THE UGANDA METEOROLOGICAL
SERVICES DATA OBSERVATIONS,
MANAGEMENT, SEASONAL FORECASTING
AND EARLY WARNING SYSTEM FOR
EXTREMES
KHALID Y. MUWEMBE
UGANDA DEPARTMENT OF METEOROLOGY
Mandate of DoM
To promote, monitor weather and climate
and provide weather forecasts and
advisories to Government and other
stakeholders for use in sustainable
development of the country
The role of Meteorological Services
• Key player in monitoring and early warning on
weather related disasters at short and medium
range basis, including extremes such as droughts,
floods and landslides which cause suffering of poor
vulnerable communities
• As natural Resources are diminishing with increased
population pressure, it is becoming increasingly
important to utilize weather and climate information
in planning
• Due to increased Climate Change and Variability,
weather information is key to future adaptation and
mitigation measures which the country will have to
adopt to survive.
DoM
Stations Data processing Training and
Forecasting
Networks And Applied Research
-Synoptic
-Data archive National
-Agromet -NMC Entebbe -Data entry Met.
-Hydromet -Soroti -Agromet Training
-Rainfall Forecast -Seasonal School
-AWS Office forecasting
etc
The Station Network
Division responsible
for the design of
optimal network
system,
implementation and
monitoring of the
networks
12 Synoptic stations
12 Hydromet stations
10 Agromet stations
300 rainfall stations
1 upper-air station
Data management
• Data processing software
– CLICOM
– ClimSoft
• Data rescue efforts ongoing
– Data Lab established with 10 new computers and a
main server
– Digitisation ongoing for manned observations
– AWS data separated archives
Services under applied Meteorology
• Seasonal climate outlook plus monthly reviews
and updates
• Agro-meteorological bulletins on dekadal (10-
days) basis
• Climatological data for different users and
clients
• User tailored information mainly for
construction and insurance companies
Seasonal climate Outlooks
• In Uganda climate related disasters are mainly
associated with seasonal rainfall extremes
resulting in droughts and in situations of enhanced
rainfall it may results into floods and landslides
• Extremes annually destroy an average of 800,000
hectares of crops leading to huge economic losses
especially among poor communities
• Droughts also results into epidemic outbreaks and
climate related conflicts
8
Key Initiatives and Programmes
• Continuous support form IGAD Climate Prediction
and Application Centre (ICPAC) in preparation of
seasonal forecasts during PRE-COFs and COFs.
Regional hub for access to climate products and
data from international forecast centres
• Support from NOAA climate Centre for training two
staff in Numerical Weather Prediction and a work
station for running the WRF Model for short range
forecasting.
DELINEATION OF HOMOGENEOUS
RAINFALL ZONES
• Principal Component Analysis
(PCA), derived from Factor
Analysis
• is a statistical technique used in
identifying a relatively small
number of factors that can be
used to represent relationships
among sets of many interrelated
variables.
• This method has widely been
used in determining regional
homogeneous rainfall zones over
East Africa, Ogallo (1989),
Oludhe, (1987), Basalirwa
(1991), Bamanya (2007)
UGANDA 10
DELINEATION OF HOMOGENEOUS
RAINFALL ZONES
11
UGANDA
Searching and extraction of Predictors
a) SST Gradients
• Predictors these are Climate
Indicators used for Seasonal
Monitoring and Prediction
In case of Uganda’s climate they
include:-
(i) SST:- (SSTs and SST
Gradients, Indian Ocean
Dipole)
(ii) Quasi-Biennial Oscillation
(QBO)
(iii) Madden Julian Oscillation
(MJO)
(iv) Southern Oscillation Index
i) SSTs
Reconstructed sea surface temperature
(SST) dataset from Climate Prediction
Center (CPC) of NOAA
12
RF ANOMALIES
RF ANOMALIES
-2
-1
0
1
2
3
4
5
-3.0000
-2.0000
-1.0000
0.0000
1.0000
2.0000
3.0000
1961 4.0000
1961
1963
1963
1965
1965
1967
1967
1969
1969
1971
1971
1973 1973
1975 1975
1977 1977
TROR
1979 1979
1981 1981
Obs_(Morulem)
1983 1983
F3
YEARS
YEARS
1985 1985
1987 1987
old model
1989 1989
1991 1991
1993 1993
OBSERVED VS PREDICTED RF FOR ZONE5 (TORORO)
1995 1995
OBSERVED VS PREDICTED FOR ZONE3 (MORULEM)
1997 1997
1999 1999
2001 2001
2003 2003
2005
2005
RF ANOMALIES
R F A N O M A L IE S
-2
-1
0
1
2
3
4
-2.000
-1.500
-1.000
-0.500
0.000
0.500
1.000
1.500
2.000
2.500
3.000
19 61
1961
19 63
1963
19 65
1965
19 67
1967
19 69
1969
19 71 1971
19 73 1973
19 75 1975
19 77 1977
EBBE
19 79 1979
19 81
Obs-(Soroti)
1981
19 83 1983
YEARS
YEARS
19 85 1985
19 87 1987
F3
forecast
19 89 1989
19 91 1991
19 93 1993
OBSERVED VS PREDICTED RF FOR ZONE6 (ENTEBBE)
19 95 1995
OBSERVED VS PREDICTED RF FOR ZONE4 (SOROTI)
19 97 1997
19 99 1999
20 01 2001
2003
13
20 03
20 05 2005
Downscaling the forecast
• Interpretation and downscaling the seasonal climate
forecasts is done at national level.
• Seasonal post-COF stakeholder workshop to develop
advisories based on downscaled forecast
• Translation of the current seasonal forecast into seven local
languages. Audio and text translations are disseminated
using community FM radios and local language print media.
This initiative have been done for the last 3 seasonal
forecasts.
• Assessment of forecast performance
WMO member Access: ECMWF & RSMC Guidance
PPT
atleast 10mm
PPT Convection
atleast 25mm & Wind
EPSgrams
Entebbe
EFI
10m Wind gust
EFI
Total PPT index
Impacts of extremes: Floods
Floods in Teso turns roads into river An officer jump flood waters to get to office
Flood water left businesses closed Vulnerable school girl struggles through floods
17
Landslides as a result of excessive rainfall
Landslide causes degradation of land Rudimentary rescue efforts
Vulnerable communities displaced
Landslide destroys a homestead
OBSERVED EXTREMES
Mbarara1999 drought and SOND 2000 floods
MBARARA 1999 AND 2000 CUMULATED RAINFALL
1400
1200
1000
CUM R/Fall
800
600
400
200
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
1999 CUM TOT 2000 CUM TOT LTM CUM TOT
Analysis of recent severe droughts in Teso region
Crop failure
Soroti 1998, 1999 & 2005 drought due to late onset (2005 Drought)
Assessment shows that very late onset of
the rains was responsible for the 1998,
1999 & 2005 disaster
Analysis also clearly detects the severe
1998, 1999 & 2005 drought
To reduce on the impacts of severe
droughts, we need Early Warning System
Selected observed extremes: MAM 2013 season
• Butaleja District (Eastern Uganda) on 13 March, 2013 experienced
heavy rains, accompanied by hailstorm and strong winds. In less than
an hour, 40 houses had their roofs removed, while many crop fields
were destroyed and a 7 year boy was crushed and killed by collapsing
walls.
• On March 30, heavy storm hit Ntoroko Landing Site (western Uganda),
leaving more than 40 houses destroyed, including a school and three
churches.
• A tornado-like storm hits an Island on L. Victoria. Nearly half the
inhabitants of Lujjabwa island on Lake Victoria were rendered homeless
after a powerful storm descended from the clouds and swept over 75
shelters into the lake in early morning of Thursday 14th March.
• Floods ravage Kisoro (SW Uganda), over 150 families displaced.
Several houses were destroyed while community roads were eroded by
the floods rendering them impassable on Sunday 31 March.
Rains wreck havoc in Kampala city slums
• These were the scenes in Kannyogoga Zone Namuwongo on the eve of Easter. Rains that
pounded most parts of country left untold damage and misery in their wake.
People's bedrooms were flooded as residents spent the entire day emptying their cramped
houses of bucketfuls of rain water.
Floods in Nakaseke
• Sunday Vision on 7 April reports that floods triggered off by heavy rains cut off a police post in
Nakaseke and rendered several roads impassable cutting off several villages from the rest of
the district. The heavy rains also caused rivers like Lugogo and Lumansi to burst their banks,
washing away several feeder roads. The most affected sub-county is Kasangombe. The 2
major access feeder roads to Nakaseke Hospital were completely ruined
• We have seen and experienced
the devastating effects of the
extreme weather related risks
and hazard…
• So what next?
The People-centered Early Warning Systems
A good EWS should be an integrated part of planning as a
program designed to mitigate and respond to the disaster
Traditional Framework People-centered framework
Empowering individuals
and communities threatened
by hazards to act in sufficient time
and in appropriate manner to
reduce the possibility of personal
injury, loss of life and damage to
property and the environment
EFFECTIVE EARLY WARNING
SYSTEM
How an early warning system collects and dissemination information
Department of Meteorology
Weather and
Hydrological
data
Local
Government
National
Service
Sectors? UGANDA Civil society
Early Warning
System Communities
(Coordinating
Body - OPM)
Mass Media
Government
Newsletters Local government Agencies
Communities Civil Society
Challenges facing EWS
- Different hazards require different early warning
systems
- Effective communication with communities
- Information fatigue
- Links between analysis and action (particularly between
technical capacity to issue the warning and the public
capacity to respond effectively to warning)
- Accuracy and reliability of information
- Coverage and timeliness
- Political sensitivity
- Decentralization and local responsibility
Way Forward on challenges
Therefore, in case of Uganda:
1. We need to understand the
roles and responsibilities of:
Developing and implementing Communities
an effective early warning Local governments
system requires the Central government
contribution and coordination of Regional institutions and
a wide range of individuals and organizations
institutions. Each has a International agencies
particular function for which it Non-governmental
should be responsible and organizations
accountable The private sector
The science community
2. How should the coordination
body for early warning
system be constituted? Who
should be the members?
Conclusion
• Occurrence of disasters related to extreme
weather events cannot be stopped,
• Timely availability and application of
accurate meteorological information would
assist in making contingency plans thus
avoiding crisis management in responding
to these disasters
Conclusion Cont’d
• There is need to improve monitoring and
observations, modelling, prediction and
early warning capacities;
• Timely availability of data and information;
• Databases for development of drought and
flood indices;
• Vulnerability assessment under different
environmental conditions;
• Skilled human resources, education,
sensitisation and awareness
Thanx for your attention
