial
[1] Bender, S. M. (2022). Algorithmic elections. Mich. L. intelligence. Net politics in the era of learning algorithms.
Rev., 121, 489. Transcript, Bielefeld.
[2] Eshed, G. (2023). Unveiling the Dark Side of Artificial [15] Tomar, M., Raj, M. N., Singh, S., Marwaha, S. S., & Tiwari,
Intelligence. In Is the Chatbot a Threat or an Opportunity for M. THE ROLE OF AI-DRIVEN TOOLS IN SHAPING THE
Security Organizations? (pp. 5–9). International Institute for DEMOCRATIC PROCESS: A STUDY OF INDIAN
Counter-Terrorism (ICT). ELECTIONS AND SOCIAL MEDIA DYNAMICS.
http://www.jstor.org/stable/resrep51667.4 [16] Wright, N. (2018). How artificial intelligence will reshape the
[3] Gul, A. A., Erturk, Y. D., & Elmer, P. (Eds.). (2020). Digital global order. Foreign Affairs, 10.
Transformation in Media & Society. Istanbul University [17] Zhou, Z., &Makse, H. A. (2019). Artificial intelligence for
Press. elections: the case of 2019 Argentina primary and presidential
election. arXiv preprint arXiv:1910.11227.



Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

33
 ACCURATE BRAIN TUMOR CLASSIFICATION BY USING
MOBILENETV1
1
SUMIT KUMAR YADAV, 2ASIF KHAN

Computer Science Engineering, Greater Noida Institute of Technology Greater Noida U.P, India
E-mail: 1sumityadavb4u@gmail.com, 2csasifkhan10@gmail.com

Abstract - Brain tumors are prevalent among children and the elderly, representing a serious type of cancer characterized by
uncontrolled growth of brain cells within the skull. Due to the heterogeneity of tumor cells, classifying them accurately is
particularly challenging. Convolutional Neural Networks (CNNs) are widely employed for visual learning and brain tumor
identification. This study introduces a CNN-based model using a dense variant of EfficientNet, combined with min-max
normalization, to classify 3,260 T1-weighted contrast-enhanced brain MRI images into four categories: glioma, meningioma,
pituitary tumor, and no tumor. The proposed network is an enhanced version of EfficientNet with added dense and dropout
layers. Additionally, data augmentation paired with min-max normalization was applied to enhance the contrast of tumor
cells. The dense CNN model’s advantage lies in its ability to accurately classify a relatively small dataset of images.
Consequently, the proposed model demonstrates exceptional performance, achieving 99.97% accuracy during training and
98.78% accuracy during testing. With its high accuracy and favorable F1 score, this newly designed EfficientNet CNN
architecture proves to be a valuable tool for brain tumor diagnostic decision-making.

Keywords - Mobilenetv1, Brain Tumor, Confusion Matrix, Efficientnet, CNN, MRI

I. INTRODUCTION medical field by developing sophisticated algorithms

to automatically analyze and interpret vast amounts
Brain tumors are serious medical conditions of medical data.
characterized by the formation of abnormal masses or
tumors within the brain or its surrounding tissues. This data includes various medical imaging types
such as MRI, CT scans, X-rays, and ultrasounds, as
They are classified into two main types: primary well as genomics and electronic health records. Deep
tumors, which originate in the brain, and metastatic learning, a subset of machine learning, is particularly
tumors, which spread to the brain from other parts of notable for its ability to extract features from medical
the body. Some brain tumors are benign and non- datasets, identify new patterns, assist in disease
cancerous, while others are malignant and cancerous. diagnosis, and aid healthcare professionals in making
precise decisions.
These tumors differ in symptoms, size, and type.
Common symptoms include headaches, seizures, These algorithms are also used for classifying
memory loss, vision changes, hearing issues, difficulty medical images related to conditions like breast
speaking or understanding language, and numbness in cancer, brain cancer, and COVID-19. They are
the limbs. An example of a brain tumor specimen effective in predicting diseases such as heart disease,
from a 16-year-old patient is shown in Figure 1. diabetes, and Alzheimer’s, as well as supporting early
diagnosis and personalized treatment planning.
Treatment options for brain tumors include surgery,
radiation therapy, and chemotherapy, with the choice Deep learning algorithms have proven highly
of treatment depending on the tumor’s type and stage. effective in classifying MRI images of brain tumors,
automatically extracting features from images of
According to a 2020 report by the American Cancer various sizes and enhancing the accuracy and speed of
Society, brain tumors are among the most fatal clinical diagnoses.
medical conditions in the United States, particularly
affecting adults in their thirties and forties. Recent research highlights the effectiveness of
convolutional neural networks (CNNs) in
In the UK, patients with brain tumors incur an distinguishing between different types of tumors,
estimated annual cost of €14,783, which covers such as glioma, meningioma, and pituitary adenomas,
treatment and doctor visits, while the average cost for and differentiating between neoplastic and non-
all cancers is €16,000. Annually, 16,000 people are neoplastic images.
diagnosed with brain tumors, and this number is
rising. These advancements promise significant
improvements in the accuracy and efficiency of brain
Artificial intelligence plays a crucial role in the tumor diagnosis and treatment planning.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

34
 Accurate Brain Tumor Classification by Using MobileNetV1

Fig. 1. Brain tumor specimen from a 16-year-old child

In this article, we introduce a method for classifying developed a method for detecting meningioma tumors
brain tumor MRI images using the MobileNetV1 using fuzzy logic-based enhancement and a co-active
model. The primary innovation of our approach is adaptive neuro-fuzzy inference system, in addition to
leveraging this highly effective model, known for its U-Net convolutional neural network classification
strong performance in image classification tasks. We algorithms. The proposed detection method involves
evaluated our method with a dataset of 253 brain MRI enhancement, feature extraction, and classification.
images sourced from the Kaggle platform, categorized Fuzzy logic enhances the original brain images,
into tumor and non-tumor images. To address the followed by a dual-tree complex wavelet transform at
dataset's limited size, we employed five augmentation various scales. Features extracted from the
techniques to expand the image count to 1,265, which decomposed sub-band images are then classified
facilitated a more comprehensive analysis of the using the CAN FIS technique to distinguish
model’s classification capabilities. meningioma from non-meningioma images.
Performance metrics such as sensitivity, specificity,
We then assessed the model’s performance using segmentation accuracy, and dice coefficient index are
various evaluation metrics. The results confirm that used to evaluate the tumor detection and segmentation
our approach is effective in accurately classifying system.
brain tumor MRI images and suggests that this model
could significantly improve both the accuracy and Recent advances in deep learning have significantly
efficiency of clinical diagnoses. The remainder of the improved computer-aided brain tumor analysis,
article is organized as follows: Section 2 reviews especially for tumors with variable shape, size, and
recent studies from 2022 that have explored deep intensity. Cheng et al. utilized T1-MRI data to
learning models for brain tumor image classification address three-class brain tumor classification by
and diagnosis. Section 3 details the dataset used, applying image dilation to expand the tumor area,
along with the procedures for training and evaluating which was then divided into progressively finer sub-
the MobileNetV1 model. Section 4 presents and regions. Badza and Barjaktarovic introduced a novel
discusses the experimental results. Finally, Section 5 CNN architecture that modifies an existing pre-
provides conclusions based on our findings and trained network for classifying brain tumors using
identifies potential directions for future research. T1-weighted contrast-enhanced MRI images,
achieving a 96.56% accuracy with two 10-fold cross-
II. RELATED WORK validation techniques on augmented images. Mzough
et al. proposed a fully automated 3D CNN model for
Medical image segmentation is crucial for detecting categorizing glioma tumors into low-grade and high-
and classifying brain tumors from magnetic resonance grade gliomas, using intensity normalization and
(MR) images, as it helps determine the appropriate adaptive contrast enhancement. Their model achieved
therapy at the right time. Various techniques have a validation accuracy of 96.49% with the Brats-2018
been proposed for brain tumor classification in MRI dataset. Hashemzehi et al. evaluated brain cancer
scans. For instance, Shelhamer et al. introduced a detection using a hybrid CNN and NADE model on
dual-path CNN architecture that integrates a deep 3,064 T1-weighted contrast-enhanced images,
coarse layer with a fine layer to achieve precise and attaining a 96% accuracy rate in identifying three
detailed segmentation of brain cancer. Tumor cells distinct brain cancer types. Diaz-Pernas et al.
often exhibit high-intensity malignant fluid, making presented a fully automated brain tumor segmentation
min- max normalization an effective preprocessing and classification algorithm for meningioma, glioma,
tool for grading tumors. and pituitary tumors, utilizing a CNN with a multi-
scale approach and achieving 97% accuracy on 3,064
Numerous image processing methodologies are imaging slices from 233 patients.
employed for classifying MR images. Karunakaran Sultan et al. used a CNN structure with 16 convolution

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

35
 Accurate Brain Tumor Classification by Using MobileNetV1

layers, pooling, normalization, and dropout layers III. PROPOSED METHODOLOGY

before the fully connected layer, achieving a 96%
accuracy rate with 68% of images used for training and In this study, the authors utilized min-max
the rest for validation and testing. Abd et al. examined normalization and data augmentation techniques on a
25,000 brain MRI images using a differential deep- comprehensive dataset of 3,260 brain MRI images.
CNN to classify various brain tumors, attaining an The dataset includes 3,064 T1-weighted contrast-
impressive accuracy of 99.25% during training. Sajja enhanced MRI images sourced from Kaggle. These
et al. used the Brats dataset with 577 T1-weighted images are categorized into three types of brain
brain tumors to classify malignant and benign tumors tumors: meningioma, with 708 images; glioma, with
using the VGG16 network, achieving an accuracy of 1,426 images; and pituitary tumor, with 930 images.
96.70%. Das et al. utilized a CNN with 3,064 T1- The images were collected from 233 patients and
weighted contrast-enhanced MRI images to classify cover three planes: sagittal (1,025 images), axial (994
glioma, meningioma, and pituitary tumors, achieving images), and coronal (1,045 images). The dataset was
94% accuracy by adapting the network with split into three distinct subsets for training, validation,
convolutional filters of varying sizes. and testing. The proposed model incorporates various
stages, as illustrated in Figure 1.

Figure 1. Overview of proposed dense EfficientNet methodology.

Image Pre-Processing

The brain tumor images suffer from low quality due to noise and inadequate illumination. To address this, the
proposed method enhances the images by transforming low pixel value images into brighter ones through data
normalization using min-max normalization. This is followed by applying Gaussian and Laplacian filters.
Initially, Gaussian blur was applied to the original images, and then the blurred images were corrected by
subtracting them from a weighted version of the mask to produce de- blurred images. Subsequently, a Laplacian
filter with a 3 × 3 kernel size was used to smooth the images, as depicted in Figure 2.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

36
 Accurate Brain Tumor Classification by Using MobileNetV1

Figure 2. T1- contrast MR images of each label after filtration.

The MRI images obtained from the patient's database study is expressed as follows:
often lack clarity and contain inherent uncertainty.
Therefore, normalization of brain images is essential d−mnr=d−mn/mx−mn
before further processing.
where d represents the double-precision image, mn is
Typically, MRI images are grayscale, which the minimum value of the image, mx is the maximum
simplifies the normalization process, enhancing value of the image, and r denotes the normalized
image quality and reducing errors. image. This membership function normalizes the
image within the range of 0 to 1, a method also
Nayak et al. employed an L membership function known as max-min normalization. The image
combined with morphological concepts to detect resulting from this normalization process is illustrated
brain tumors. The membership function used in their in Figure 3.

Figure 3. T1-contrast MR images of each label after fuzzification.

Data Division and Augmentation normalization, and data warping were applied to the
original dataset to expand its size and improve
Deep neural networks typically require large datasets training efficacy..
to achieve optimal performance; however, our dataset
is relatively small, containing only 3,260 brain Dense EfficientNet CNN Model
images. To address this, the dataset is split into 80%
for training and the remainder for testing and This article introduces a novel dense CNN model that
validation. Data augmentation is therefore essential to combines a pre-trained EfficientNetB0 with
improve the model’s performance. The authors additional dense layers. The EfficientNetB0
employed several augmentation techniques, including architecture features 230 layers and 7 MBConv
rotation, width-shift, height-shift, and zoom-range blocks, organized into a dense block structure with
adjustments. four interconnected layers and a growth rate of 4.
Each layer in this setup uses the feature maps from
They augmented the original dataset 21 times to previous layers as input. The dense block concept
enhance the volume of training data. This approach employs convolutional layers of the same size as the
increases the diversity of the training set, which helps input feature maps from EfficientNet, leveraging the
the model learn more effectively and reduces the risk output feature maps of earlier layers to generate more
of overfitting. Data augmentation (DA) involves feature maps with fewer convolutional kernels.
generating additional samples through various The CNN model processes enhanced MRI image data
transformations to enrich the existing dataset. at a resolution of 150 × 150 pixels. The dense
Techniques such as dropout regularization, batch EfficientNet network incorporates alternating dense
Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024
37
 Accurate Brain Tumor Classification by Using MobileNetV1

and dropout layers. A dense layer connects all outputs 360, and 180 neurons, respectively, with dropout
from the previous layer to every neuron in the current rates of 0.25, 0.25, and 0.5. The final layer is a dense
layer, with each neuron passing one output to the next layer with four fully connected neurons, coupled with
layer. Dropout layers are employed to reduce network a Softmax output layer to compute and classify the
capacity during training and mitigate overfitting. The probability score for each class. Figure 4 provides a
model starts with a pooling layer, followed by four detailed illustration of the proposed EfficientNet
dense layers and three dropout layers to ensure architecture.
smooth operation. The dense layers contain 720, 360,

Figure 4. Proposed dense EfficientNet CNN model architecture.

IV. RESULTS AND DISCUSSION

Extensive experimental evaluations have been carried out to validate the proposed dense CNN model. These
assessments were performed in a Python programming environment with GPU support. Initially, the MRI images
underwent pre-processing to improve contrast through max-min normalization, followed by data augmentation
for training purposes. The augmented data was used to activate the dense-CNN model, enhancing accuracy. The
model demonstrated a remarkable 99.97% accuracy on the training data and 98.78% accuracy on the testing
dataset, as illustrated in Figure 5.

Figure 5. Graph representing model accuracy and model loss for training and validation set using the dense

The EfficientNet approach was evaluated over 20 epochs with a batch size of 32 and an image size of 150, using
verbose mode 1. Initially, the validation accuracy was below 0.75, but it rose significantly to nearly 0.88 after
Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024
38
 Accurate Brain Tumor Classification by Using MobileNetV1

just one epoch. Similarly, the validation loss started above 0.8 and decreased to below 0.4 after the first epoch.
Figure 5 illustrates a positive trend, showing improved accuracy and reduced loss over time. Validation
accuracy, which began at a lower level, progressively increased to approximately 97.5%. Further evaluation was
conducted using the ResNet50 model, MobileNet, and MobileNetV2 models, with results depicted in Figures
6,7,and 8 respectively.

Figure 6. Graph representing model accuracy and model loss for training and validation set using the ResNet50 approach.

Figure 7. Graph representing model accuracy and model loss for training and validation set using the MobileNet approach.

Figure 8. Graph representing model accuracy and model loss for training and validation set using the MobileNetV2 approach.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

39
 Accurate Brain Tumor Classification by Using MobileNetV1

Based on the accuracy and loss graphs for the various 0.0645. MobileNetV2, however, showed a testing
models, the authors observed that MobileNet's accuracy of 96.94% and a testing loss of 0.2452. The
performance was inconsistent, with a significant MobileNet model had a testing accuracy of 96.94%
disparity between loss and accuracy values, resulting and a testing loss of 0.1339. ResNet’s performance
in lower accuracy compared to other models. In was slightly lower than MobileNet. A detailed
contrast, the accuracy and loss graphs for Dense comparison of the test accuracy and loss for the
EfficientNet, ResNet, and MobileNet are relatively different models is presented in Table 1, with
similar. The Dense EfficientNet model achieved a performance analysis illustrated in Figure 9.
testing accuracy of 98.78% and a testing loss of

Figure 9. Comparison of accuracy and loss among different pre-trained deep-learning-based techniques.

Model Dataset Testing Loss Testing Accuracy

Proposed dense EfficientNet T1 contrast brain tumors 0.0645 98.78%
ResNet50 T1 contrast brain tumors 0.1337 96.33%
MobileNet T1 contrast brain tumors 0.1339 96.94%
MobileNetV2 T1 contrast brain tumors 0.2452 94.80%
Table 1. Comparison of accuracy and loss among different pre-trained deep-learning-based techniques.

To evaluate the performance of the proposed model, misclassifications occurred within the ―glioma‖ class,
several metrics—accuracy, precision, recall, and F1- which struggled to learn as effectively as the other
score—were utilized, all of which are derived from tumor types.
the confusion matrix. Figure 10 shows the confusion
matrix, which highlights misclassifications resulting For comparing different techniques, precision, recall,
from overfitting, based on 10% of the testing data and F1-score were the key metrics considered. These
from the original dataset of 3,260 images. The matrix metrics forall CNN models are detailed in Table 2 and
reveals that the proposed Dense EfficientNet model illustrated in Figure 11. The evaluation metrics are
misclassified 4 tumors, ResNet50 misclassified 12, based on the following parameters:
MobileNet misclassified 10, and MobileNetV2  True Positive (TP): Correctly identified tumor
misclassified 15 out of 326 testing images. The samples.
reduced number of misclassified images contributes  True Negative (TN): Correctly identified healthy
to the higher accuracy of the Dense EfficientNet samples.
model compared to the others.  False Positive (FP): Healthy samples incorrectly
Notably, MobileNetV2 had the lowest confidence classified as tumor.
level for pituitary tumors relative to the other models.  False Negative (FN): Tumor samples incorrectly
All CNN models performed well in classifying classified as healthy.
meningioma tumors. However, the majority of

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

40
 Accurate Brain Tumor Classification by Using MobileNetV1

(a) (b)

(c) (d)
Figure 10. Confusion matrix of: (a) proposed dense EfficientNet model; (b) ResNet50 model; (c) MobileNet model; (d) MobileNetV2
model.

Types of
CNN Dense EfficientNet ResNet50 MobileNet MobileNetV2

Different
types of F1- F1- F1- F1-
Precision Recall Precision Recall Precision Recall Precision Recall
tumors Score Score Score Score
No tumor 1 0.98 0.99 1 0.98 0.99 0.98 0.98 0.98 0.93 0.96 0.95
Pituitary
tumor 0.99 1 1 0.97 1 0.99 0.97 1 0.99 1 0.9 0.95

Meningioma 0.96 1 0.98 0.91 0.98 0.94 0.95 0.95 0.95 0.93 0.99 0.96
Glioma
tumor 1 0.97 0.98 0.99 0.9 0.94 0.98 0.94 0.96 0.92 0.95 0.94
Table 2. Class-specific evaluation of brain tumors using different CNN.

Figure 11. Analysis: class-specific evaluation of brain tumor using different CNN.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

41
 Accurate Brain Tumor Classification by Using MobileNetV1

These parameters are calculated from the confusion precision, recall, and F1- score compared to the other
matrix, which is shown in Figure 10 three models. Among the different tumor types,
pituitary tumors show the best performance across all
Hence, the different measures can be defined as metrics, with high values for precision, recall, and
follows: F1-score.

Overall, Dense EfficientNet demonstrates exceptional

results. For comparative purposes, the authors also
reviewed recent performance metrics of modified
CNN structures from other researchers, as detailed in
Table 3 and illustrated in Figure 12.

The proposed method boasts accuracy, precision, and

F1-score values of 98.78%, 98.75%, and 98.75%,
respectively, surpassing other methods.

According to Table 3, the proposed deep learning

segmentation algorithm outperforms existing state-of-
the-art techniques.

The authors conclude that Dense EfficientNet excels

because deep learning approaches are more effective
Where Recall is the same as Sensitivity as shown in
and capable of managing extensive data for
Equation (2).
classification tasks.
From Table 2 and the analysis graph in Figure 11, it is
evident that Dense EfficientNet achieves the highest

F1-
Authors Year Dataset Model Accuracy Precision
Score
T1
contrast
Badza et al. [12] 2020 brain CNN 96.56% 94.81% 94.94%
tumors
Mizoguchi et al. Brats- 3D
2020 96.49% - -
[13] 2018 CNN
T1
contrast
Hashemzehi et al.
2020 brain CNN and NAND 96.00% 94.49% 94.56%
[14]
tumors
T1
Díaz-Pernas et al. contrast
2021 brain Multi-scale CNN 97.00% 95.80% 96.07%
[15]
tumors
T1
contrast
Sajja et al. [18] 2021 brain Deep-CNN 96.70% 97.05% 97.05%
tumors
T1
contrast
Proposed method Present brain Dense EfficientNet 98.78% 98.75% 98.75%
tumors
Table 3. Comparison of performance among different deep-learning-based techniques.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

42
 Accurate Brain Tumor Classification by Using MobileNetV1

Figure 12. Comparison of performance among different deep-learning-based techniques.

Figure 12 illustrates that all mentioned authors 2095.

[2] Reddy, A.V.N.; Krishna, C.P.; Mallick, P.K.; Satapathy, S.K.;
used contrast brain tumors for their experiments. Tiwari, P.; Zymbler, M.; Kumar, S. Analyzing MRI scans to
The proposed dense EfficientNet method has detect glioblastoma tu-mor using hybrid deep belief networks.
higher accuracy, at nearly 99%, than the others J. Big Data 2020, 7, 35.
[3] Nayak, D.R.; Padhy, N.; Mallick, P.K.; Bagal, D.K.; Kumar, S.
do. Brain Tumour Classification Using Noble Deep Learning
Approach with Parametric Optimization through
V. CONCLUSIONS Metaheuristics Approaches. Computers 2022, 11, 10.
[4] Mansour, R.F.; Escorcia-Gutierrez, J.; Gamarra, M.; Díaz,
V.G.; Gupta, D.; Kumar, S. Ar-tificial intelligence with big
In this paper, the authors have implemented Dense data analytics-based brain intracranial hemorrhage e-
EfficientNet with min-max normalization to classify diagnosis us-ing CT images. Neural Comput. Appl. 2021.
various types of brain tumors, achieving an impressive [5] Rehman, A.; Naz, S.; Razzak, M.I.; Akram, F.; Imran, M.A.
Deep learning-based framework for automatic brain tumors
accuracy of 98.78%, surpassing other methods that classification using transfer learning. Circuits Syst. Signal
used the same dataset. This technique outperforms Processing 2020, 39, 757–775.
existing deep learning models in terms of accuracy, [6] Long, J.; Shelhamer, E.; Darrell, T. Fully convolutional
precision, and F1-score. The proposed approach holds networks for semantic segmentation. In Proceedings of the
IEEE Conference on Computer Vision and Pattern
promise for improving tumor detection in the brain, Recognition, Boston, MA, USA, 7–12 June 2015; pp. 3431–
with glioma exhibiting the lowest detection rate and 3440.
an F1-score of 98%, while pituitary tumors show the [7] Ozyurt, F.; Sert, E.; Avci, D. An expert system for brain
highest detection rate with an F1-score of 100%. tumor detection, Fuzzy C-means with super-resolution and
convolutional neural network with extreme learning machine.
Among deep learning methods, Dense CNN Med. Hypotheses 2020, 134, 109433.
demonstratessuperior performance with rapid [8] Hu, M.; Zhong, Y.; Xie, S.; Lv, H.; Lv, Z. Fuzzy System
processing and high classification accuracy, making it Based Medical Image Processing for Brain Disease
effective for tumor localization and detection. Future Prediction. Front. Neurosci. 2021, 15, 714318.
[9] Maqsood, S.; Damasevicius, R.; Shah, F.M. An Efficient
research will explore advanced pre-processing Approach for the Detection of Brain Tumor Using Fuzzy
techniques, such as fuzzy thresholding or nature- Logic and U-Net CNN Classification. In Lecture Notes in
inspired algorithms, to enhance early diagnosis of Computer Science; Springer: Berlin/Heidelberg, Germany,
critical medical conditions. Additionally, efforts will 2021; Volume 12953.
[10] Ragupathy, B.; Karunakaran, M. A fuzzy logic-based
focus on reducing the number of parameters and meningioma tumor detection in magnetic resonance brain
computing time required for the model, aiming to images using CANFIS and U- Net CNN classification. Int. J.
maintain performance while optimizing efficiency. Imaging Syst. Technol. 2021, 31, 379–390.
[11] Cheng, J.; Huang, W.; Cao, S.; Yang, R.; Yang, W.; Yun, Z.;
Wang, Z.; Feng, Q. Correction, enhanced performance of
REFERENCES brain tumor classification via tumor region augmentation and
partition. PLoS ONE 2015, 10, e0144479.
[1] Pradhan, A.; Mishra, D.; Das, K.; Panda, G.; Kumar, S.; [12] Badža, M.M.; Barjaktarovic´, M.C. Classification of brain
Zymbler, M. On the Classification of MR Images Using tumors from MRI images using a convolutional neural
―ELM-SSA‖ Coated Hybrid Model. Mathematics 2021, 9, network.ˇ Appl. Sci. 2020, 10, 1999.

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

43
 Accurate Brain Tumor Classification by Using MobileNetV1
[13] Mzoughi, H.; Njeh, I.; Wali, A.; Slima, M.B.; BenHamida, (TURCOMAT) 2021, 12, 2103–2113.
A.; Mhiri, C.; Mahfoudhe, K.B. Deep multi-scale 3D [19] Das, S.; Aranya, O.R.; Labiba, N.N. Brain tumor
convolutional neural network (CNN) for MRI gliomas brain classification using a convolutional neural network. In
tumor classification. J. Digit. Imaging 2020, 33, 903–915. Proceedings of the 2019 1st International Conference on
[14] Hashemzehi, R.; Mahdavi, S.J.S.; Kheirabadi, M.; Kamel, Advances in Science, Engineering and Robotics Technology
S.R. Detection of brain tumors from MRI images base on deep (ICASERT), Dhaka, Bangladesh, 3 May 2019; IEEE: New
learning using hybrid model CNN and NADE. Biocybern. York, NY, USA, 2019; pp. 1–5.
Biomed. Eng. 2020, 40, 1225–1232. [20] Cheng, J. Brain tumor dataset. Figshare. Dataset. 2017,
[15] Díaz-Pernas, F.J.; Martínez-Zarzuela, M.; Antón-Rodríguez, 1512427/5. Available online:
M.; González-Ortega, D. A deep learning approach for brain https://figshare.com/articles/dataset/
tumor classification and segmentation using a multiscale brain_tumor_dataset/1512427 (accessed on 12 October
convolutional neural network. Healthcare 2021, 9, 153. 2021).
[16] Sultan, H.H.; Salem, N.M.; Al-Atabany, W. Multi- [21] Nayak, D.R.; Padhy, N.; Swain, B.K. Brain Tumor Detection
classification of brain tumor images using deep neural and Extraction using Type-2 Fuzzy with Morphology. Int. J.
network. IEEE Access 2019, 7, 69215– 69225. Emerg. Technol. 2020,11, 840–844.
[17] Abd El Kader, I.; Xu, G.; Shuai, Z.; Saminu, S.; Javaid, I.; [22] Tan, M.; Le, Q.V. EfficientNet, Rethinking Model Scaling for
Salim Ahmad, I. Differential deep convolutional neural Convolutional Neural Networks. arXiv 2019,
network model for brain tumor classification. Brain Sci. 2021, arXiv:1905.11946.
11, 352. [23] Available online: https://towardsdatascience.com/complete-
[18] Sajja, V.R. Classification of Brain tumors using Fuzzy C- architectural-details-of-all-efficientnet-models-5fd5b736142
means and VGG16. Turk. J. Comput. Math. Educ. (accessed on 16October 2021).



Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

44
 A COMPREHENSIVE REVIEW OF OFFSHORE CARBON CAPTURE
AND STORAGE TECHNOLOGIES INNOVATION CHALLENGES AND
FUTURE HORIZON
DHIVAKAR POOSAPADI

Lead Engineer (Mechanical), Quest Global Services North America

E-mail: dhivap05@gmail.com

Abstract - Carbon capture and sequestration (CCS) is a very crucial strategy aimed at reducing CO2 emissions from
industrial activities and power plants. This process involves capturing CO2 before it enters the atmosphere, transporting it to
a storage site, and securely storing it underground in geological formations or using other methods like human-engineered
technologies as well. The two methods referred to here are likely natural as well as technological forms of carbon
sequestration, both of which remove CO2 from the atmosphere and help combat climate change. Many countries are
committing to achieve "net zero" emissions, meaning they aim to balance the amount of CO2 they emit with the amount they
remove from the atmosphere. This commitment is mainly considered as a strategic move for companies to comply with
regulations, improve their public image, and align with the growing demand for sustainability from consumers and investors
as well. Technological carbon sequestration methods, which are human-engineered solutions have been thoroughly reviewed
in this paper to capture and store CO2. Among various technologies, Carbon Capture and Storage (CCS) is the most well-
known method for sequestering carbon. CCS involves capturing CO2 emissions very effectively from large-scale sources,
like power plants, factories, and industrial facilities, which are major contributors to CO2 emissions.

Keywords - Carbon Capture and Storage, CO2 Emissions, Carbon Sequestration, Extraction Method.

I. INTRODUCTION countries to implement [2]. Thermal efficiency has to

be improved in order to reduce CO2 emissions from
The primary aim of carbon sequestration (CCS) coal-fired power stations very effectively. This means
process is to lower atmospheric CO2 levels to combat making the plant more efficient so less coal is needed
global climate change. Hence, by reducing the amount for the same energy output. Over time, advancements
of CO2 in the air, carbon sequestration helps to reduce have increased efficiency from around 5% a century
the impact of global warming, such as rising ago to over 38% today.
temperatures, more extreme weather and also sea- Supercritical plants, which operate at higher
level rise in a very effective manner. It is highly temperatures and pressures than conventional ones,
notable that the United States Geological Survey can achieve even greater efficiencies [3]. CO2 levels
(USGS) is evaluating two key types of carbon in the atmosphere have been increasing a lot due to
sequestration: They are Geologic Sequestration and fossil fuel combustion, which is a very important
Biologic Sequestration. McKinsey’s Perspective on energy source due to its advantages. Since fossil fuels
Stabilizing Greenhouse Gases, the management will likely remain a key part of global energy, it is
consultancy McKinsey & Company notes that a major very crucial to implement measures to reduce human-
challenge for businesses, governments, and non- caused CO2 emissions [4]. Atmospheric CO2 can be
profits is finding ways to reduce greenhouse gases reduced by improving energy efficiency and using
without harming economic growth [1]. CCS and lower-carbon or non-carbon energy sources are
Climate Change is the report refers to Carbon Capture crucial. Mainly, CO2 can be stored in deep geological
and Storage (CCS) as a key approach in the fight formations, oceans, or converted into stable
against climate change. CCS has been involved at carbonates. In the paper, it has been mainly focused
capturing carbon dioxide (CO2) emissions from on geological storage options, thereby evaluating their
industrial processes, power plants, or directly from the feasibility, cost, capacity, and safety as well. Notably,
air, and storing it underground to prevent it from stable isotopes of CO2 contain very important
entering the atmosphere. information regarding interactions between the
Despite its potential, it has been mentioned in the biosphere and atmosphere as well. The carbon isotope
report that there are several barriers that make it ratio (δ¹³C) of CO2 provides very crucial information
difficult to implement CCS on a large scale: They are regarding the terrestrial carbon cycle and also about
Economic barriers, technical barriers and legal the processes of photosynthesis, respiration, and
barriers. In order to meet Kyoto Protocol emission decomposition as well. The oxygen isotope ratio
reduction targets using cropland carbon sequestration, (δ¹⁸O) reveals interactions between carbon and water
changes in soil carbon need to be measurable and through CO2-H2O oxygen exchange. Researchers can
verifiable. However, measuring soil carbon over a be able to understand and quantify how carbon and
five-year period is challenging and may result in only water cycle function if the carbon isotopic
low to intermediate levels of verifiability. Strict measurements are analyzed thoroughly and ecosystem
verification requirements could be too costly for many models are used very effectively and they can use this
Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024
45
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon

information for hypothesis testing, data interpretation, carbon fluxes (i.e., the flow of carbon into and out of a
and future predictions in a very effective manner [5].A region). It is very important to note that upcoming
literature survey reviewing offshore carbon capture satellites are equipped with lidar, radar, and imaging
and storage (CCS) technologies will be focused on spectroscopy will provide additional data, such as tree
Session 2. The conclusions drawn from the review and biomass and canopy biochemistry can improve spatial
discuss directions will be presented in Session 3 for extrapolation methods and improve the accuracy of
future work in offshore CCS technologies. carbon stock as well as sequestration predictions. To
further refine carbon simulations, improvements are
II. LITERATURE SURVEY needed in two areas: 1) Soil Depth and Soil Carbon
Information and 2) Incorporation of Cropland
The technology for capturing CO2 is not new; it has Alexander S. Antonarakis et al., [7].
been used for many years in industries like food A method has been developed to address challenges in
processing and chemicals, and petroleum companies mapping soil organic carbon (SOC) in semiarid
frequently separate CO2 from natural gas. However, ecosystems using field data and imaging spectroscopy
managing the large volumes of CO2 from major (IS) as well. Field data and IS approach integrates
power stations requires extensive and secure storage field measurements with hyperspectral imagery to
solutions. Saline Aquifers are a very good storage improve the mapping of SOC. Hyperspectral images
option. Notably, Saline aquifers are underground rock have been captured by HyMap sensor, which can be
formations saturated with saline water, capable of affected by non-soil materials like vegetation. It is
storing CO2 securely. For example, in the North Sea, highly notable that Bayer's method reduces the
the Norwegian company Statoil is undertaking a influence of these materials on the soil's spectral
significant CCS project, injecting about a million tons signature, thereby allowing for a very clear
of CO2 annually into aquifers located 800-1000 identification of the residual soil signature. Mainly,
meters below the seabed. Various methods of carbon Extended Mapping Capability method is effective
sequestration have been discussed and different even with up to 40% vegetation coverage, enhancing
approaches to CO2 extraction and storage have been the ability to map SOC over large areas. For accurate
examined very clearly. spatial distribution of ecosystem parameters,
hyperspectral sensors need to be paired with well-
A distance-constrained (DC) zonal analysis method calibrated models and methods to address spectral
has been developed to assess the potential for mixtures (where multiple materials affect the same
increasing carbon sequestration by vegetation in pixel) Anita D. Bayer et al., [8].
mainland China. Zonal analysis approach divides the A transition to a low-carbon economy can be
land into homogeneous zones based on landform, supported through CO2 capture and storage (CCS).
vegetation, and soil (LVS) and also identifies areas Various methods have been reviewed in the study for
where good land management practices (GLMP) can capturing and storing CO2, including in the terrestrial
improve carbon sequestration to meet target levels. By biosphere (e.g., forests), oceans, and deep geological
comparing current carbon sequestration levels with formations as well. Especially, the main focus is on
target levels, the study estimates that about 25% more CCS in deep geological formations, which is
carbon could be sequestered if GLMPs are applied in considered the very important solution for reducing
areas where current vegetation does not meet the CO2 emissions. CCS in deep geological formations as
target. Grasslands have the highest potential for a key strategy for large-scale CO2 reduction, detailing
additional carbon capture, followed by savannas the technology, costs, risks, and future research
(including woody savannas) and croplands. In the directions have been highlighted very clearly in the
study, the regions where carbon sequestration could be study Sally M. Benson et al., [9].
most effectively improved have also been The cost of capturing CO2 from industrial processes
identifiedZongyao Sha et al., [6]. or power plants is expensive, which can be a barrier to
widespread adoption. It is very essential to note that
Large uncertainties in predicting terrestrial carbon CO2 storage in geological formations requires large-
stocks and sequestration arise from a lack of detailed scale infrastructure, which poses logistical and
regional data on forest structure.Satellite waveform financial challenges as well. The current energy
lidar data has been utilized from ICESat to estimate systems have to be adapted to integrate CCS
the forest structure in central New England. ICESat’s technologies, which involves additional costs and
waveform lidar provides fine-scale data on forest technical adjustments.A specific method has been
heterogeneity but does not cover the entire globe. To developed in order to detect gas leaks from
achieve comprehensive coverage, the data from underground sources by observing changes in
ICESat is extrapolated across regions using the vegetation stress using spectral imaging. Visible/Near-
random forest machine-learning algorithm. The Infrared (Vis/NIR) Reflection measures how light is
detailed forest descriptions obtained through this reflected from vegetation to detect changes in plant
method are used to initialize individual-based health or stress. Long-Wave Infrared (LWIR) Thermal
terrestrial biosphere models, which predict regional Emission measures heat emissions from vegetation to

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

46
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon

identify temperature changes that may indicate the been used very effectively to assess the amount of
presence of escaping gas. Practical Use: Elevated carbon that vegetation absorbs and stores, which is
LWIR Brightness: Higher-than-expected thermal very crucial for understanding the carbon
emissions can signal a potential gas leak. Rapid sequestration capacity of ecosystems and tracking
Changes in Vis/NIR Reflectance: Sudden changes in their ecological health as well. Existing products that
how vegetation reflects light can also indicate stress measure vegetation productivity often have limited
caused by gas leaks. Ground-Based Confirmation: spatial (area) and temporal (time) resolution, making it
Once a potential leak is identified through spectral challenging to use them effectively for comprehensive
imaging, ground-based measurements are used to ecosystem service assessments. Warmer temperatures
confirm the presence of gas Jennifer E. Johnson et al., can increase the rate of water loss from vegetation
[10]. (evapotranspiration), which may exacerbate water
For assessing forest carbon sequestration capability, it stress if there is no significant increase in
is very crucial to estimate the carbon stock of forests, precipitation. Higher temperatures can boost the
and to study their spatiotemporal variations further. A respiration rate of plants (autotrophic respiration),
study has been conducted to simulate and analyze the which consumes some of the carbon that plants would
aboveground carbon (AGC) in bamboo forests in otherwise store. This can reduce net primary
Zhejiang Province, China, from 2003 to 2014. An productivity (NPP), which is the net amount of carbon
improved BIOME-BGC (BioGeochemical Cycles) that plants fix after accounting for their own
model has been used for simulating carbon dynamics respirationQinru Liu et al., [13].
in vegetation. It is highly notable that this model has The "dual carbon" goal has been discussed very
been combined with a bamboo forest map derived clearly, which is focused on achieving sustainable
from Moderate Resolution Imaging Spectroradiometer development by reducing carbon emissions and
(MODIS) data. Spatiotemporal variations (changes increasing carbon sequestration. It has been
over time and space) in AGC have been analyzed highlighted in the study that this goal has become a
using geostatistical methods, which involve statistical key policy direction across industries during China's
techniques to analyze spatial data patterns. It has been 14th Five Year Plan and beyond. The Intuitionistic
mentioned in the study that special attention should be Fuzzy Number EDAS (IFN-EDAS) technique has
given to areas within the bamboo forest that have been introduced for managing Multi-Attribute Group
lower carbon storage. The importance of targeted Decision-Making (MAGDM). Hamming distance and
management practices to improve carbon storage in Logarithmic distance have been used by this
bamboo forests, based on simulations and spatial technique under intuitionistic fuzzy sets (IFSs) to
analysis of carbon dynamics have been highlighted evaluate options. IFN-EDAS technique has been
very clearly in the studyHuaqiang Du et al., [11]. applied very effectively to evaluate the sustainable
Advancements in technology have been discussed potential of tourism environments, thereby
very effectively for monitoring CO2 leakage, demonstrating its effectiveness in practical
especially focusing on distributed feedback quantum evaluationsRONG YAN et al., [14].
cascade (QC) lasers that are used for sensing CO2
isotopic ratios. QC lasers can measure different The uncertainties in assessing carbon accumulation in
isotopes of CO2, which helps in detecting as well as agroforestry systems (AFS) have been discussed very
monitoring CO2 leaks. It is highly notable that clearly in the study. It has been highlighted that
effective monitoring of CO2 leakage is very crucial previous land uses were often not considered in many
for the widespread adoption of carbon capture and studies. As a result, it is unclear whether these systems
storage (CCS) technology and ensures that whether always result in a positive carbon balance. Several
pilot projects are validated, builds confidence among barriers have been studied to implement AFS, such
investors, and helps control costs associated with as: Felling regimes: Regulations around cutting down
capital, insurance, and liability. For CCS to be trees, Transit permits: Permissions needed for
effective over long periods (centuries to millennia), transporting timber, Market availability: Challenges in
accurate and continuous leakage monitoring is finding markets for agroforestry products. Despite
essential. Risk-Averse Approach, a safety-first these uncertainties related to factors like region,
approach, which is similar to those adopted in the species richness, tree density, rainfall, and local
nuclear and oil industries, is very vital for the long- choices—the data on biomass, carbon stock, and soil
term acceptance and success of CCS technology. organic carbon (SOC) can guide the development of
Continued development of QC laser-based sensors policies for carbon neutrality and financing in the
may become a key tool for monitoring CO2 and future. The need for careful evaluation of agroforestry
ensuring the safe and effective implementation of systems and the role they can play in achieving carbon
CCS technology Matthew D. Escarra et al., [12]. neutrality while addressing various implementation
The use of quantitative measures of vegetation challenges have been highlighted very clearly in the
productivity has been discussed to evaluate how well studyPankaj Panwar et al., [15].
ecosystems sequester carbon and to monitor A polynomial model has been proposed, which
environmental health. Quantitative Measures have predicts the geo-electrical properties of the CO2–

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

47
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon

water–porous media system, focusing on water been especially aimed to compare different methods
saturation levels. The model is reliable at higher water for estimating biomass changes, which include: field
saturation, while a mixing model from literature is surveys, LiDAR analysis, and optical remote sensing.
more accurate at lower water saturation. It is highly It has been highlighted that while combining data
notable that this model considers fluid pressure and from different sources requires careful consideration
the initial permittivity before CO2 injectionLuqman due to advancing sensor technology, periodic LiDAR
Kolawole Abidoye et al., [16].A geostatistical observations could effectively track biomass changes
inversion approach has been proposed to predict how over time. This approach can be further refined by
CO2 saturation changes over time using geophysical integrating data from other sources, like Landsat
data. The Ensemble Smoother, a stochastic disturbance maps, to improve accuracy and detail in
optimization method has been applied to solve the monitoringJacqueline Rosette et al., [21].
inverse problem, integrating rock physics and
geophysical models. It is very essential to note that The study has been thoroughly examined that carbon
this method was tested on a deep saline aquifer but offset credit payments for soil carbon sequestration in
can also be used for CO2 sequestration in depleted agriculture are mainly benefiting farmers who were
hydrocarbon reservoirs and for Enhanced Oil already using or interested in such practices. It is very
Recovery (EOR) applications involving CO2 injection important to note that carbon offset credit payments
and storageDario Grana et al., [17]. are seen as an extra benefit rather than a primary
incentive, thereby highlighting challenges in ensuring
Carbon utilization technology, which involves that these offset markets achieve true additionality—a
converting CO2 into other products has been very crucial factor for effective climate
discussed very clearly in the study. An example is mitigation Clare T. Barbato et al., [22].
using exhaust gases to grow algae and produce
biofuels. Technologies that purify CO2 can also be When CO2 is injected into the Earth’s subsurface, it
used to create products, but the current high creates significant chemical and mechanical
operational and capital costs mean only high-value imbalances in the system, pushing it far from its
products are feasible. However, as these technologies natural equilibrium.A predictive simulation technique
evolve and costs decrease, their broader use may that models the flow field at an extremely high
become very successfulJohn Kline et al., resolution has been described in the study and this
[18].A closed-loop process has been proposed for technique mainly serves as a foundation for studying
capturing CO2 from cement plants and using it in reactive transport processes involved in carbon
ready-mix concrete as well as concrete products. It is sequestration. Particularly, the simulation technique is
highly notable that this approach reduces CO2 very beneficial to explore how physical parameters
emissions by lowering the amount of cement needed, change with scale, thereby understanding interactions
as injecting liquid CO2 into concrete increases its across different scales as well as identifying broader
strength. Only a small amount of liquid CO2 is physical laws based on detailed small-scale
required, making it an efficient way to both utilize and simulations in a very effective manner. It is very
reduce CO2 emissions in the cement industryJohn important to note that this simulation technique can be
Kline et al., [19]. used by researchers to thoroughly examine how
diffusion influences chemical reaction rates at the
Research has been conducted very effectively in order interfaces between solids and fluids, which is
to measure the thickness of polystyrene coatings considered as very crucial for improving carbon
applied to optical fibers using the dip-coating sequestration processesDavid Trebotich et al., [23].
technique. Notably, the thickness was analyzed Carbon capture and storage (CCS) concept has been
through scanning electron microscopy (SEM). It is discussed very clearly in the study. Initiatives like the
highly notable that these fibers, equipped with LPG Carbon XPrize have been created in order to
sensors, were then used to monitor the dissolution of encourage the transformation of CO2 into valuable
liquid CO2 in deionized water at high pressure, products. Notably, one example is the Air Company,
simulating CO2 sequestration in deep aquifers. A which uses liquefied carbon dioxide and distills it into
correlation between the concentration of the ethanol, later refining it into vodka. It is very
polystyrene/chloroform solution used during coating important to note that this process not only neutralizes
and the thickness of the applied coating, which is CO2 emissions but actually results in a carbon-
important for optimizing the sensors used in CO2 negative product. The scientific process behind this,
monitoring applications, has been established in the led by Stafford Sheehan, begins with liquefied CO2 as
study Luis Melo et al., [20]. a key input Steven Cherry et al., [24].
Research has been conducted as part of NASA’s
Carbon Monitoring System (CMS) initiative, thereby The process of combining convolutional neural
focusing mainly on local-scale mapping of biomass networks (CNN) with the random forest (RF)
and stock changes relevant for forest management as algorithm has been proposed in the study in order to
well. It is very essential to note that this research has leverage CNN's ability to capture texture details in

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

48
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon

high-resolution images and RF's robustness as a global carbon and mitigating climate change as well.
classifier. However, this fusion model is At first, the concept of carbon sequestration has been
computationally demanding, and improving hardware explained clearly and then a different strategy to
(such as memory and GPUs) could help. For tasks like improve the concept has been thoroughly analyzed.
urban land use classification, where shapes and edges The challenges that could prevent the widespread
are key, the ConvRF model might not perform adoption of these measures, such as data overlap,
optimally. It is very important to note that the which complicates understanding the full carbon
researchers suggest focusing on the size of image sequestration potential of wetlands, have been
patches and exploring advanced CNN models as well discussed very clearly. It is highly notable that the
as incorporating multilevel feature representations to study has been especially aimed at addressing
improve the model's effectivenessLuofan Dong et al., concerns from wetland scientists, managers, and
[25]. stakeholders about how enhancing wetland carbon
A new non-destructive method has been proposed for storage can contribute to climate mitigation David
analyzing carbon content in soil using the Inelastic Were1 et al., [29].
Neutron Scattering (INS) reaction. It is highly notable
that this method employs a D-T neutron generator to Biochar-based carbon management networks (CMNs)
measure carbon in soil in situ. The system, optimized have been thoroughly explored as a scalablenegative
in terms of detectors, configuration, and shielding, can emissions solution, especially for agriculture-driven
achieve enough precision to detect changes in soil developing countries. The need for computer-aided
carbon content of about 1%, which is the expected tools like simulation models, optimization techniques,
annual variation. The measurement process takes and remote sensing for better planning and
around one hour per site, allowing multiple sites to be management of these networks have been highlighted
analyzed daily. It is highly notable that this approach very clearly. Although system-level integration is
enables efficient and accurate carbon monitoring in unproven, the required technologies are already
soil Lucian Wielopolski et al., [26]. mature, making biochar-based CMNs a viable option
Landsat Thematic Mapper (TM) imagery has been for reducing carbon emissions Raymond R. Tan [30].
selected for the study on forest carbon stock mapping
due to two key reasons: the relatively low forest The growing global recognition of the need for
carbon stock values helped minimize issues with the sustainable land use and food production has been
saturation of spectral reflectance, and TM images are highlighted very clearly in the study, particularly as
freely available and widely used, especially in climate change impacts become more severe. It is very
developing countries. The study specifically focused important to note that there is increasing interest from
on addressing plot location errors and deliberately businesses in the food and agriculture sectors to assess
excluded uncertainties caused by discrepancies as well as improve their sustainability practices.
between image data and plot measurements, Particularly, the urgency for national as well as
particularly where parts of trees located at plot international guidelines to manage carbon in
boundaries were not fully captured within the agricultural supply chains is also rising, with a
plotsMaozhen Zhang et al., [27]. specific focus on improving policies and laws that
An electrochemical-biological hybrid system has been support soil carbon sequestration, which is a very
developed to improve the speed of CO2 fixation, crucial strategy for climate change
which is naturally a slow process in biological mitigationHaraldGinzky et al., [31].
systems. Specifically, by using electrocatalysts, CO2
has been reduced to formate, which was then The impact of agroforestry on soil carbon stocks using
integrated into the metabolism of E coli. Notably, meta-analytical techniques has been thoroughly
successful carbon fixation and integration of the analyzed, thereby allowing for the comparison of
products into central metabolism have been results from various studies to identify patterns,
demonstrated very effectively, thereby showing differences, or relationships as well. It has been
promise for biochemical production. Specifically, founded in the study that the conversion of land from
further improvements are very much required, agricultural systems to agroforestry generally led to an
particularly in enhancing the reductive glycine increase in soil organic carbon (SOC)
pathway (RGP) in E coli to optimize carbon flux and stocks. However, when converting from uncultivated
improve efficiency as well. Mainly, an or other land uses to agroforestry, the results were
electrochemical-biological hybrid system holds inconsistent, likely due to variability in land-use
significant potential for more renewable and efficient history and limited data. Therefore, agroforestry
biochemical production methods compared to current intends to boost SOC in simpler agricultural
systems Yohei Tashiro et al., [28]. systemsAndrea De Stefano et al., [32].

Numerous ways to improve carbon sequestration (CS) The balance between carbon sequestration and other
in wetlands has been evaluated. It is veryessential to ecosystem services have been evaluated in the study,
note that these ways have a key role in balancing noting that the field of ecosystem services valuation is

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

49
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon

advancing as well as markets is emerging for various sequestration has been highlighted very clearly in the
services beyond carbon. When plantations are used for study. Notably, the inclusion of trees, especially
biological carbon sequestration, there are trade-offs to nitrogen-fixing species, may enhance SOC storage in
consider, such as potential reductions in stream flow these systems. Agroforestry has the potential to
and changes in soil and water quality. It is highly promote rural development very effectively. However,
notable that these co-benefits and trade-offs must be the rates of SOC sequestration in agroforestry systems
carefully considered during negotiations for carbon are highly variable, and only a limited number of field
exchange agreements, as they may impact the overall experiments have rigorously tested these effects. A
value of the ecosystemRobert B. Jackson et al., [33]. very clear understanding of SOC processes in soil
profiles is necessary before recommending specific
In the study, it has been demonstrated very clearly that practices for site-specific SOC sequestration in
buffer zones have a higher root length density(RLD) agroforestryKlaus Lorenz et al., [37].
and root carbon compared to pastures. Buffer zones The managing processes of agricultural systems have
allow them to extract water and nutrients from deeper been thoroughly explored in the study in order to
soil layers, which shallow-rooted crops cannot access. improve soil organic carbon (SOC) sequestration as
Notably, the increased root carbon from buffers well as reduce greenhouse gas (GHG) emissions that
improves soil structure, which in turn improves soil can contribute to mitigating climate change.
hydraulic properties as well. It is very important to Specifically, the role of agriculture in the United
note that the hydraulic properties of soil have been States in this context has been examined and key areas
especially aimed at reducing surface water runoff and have been highlighted very clearly in order to find out
sediment loss from watersheds, thereby contributing where more research is needed to fully understand and
to better soil and water management as wellSandeep optimize carbon sequestration practicesJack A.
Kumar et al., [34]. Morgan et al., [38].
Model simulations have been utilized in the study in
The study has been highlighted very clearly that while order to analyze the carbon sequestration potential of
developing alternatives to fossil fuels is the ultimate different tree species, sites, and silvicultural regimes
long-term solution, soil organic carbon (SOC) very effectively. It has been highlighted in the study
sequestration provides a valuable interim measure. It that certain species, such as the indigenous
is very crucial to note that SOC allows for the removal Podocarpus totara and the exotic Sequoia
of atmospheric CO2 through plant growth, storing sempervirens, are suitable for long-term carbon
carbon in the soil as organic matter, which buys time sequestration. In contrast, species like Pinus radiata,
for the development of renewable energy sources. It is Pseudotsuga menziesii, and Eucalyptus fastigata are
highly notable that SOC sequestration improves soil better for rapid short-term sequestration. It is very
quality by increasing SOC density, enhancing its important to note that future research has been
depth distribution, and stabilizing it in forms that are suggested in the study to explore how various climate
less accessible to microbial processes and this change scenarios might affect the carbon sequestration
protection helps preserve carbon for longer periods, potential of these tree species and whether the optimal
making agroecosystem management crucial for choices for species, sites, and management practices
effective carbon sequestrationR. Lal [35]. would shift under changing climate conditionsSerajis
Salekin etal., [39].
Three key strategies have been proposed in the study A method has been proposed for calculating salinity-
for addressing global carbon emissions: reducing normalized total alkalinity (TA) and dissolved
energy use, developing carbon-neutral fuels, and inorganic carbon (DIC). The researchers achieved this
sequestering carbon as well. It is very important to by using the mean salinity and the intercept from the
note that the manuscript has been mainly focused on regression of TA or DIC against salinity for each site.
carbon sequestration, which involves transferring The normalization process that has been demonstrated
atmospheric CO2 into long-term storage pools, such in the study especially helps to correct for the
as the ocean, soil, living organisms as well as influence of salinity on these measurements. The
geological formations in a very effective identification of dominant biogeochemical pathways
manner. Carbon sequestration process especially helps (like aerobic respiration or sulfate reduction) at each
to slowdown the rise of CO2 levels in the atmosphere. site, has not been addressed in the study, as such
While carbon sequestration is a very crucial method analysis would require more detailed examination of
for mitigating climate change, the development of respiration pathways and carbon isotopic signatures as
carbon-neutral technologies remains equally vital in wellGloria M. S. Reithmaier et al., [40].
reducing overall emissionsRattan Lal et al., [36]. In the study, it has been highlighted very clearly that
estimating forest carbon sequestration in China is
In the study, the potential of agroforestry systems for challenging due to the underrepresentation of tree
improving resource efficiency—such as nutrient, light, demographic dynamics and the impact of harvesting.
and water use—while contributing to climate change It has been founded in the study that statistical models
mitigation through soil organic carbon (SOC) might significantly misestimate long-term carbon

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

50
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon

projections if these models overlook the effects of for this effect is still inconclusivePETE SMITH et al.,
wood harvest and changes in forest demographics. [44].
Ignoring these factors can lead to an early shift in the The characteristics of biochar produced from various
timing of the carbon sink peak by 1–3 decades. To feedstocks and assess its potential for maintaining soil
enhance carbon sequestration, implementing better qualities as well as sequestering carbon have been
forest management practices have been recommended, explored in the study. Particularly, biochar properties,
including postponing harvesting to retain more carbon including elemental composition, pH, surface area,
in timber forests. It has been suggested in the study and cation exchange capacity have been analyzed in
that reducing harvesting intensity and delaying order to understand how different sources affect its
harvests by 5 years to mitigate the potential for grid performance. It is highly notable that the effects of
cells with high carbon removals to become net carbon biochar applications in both pot and field experiments
sourcesZhen Yu et al, [41]. have been compared very effectively in the study. It is
A large-scale study has been conducted very very important to note that soil pH, organic carbon
effectively in order to compare the effects of active levels and soil fertility has been improved by biochar
versus natural restoration on soil carbon sequestration through its large surface area and high cation
across China. It has been found that both restoration exchange capacityTao Xie et al., [45].
strategies generally improved soil carbon levels Generally, in the study, it has been indicated very
compared to croplands. However, the effectiveness of clearly that current research on biochar is limited,
active restoration compared to natural regeneration thereby focusing primarily on short-term pot or field
varied based on context. Active restoration was more studies (1-2 years). Particularly, there is a need for
effective in sequestering carbon in soils that were more comprehensive research that: The study should
initially low in carbon but less effective in carbon-rich integrate information about both the characteristics of
soils. Additionally, active restoration was found to biochar and the soil being amended. Mainly, longer-
increase carbon storage in the topsoil but resulted in term studies are very necessary to understand the
less carbon sequestration in the subsoil compared to sustained effects of biochar. Current models need
natural regenerationDashuan Tian et al., [42]. better representation of how carbon is distributed
The study thoroughly explored as of how adopting among plant parts, soil pools, and plant respiration. It
farming systems have been designed to sequester is very crucial to note that this aspect is not well
carbon in agricultural soils can help mitigate climate understood and requires more research. Models often
change. In the study, it has been specifically focused overlook how changes in nitrogen availability (which
on estimating the potential for organic carbon regulates CO2 assimilation) respond to climate and
sequestration in the upper 0-30 cm soil layer of atmospheric changes as well.
Russian croplands. Maps using data from global and
national databases have been created, aiming to In biogeochemical modeling, a major limitation is the
provide a reproducible method that can be refined insufficient understanding of how nitrogen availability
with more detailed and region-specific data. It is very interacts with climate change. Nitrogen is very
important to note that the main goal is to improve the essential for plant CO2 absorption, but models often
accuracy of predictions regarding soil carbon content lack very crucial insights into how changes in nitrogen
as well as reduce uncertainty in these due to climate and atmospheric conditions affect
predictions. Notably, preliminary data have been carbon dynamics.In order to improve biogeochemical
obtained, and further computations and assessments modeling, it is crucial to understand very clearly as of
are on the process in order to refine the sequestration how root deployment affects soil processes at different
potential and evaluate uncertainties across different depths, including nutrient and water uptake,
scenarios as wellV. A. Romanenkova et al., [43]. decomposition, and carbon and nitrogen sequestration
The potential for negative emissions through soil as well. It is highly notable that more data has been
carbon sequestration and the addition of biochar to required on nitrogen fixation in natural ecosystems
land has been evaluated very effectively in the study. and its response to elevated CO2, climate change, and
It has been evaluated in the study as of how these disturbances as well. It is very important to understand
methods impact global land use, water resources, very clearly that how nitrogen loss processes like
nutrients, albedo (reflectivity), energy, and costs. denitrification, leaching, and volatilization are
While soil carbon sinks may face saturation issues affected by increased CO2 and climate change.
over time, biochar is more stable and less likely to
reach equilibrium quickly. Thus, biochar could remain The potential for carbon sequestration in European
effective as a negative emissions technology (NET) croplands is very important, but its overall impact
well into the second half of the century if mainly depends on economic, political, and cultural
implemented soon. However, there is some major factors and also environmental impacts like non-CO2
concern about a potential "priming effect," where greenhouse gas emissions as well. Since carbon
biochar might accelerate the decomposition of existing sequestration is temporary and non-permanent, it
organic matter. It is highly notable that the evidence should not replace efforts to reduce emissions. By
2100, the carbon emission gap might be too large for

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

51
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon

soil carbon sequestration to have a major role, but it Sequestration Potential in Vegetation by Distance-Constrained
could be useful as a short- or medium-term solution Zonal Analysis. IEEE GEOSCIENCE AND REMOTE
SENSING LETTERS, Digital Object Identifier
while more permanent technologies, like non-CO2 10.1109/LGRS.2020.3003448, 1545-598X © 2020 IEEE.
emitting energy sources, are developed. It is highly [7] Alexander S. Antonarakis and Alejandro Guizar Couti˜no.
notable that european croplands could help meet Regional Carbon Predictions in a Temperate Forest Using
short-term emission targets, especially if integrated Satellite Lidar. IEEE JOURNAL OF SELECTED TOPICS IN
with broader agricultural and sustainability policies. APPLIED EARTH OBSERVATIONS AND REMOTE
SENSING, VOL. 10, NO. 11, NOVEMBER 2017, Digital
Object Identifier 10.1109/JSTARS.2017.2732058, 1939-1404
III. CONCLUSION
© 2017 IEEE.
[8] Anita D. Bayer, Martin Bachmann, Derek Rogge, Andreas
The growing importance of sustainable environmental M¨uller, and Hermann Kaufmann. Combining Field and
planning has been highlighted very clearly in the study Imaging Spectroscopy to Map Soil Organic Carbon in a
as demand for land, particularly for plantations, rises. Semiarid Environment. IEEE JOURNAL OF SELECTED
It is very essential to note that with more land being TOPICS IN APPLIED EARTH OBSERVATIONS AND
REMOTE SENSING. Digital Object Identifier
dedicated to plantations, there is a need to carefully
10.1109/JSTARS.2016.2585674, 1939-1404 © 2016 IEEE.
plan land use to prevent environmental damage and [9] Sally M. Benson and Terry Surles. Carbon Dioxide Capture
ensure long-term sustainability. Ecosystem services and Storage: An Overview With Emphasis on Capture and
(such as water purification, biodiversity, and Storage in Deep Geological Formations. Vol. 0018-
recreational benefits) are becoming more quantifiable 9219/$20.00 2006 IEEE 94, No. 10, October 2006 |
and valuable in economic terms. It has been Proceedings of the IEEE.0018-9219/$20.00 2006 IEEE,
highlighted in the study that markets are emerging for Digital Object Identifier: 10.1109/JPROC.2006.883718.
[10] Jennifer E. Johnson, Joseph A. Shaw, Senior Member, IEEE,
these services, making it easier to compare their value Rick L. Lawrence, Paul W. Nugent, Justin A. Hogan, Laura
with that of carbon sequestration. Decision-making for M. Dobeck, and Lee H. Spangler. Comparison of Long-Wave
land use must consider these co-benefits and trade- Infrared Imaging and Visible/Near-Infrared Imaging of
offs when negotiating policies or deals for land use. Vegetation for Detecting Leaking CO 2 Gas. IEEE JOURNAL
Carbon Dioxide (CO2) Isotope fingerprinting OF SELECTED TOPICS IN APPLIED EARTH
technology is discussed as a future tool for more OBSERVATIONS AND REMOTE SENSING, VOL. 7, NO.
5, MAY 2014, Digital Object Identifier
accurate monitoring of carbon sequestration sites.
10.1109/JSTARS.2013.2295760 ,IEEE JOURNAL OF
CO2 isotope fingerprinting could be used to detect SELECTED TOPICS IN APPLIED EARTH
leaks early, preventing potentially hazardous OBSERVATIONS AND REMOTE SENSING, VOL. 7, NO.
situations for human health if gases were to escape to 5, MAY 2014 1651 1939-1404 © 2014 IEEE.
the surface. This early detection would improve the [11] HuaqiangDu ,Fangjie Mao, Guomo Zhou , Xuejian Li,
safety and reliability of carbon sequestration as a Xiaojun Xu, Hongli Ge, Lu Cui, Yuli Liu, Di’en Zhu, and
Yangguang Li. Estimating and Analyzing the Spatiotemporal
climate mitigation strategy.
Pattern of Aboveground Carbon in Bamboo Forest by
Combining Remote Sensing Data and Improved BIOME-BGC
REFERENCES Model. IEEE JOURNAL OF SELECTED TOPICS IN
APPLIED EARTH OBSERVATIONS AND REMOTE
[1] Wilf Altman. A leading management consultant has said that SENSING. Digital Object Identifier
we can cut our global energy dependency by half in the next 10.1109/JSTARS.2018.2817344, 1939-1404 © 2018 IEEE.
decade or so. So why do only half of managers take [12] Matthew D. Escarra, Loan T. Le, Nathan M. Urban, Michael
sustainable energy seriously?. management sustainability Oppenheimer, and Claire F. Gmachl. Quantum Cascade Laser-
image source, reuters/corbis. engineering & technology 6 June Based Sensing for Carbon Sequestration Leakage Monitoring.
- 19 June 2009 www.theiet.org/magazine. IEEE SENSORS JOURNAL, VOL. 13, NO. 6, JUNE 2013,
[2] Pete Smith∗. Carbon sequestration in croplands: the potential Digital Object Identifier 10.1109/JSEN.2013.2253731, 1530-
in Europe and the global context. Europ. J. Agronomy 20 437X/$31.00 © 2013 IEEE.
(2004) 229–236, 1161-0301/$ – see front matter © 2003 [13] Qinru Liu, Liang Zhao , Rui Sun , Tao Yu, Shun Cheng,
Elsevier B.V. All rights reserved. Mengjia Wang, Anran Zhu, and Qi Li. Estimation and
doi:10.1016/j.eja.2003.08.002 Spatiotemporal Variation Analysis of Net Primary
[3] Bond-Lamberty, B., C. Wang, and S.T. Gower. 2002. ―Coarse Productivity in the Upper Luanhe River Basin in China From
Woody Debris and Its Annual Carbon Flux for a Boreal Black 2001 to 2017 Combining With a Downscaling Method. IEEE
Spruce Fire Chronosequence,‖ Journal of Geophysical JOURNAL OF SELECTED TOPICS IN APPLIED EARTH
Research 108(D3), 8220. OBSERVATIONS AND REMOTE SENSING, VOL. 15,
[4] G. Aydin, I. Karakurt, and K. Aydiner, ―Evaluation of 2022. Digital Object Identifier
geologic storage options of CO2: Applicability, cost, storage 10.1109/JSTARS.2021.3132723.
capacity and safety,‖ Energy Policy, vol. 38, no. 9, pp. 5072– [14] RONG YAN1, YONGGUANG HAN1, FANG LI 2,3, AND
5080, Sep. 2010. PENGFEI LI3. Evaluation of Sustainable Potential Bearing
[5] N. McDowell, D. Baldocchi, J. Barbour, C. Bickford, M. Capacity of Tourism Environment Under Uncertainty: A
Cuntz, D. Hanson, A. Knohl, H. Powers, T. Rahn, J. Multiphase Intuitionistic Fuzzy EDAS Technique Based on
Randerson, W. J. Riley, C. Still, K. Tu, and A. Walcroft, Hamming Distance and Logarithmic Distance Measures.
―Understanding the stable isotope composition of biosphere- Received 26 December 2023, accepted 6 January 2024, date of
atmosphere CO2 exchange,‖ Eos, Trans. Amer. Geophys. publication 11 January 2024, date of current version 19
Union, vol. 89, no. 10, pp. 94–95, Mar. 2008. January 2024. Digital Object Identifier
[6] Zongyao Sha, Ruren Li, Jonathan Li , Senior Member, IEEE, 10.1109/ACCESS.2024.3352732.
and Yichun Xie , Member, IEEE. Estimating Carbon

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

52
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon
[15] Pankaj Panwar 1,*, Devagiri G. Mahalingappa 2, Rajesh SENSING, VOL. 13, 2020, Digital Object Identifier
Kaushal 3, Daulat Ram Bhardwaj 4 , Sumit Chakravarty 5, 10.1109/JSTARS.2019.2953234.
Gopal Shukla 5 , Narender Singh Thakur 6 , Sangram [26] Lucian Wielopolski', Itzhak Orion', George Hendrey', and
Bhanudas Chavan 7 , Sharmistha Pal 1, Baliram G. Nayak 2, Hugo Roge?. Soil Carbon Measurements Using Inelastic
Hareesh T. Srinivasaiah 2, Ravikumar Dharmaraj 2, Naveen Neutron Scattering. IEEE TIUNSAC~ONS ON NUCLEAK
Veerabhadraswamy 2, KhulakpamApshahana 8, SCIENCE, VOL. 47, NO. 3, J U N z~nn o,
ChellackanPerinba Suresh 8, Dhirender Kumar 4, Prashant [27] Maozhen Zhang, Hui Lin, Siqi Zeng, Jiping Li, Junnan Shi,
Sharma 4 , Vijaysinha Kakade 7, Mavinakoppa S. Nagaraja 9 , and Guangxing Wang. Impacts of Plot Location Errors on
Accuracy of Mapping and Scaling Up Aboveground Forest
Manendra Singh 5 , Subrata Das 5, Mendup Tamang 5,
Carbon Using Sample Plot and Landsat TM Data. IEEE
Kanchan 5, Abhilash Dutta Roy 5 and Trishala Gurung 5.
GEOSCIENCE AND REMOTE SENSING LETTERS, VOL.
Biomass Production and Carbon Sequestration Potential of 10, NO. 6, NOVEMBER 2013, Digital Object Identifier
Different Agroforestry Systems in India: A Critical Review. 10.1109/LGRS.2013.2260719.
Forests 2022, 13, 1274. https://doi.org/10.3390/f13081274. [28] Yohei Tashiroa,b1, Shinichi Hiranoc, Morgan M. Matsonb,
[16] Luqman Kolawole Abidoye and Diganta Bhusan Das. Shota Atsumib*, Akihiko Kondoa,d*. Electrical-biological
Geoelectrical characterization of carbonate and silicate porous hybrid system for CO2 reduction.
media in the presence of supercritical CO2–water flow. https://www.sciencedirect.com/science/article/pii/S109671761
Geophys. J. Int. (2015) 203, 79–91 doi: 10.1093/gji/ggv283 8300478, © 2018 published by Elsevier
,GJI Marine geosciences and applied geophysics. Manuscript_593378a7c5dfa1d649ec16f4110d16b1.
[17] Dario Grana ,Mingliang Liu, and Mohit Ayani.Prediction of [29] David Were1,2 · Frank Kansiime2 · Tadesse Fetahi1 · Ashley
CO2 Saturation Spatial Distribution Using Geostatistical Cooper3 · Charles Jju. Carbon Sequestration by Wetlands: A
Inversion of Time-Lapse Geophysical Data. IEEE Critical Review of Enhancement Measures for Climate
TRANSACTIONS ON GEOSCIENCE AND REMOTE Change Mitigation. Earth Systems and Environment
SENSING,Digital Object Identifier https://doi.org/10.1007/s41748-019-00094-0.
10.1109/TGRS.2020.3018910,0196-2892 © 2020 IEEE. [30] Raymond R. Tan. Data challenges in optimizing biochar-based
[18] John Kline and Charles Kline. Cement and CO2: What is carbon sequestration. Renewable and Sustainable Energy
Happening. IEEE TRANSACTIONS ON INDUSTRY Reviews 104 (2019) 174–177.
APPLICATIONS, VOL. 51, NO. 2, MARCH/APRIL 2015. https://doi.org/10.1016/j.rser.2019.01.032.
Digital Object Identifier 10.1109/TIA.2014.2339396, 0093- [31] Harald Ginzky•Fabiano De Andrade
9994 © 2014 IEEE. Corrêa•ElizabethDooley•Irene L. Heuser•PatriciaKameri-
[19] John Kline and Charles Kline. CO2 Capture From Cement Mbote•RobertKibugi•Oliver C. Ruppel. International
Manufacture and Reuse in Concrete. THE CARBON Yearbook of Soil Law and Policy 2022.
XPRIZE. IEEE Industry Applications Magazine, Digital https://doi.org/10.1007/978-3-031-40609-6.
Object Identifier 10.1109/MIAS.2019.2943659, Date of [32] Andrea De Stefano . Michael G. Jacobson. Soil carbon
current version: 11 November 2019. sequestration in agroforestry systems: a meta-analysis.
[20] Luis Melo*1, Geoff Burton1, Stephen Warwick1 and Peter Agroforest Syst , DOI 10.1007/s10457-017-0147-9.
Wild1. Experimental investigation of long period grating [33] Robert B. Jackson,1* Esteban G. Jobba´gy,1,2 Roni Avissar,3
transition modes to monitor CO2 in high pressure aqueous Somnath Baidya Roy,3 Damian J. Barrett,4 Charles W.
solutions. Citation information: DOI Cook,1 Kathleen A. Farley,1 David C. le Maitre,5 Bruce A.
10.1109/JLT.2014.2383356, Journal of Light wave McCarl,6 Brian C. Murray. Trading Water for Carbon with
Technology. 0733-8724 (c) 2013 IEEE. Biological Carbon Sequestration. 23 DECEMBER 2005 VOL
[21] Jacqueline Rosette, Bruce Cook, Ross Nelson, Chengquan 310 SCIENCE www.sciencemag.org.
Huang, Jeff Masek, Compton Tucker,Guoqing Sun, Wenli [34] Sandeep Kumar • Ranjith P. Udawatta • Stephen H. Anderson.
Huang, Paul Montesano, Jérémy Rubio-Gil, and Jon Ranson. Root length density and carbon content of agroforestry and
Sensor Compatibility for Biomass Change Estimation Using grass buffers under grazed pasture systems in a Hapludalf.
Remote Sensing Data Sets: Part of NASA’s Carbon Agroforest Syst (2010) 80:85–96DOI 10.1007/s10457-010-
Monitoring System Initiative. IEEE GEOSCIENCE AND 9312-0.
REMOTE SENSING LETTERS, VOL. 12, NO. 7, JULY [35] R. Lal*. Soil carbon sequestration to mitigate climate change.
2015, Digital Object Identifier 10.1109/LGRS.2015.2411262, Geoderma 123 (2004) 1 –22,
1545-598X © 2015 IEEE. doi:10.1016/j.geoderma.2004.01.032.
[22] Clare T. Barbato1 and Aaron L. Strong. Farmer perspectives [36] Rattan Lal* Carbon sequestration. Phil. Trans. R. Soc. B
on carbon markets incentivizing agricultural soil carbon (2008) 363, 815–830, doi:10.1098/rstb.2007.2185Published
sequestration. npj Climate Action (2023)2:26 ; online 30 August 2007.
https://doi.org/10.1038/s44168-023-00055-4. [37] Klaus Lorenz & Rattan Lal.Soil organic carbon sequestration
[23] David Trebotich, Mark F. Adams, Sergi Molins, and Carl I. in agroforestry systems. A review. Agron. Sustain. Dev. DOI
Steefel|,Chaopeng Shen |.High-Resolution Simulation of Pore- 10.1007/s13593-014-0212-y.
Scale Reactive Transport Processes Associated with Carbon [38] Jack A. Morgan, Ronald F. Follett, Leon Hartwell Allen, Jr.,
Sequestration. Computing in Science & Engineering 1521- Stephen Del Grosso, Justin D. Derner, Feike Dijkstra, Alan
9615/14/$31.00 © 2014 IEEE Copublished by the IEEE CS Franzluebbers, Robert Fry, Keith Paustian, and Michele M.
Schoeneberger. Carbon sequestration in agricultural lands of
and the AIP November/December 2014.
the United States. Jan/feb 2010—vol. 65, no. 1, journal of soil
[24] Steven Cherry , Stafford Sheehan .TURNING CARBON and water conservation,doi:10.2489/jswc.65.1.6A.
DIOXIDE INTO VODKA A BROOKLYN STARTUP IS AN [39] Serajis Salekin1*, Yvette L. Dickinson1, Mark Bloomberg2
XPRIZE FINALIST FOR ITS BOOZY CARBON-CAPTURE and Dean F. Meason1. Carbon sequestration potential of
TECHNOLOGY. SPECTRUM.IEEE.ORG | DEC 2020 | 21. plantation forests in New Zealand - no single tree species is
[25] Luofan Dong, HuaqiangDu ,Fangjie Mao , Ning Han, Xuejian universally best. Carbon Balance and Management (2024)
Li, Guomo Zhou , Di’en Zhu, Junlong Zheng, Meng Zhang, 19:11, https://doi.org/10.1186/s13021-024-00257-1.
Luqi Xing, and Tengyan Liu. Very High Resolution Remote [40] Gloria M. S. Reithmaier1 , Alex Cabral 1, Anirban
Sensing Imagery Classification Using a Fusion of Random Akhand2,3, Matthew J. Bogard 4, Alberto V. Borges 5, Steven
Forest and Deep Learning Technique—Subtropical Area for Bouillon6, David J. Burdige 7, Mitchel Call8, Nengwang
Example. IEEE JOURNAL OF SELECTED TOPICS IN Chen9, Xiaogang Chen 10, Luiz C. Cotovicz Jr11,12, Meagan
APPLIED EARTH OBSERVATIONS AND REMOTE J. Eagle13, Erik Kristensen 14, KevinD. Kroeger13, Zeyang

Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

53
 A Comprehensive Review of Offshore Carbon Capture and Storage Technologies Innovation Challenges and Future Horizon
Lu15, Damien T. Maher 8, J. Lucas Pérez-Lloréns 16, Raghab with initial carbon content and soil layer. ARTICLE
Ray17, Pierre Taillardat18, Joseph J. Tamborski7, Rob C. COMMUNICATIONS EARTH & ENVIRONMENT |
Upstill-Goddard19, https://doi.org/10.1038/s43247-023-00737-1.
[41] Faming Wang 20, Zhaohui Aleck Wang21, Kai Xiao22, [44] V. A. Romanenkova, b, J. L. Meshalkinaa, *, A. Yu.
Yvonne Y. Y. Yau1 & Isaac R. Santos 1. Carbonate chemistry Gorbachevaa, A. N. Krenkec, I. K. Petrovd, O. M.
and carbon sequestration driven by inorganic carbon Golozubova, and D. I. Rukhoviche. Maps of Soil Organic
outwellingfrommangroves and saltmarshes. Nature Carbon Sequestration Potential in the Russian Croplands.
Communications | ( 2023) 14:8196, ISSN 1064-2293, Eurasian Soil Science, 2024, Vol. 57, No. 5,
https://doi.org/10.1038/s41467-023-44037-w. pp. 737–750. © The Author(s), 2024. This article is an open
[42] Zhen Yu 1,2 , Shirong Liu 2 , Haikui Li3, Jingjing Liang4, access publication.
Weiguo Liu 5, Shilong Piao 6, Hanqin Tian 7, Guoyi Zhou 1, [45] PETE SMITH. Soil carbon sequestration and biochar as
Chaoqun Lu 8, Weibin You9, Pengsen Sun2, Yanli Dong1, negative emission technologies. Global Change Biology
Stephen Sitch 10 & Evgenios Agathokleous. Maximizing (2016) 22, 1315–1324, doi: 10.1111/gcb.13178.
carbon sequestration potential in Chinese forests through [46] Tao Xie1; Bala Yamini Sadasivam, S.M.ASCE2; Krishna R.
optimal management. Nature Communications | ( 2024) Reddy, F.ASCE3; Chengwen Wang4; and Kurt Spokas5.
15:3154, https://doi.org/10.1038/s41467-024-47143-5. Review of the Effects of Biochar Amendment on Soil
[43] Dashuan Tian 1,2, Yangzhou Xiang3, Eric Seabloom4, Properties and Carbon Sequestration. Journal of Hazardous,
Jinsong Wang1,2, Xiaoxu Jia1,2, Tingting Li1,2, Zhaolei Li5, Toxic, and Radioactive Waste, © ASCE, ISSN 2153-
Jian Yang6, Hongbo Guo1 & Shuli Niu. Soil carbon 5493/04015013(14)/$25.00.
sequestration benefits of active versus natural restoration vary



Proceedings of SARC International Conference, Madurai, India, 03rd November, 2024

54