ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185 – 198

www.elsevier.com/locate/isprsjprs

Satellite remote sensing of earthquake, volcano, flood,

landslide and coastal inundation hazards
David M. TralliT, Ronald G. Blom, Victor Zlotnicki, Andrea Donnellan, Diane L. Evans
Earth Science and Technology Directorate, Jet Propulsion Laboratory, California Institute of Technology, 4800 Oak Grove Drive,
Pasadena, CA 91109, USA
Received 7 June 2004; accepted 2 February 2005
Available online 1 April 2005

Abstract

Satellite remote sensing is providing a systematic, synoptic framework for advancing scientific knowledge of the Earth as a
complex system of geophysical phenomena that, directly and through interacting processes, often lead to natural hazards.
Improved and integrated measurements along with numerical modeling are enabling a greater understanding of where and when
a particular hazard event is most likely to occur and result in significant socioeconomic impact. Geospatial information products
derived from this research increasingly are addressing the operational requirements of decision support systems used by policy
makers, emergency managers and responders from international and federal to regional, state and local jurisdictions. This forms
the basis for comprehensive risk assessments and better-informed mitigation planning, disaster assessment and response
prioritization. Space-based geodetic measurements of the solid Earth with the Global Positioning System, for example,
combined with ground-based seismological measurements, are yielding the principal data for modeling lithospheric processes
and for accurately estimating the distribution of potentially damaging strong ground motions which is critical for earthquake
engineering applications. Moreover, integrated with interferometric synthetic aperture radar, these measurements provide
spatially continuous observations of deformation with sub-centimeter accuracy. Seismic and in situ monitoring, geodetic
measurements, high-resolution digital elevation models (e.g. from InSAR, Lidar and digital photogrammetry) and imaging
spectroscopy (e.g. using ASTER, MODIS and Hyperion) are contributing significantly to volcanic hazard risk assessment, with
the potential to aid land use planning in developing countries where the impact of volcanic hazards to populations and lifelines
is continually increasing. Remotely sensed data play an integral role in reconstructing the recent history of the land surface and
in predicting hazards due to flood and landslide events. Satellite data are addressing diverse observational requirements that are
imposed by the need for surface, subsurface and hydrologic characterization, including the delineation of flood and landslide
zones for risk assessments. Short- and long-term sea-level change and the impact of ocean-atmosphere processes on the coastal
land environment, through flooding, erosion and storm surge for example, define further requirements for hazard monitoring
and mitigation planning. The continued development and application of a broad spectrum of satellite remote sensing systems
and attendant data management infrastructure will contribute needed baseline and time series data, as part of an integrated
global observation strategy that includes airborne and in situ measurements of the solid Earth. Multi-hazard modeling

T Corresponding author. Tel.: +1 818 354 1835.

E-mail address: david.m.tralli@jpl.nasa.gov (D.M. Tralli).

0924-2716/$ - see front matter D 2005 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS). Published by Elsevier B.V.
All rights reserved.
doi:10.1016/j.isprsjprs.2005.02.002
186 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198

capabilities, in turn, will result in more accurate forecasting and visualizations for improving the decision support tools and
systems used by the international disaster management community.
D 2005 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS). Published by Elsevier B.V. All rights
reserved.

Keywords: Satellite remote sensing; Imaging spectroscopy; Interferometric synthetic aperture radar (InSAR); Global positioning system (GPS);
risk mitigation; Disaster response; Decision support system (DSS); Natural hazards

1. Introduction decision-making. The key to understanding the

Earth’s dynamics and system complexity is to
The last two decades have witnessed the increasing integrate observations at local, regional and global
use of remote sensing for understanding the geo- scales, over a broad portion of the electromagnetic
physical phenomena underlying natural hazards. The spectrum with increasingly refined spectral resolution,
scientific knowledge gained along with the ability to spatial resolution and over time scales that encompass
disseminate timely geospatial information that can be phenomenological lifecycles with requisite sampling
integrated with demographic and socioeconomic data frequency. Advances in computational science and
are contributing to comprehensive risk mitigation numerical simulations are allowing the study of
planning and improved disaster response. As the correlated systems, recognition of subtle patterns in
discipline of Earth science began to recognize the large data volumes, and are speeding up the time
interactions between the hydrosphere, atmosphere, necessary to study long-term processes using obser-
biosphere and solid Earth as a complex system, satellite vational data for constraints and validation (Donnellan
remote sensing uniquely has provided the synoptic et al., 2004). Integrating remotely sensed data into
perspective for in situ measurements on local spatial predictive models requires measurements at resolu-
scales and variable temporal resolution. Observations tions substantially superior to those made in the past
from Earth orbiting satellites are complementary to when the observational systems and the discipline of
local and regional airborne observations, and to tradi- natural hazards research were less mature than they
tional in situ field measurements and ground-based are today. Furthermore, assimilation of data and
sensor networks in seismology, volcanology, geo- model outputs into decision support systems must
morphology and hydrology. The contributions of meet operational requirements for accuracy, spatial
satellite remote sensing to solid Earth science, ranging coverage and timeliness in order to have positive
from high-resolution topography (using e.g. Interfero- impact on disaster risk management.
metric SAR, Lidar and digital photogrammetry) and In the following sections, satellite remote sensing
geodesy to passive hyperspectral (such as ASTER, systems and integrated observational and modeling
MODIS and Hyperion) and active microwave imaging, approaches that are being used to study and assess risks
have transformed the discipline. This transformation due to earthquakes, volcanoes, floods, landslides and
has helped to define a rapidly growing field of applied coastal inundation are reviewed briefly. Only hazards
research that increasingly will provide geospatial manifested through solid Earth processes are the focus
information products addressing the operational herein, although significant advances in the application
requirements of multi-hazard decision support tools of satellite remote sensing to severe weather events and
and systems. Policy makers, emergency managers and wildfires, for example, indeed are being made but left
responders from international and federal to state, for review elsewhere. This paper is not intended as a
regional and local jurisdictions use these tools and comprehensive treatment of the application of remote
systems to generate scenarios, devise mitigation plans sensing to the selected natural hazards but rather as a
and implement effective response measures. recognition of the contributions of satellite remote
Satellite remote sensing, in particular, is providing sensing to understanding underlying phenomena and
a systematic framework for scientific knowledge of providing critical information for decision support by
the solid Earth that is the basis for better-informed emergency managers and the disaster response com-
 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198 187

munity. As such, it is a summary of work reported by standing in constructing probabilities of occurrence;

the solid Earth science community in natural hazards characterizing the three-dimensional material proper-
research and applications (NRC, 2003; SESWG ties of fault systems, their response to deformation and
Report, 2002), with reference to the latest results in the physics of earthquake nucleation, propagation and
the literature and to seminal planning documents arrest; and predicting strong ground motions and non-
developed by the Subcommittee on Disaster Reduction linear surface layer response, including fault rupture,
(SDR), National Science and Technology Council of landsliding and liquefaction (NRC, 2003).
the U.S. Committee on the Environment and Natural Integrated ground measurements and satellite
Resources (CENR) (CENR, 2003), the Geohazards remote sensing can help meet these various require-
Theme Report of the Integrated Global Observing ments for baseline and time-series data. The Global
Strategy (IGOS) international partnership (IGOS Geo- Positioning System (GPS), for example, used for
hazards Theme Report, 2004), and the Disaster navigation and positioning in civilian and military
Management Support Group (DMSG) Report of the applications, provides the millimeter-level differential
Committee on Earth Observation Satellites (CEOS) accuracy that is used by regional ground deformation
within IGOS (CEOS, 2003). The intent is to underscore networks to monitor interseismic ground deformation
the view of the Earth as a complex system of forcings and co-seismic displacements. GPS monitoring net-
and responses, the increasingly recognized potential of works are leading to better definition of off-fault
satellite remote sensing to assess the consequences of surface deformation rates; timely detection of diag-
resultant changes, and the need for advanced observa- nostic changes in the fault environment; and constraints
tion platforms and monitoring systems for sustained on the extent of surficial fault creep and its significance
measurements of the solid Earth relevant to disaster to potentially significant earthquakes. Accurate esti-
management. mates of the distribution of potentially damaging
ground motions from such earthquakes enable ground
motion modeling, structural design planning and risk
2. Earthquake hazards assessment for loss estimation. Earthquake forecasting
efforts are related intimately to generating this
Deformation at the Earth’s surface, predominantly increased understanding of the fundamental dynamics
adjacent to tectonic plate boundaries, is the manifes- of major faults, with fault segment definition leading to
tation of forces acting deep within its interior. Geo- a better description of the expected details of earth-
detic and seismological measurements provide the quake faulting and rupturing. Moreover, GPS and
principal data for understanding mantle dynamics, modern digital seismic data can be combined with
lithospheric processes and crustal response, and for satellite remote sensing, such as interferometric syn-
improving numerical modeling for forecasting cata- thetic aperture radar (InSAR) to provide spatially
strophic events such as earthquakes and volcanic continuous deformation with sub-centimeter accuracy.
eruptions. Major advances have been made in earth- Fig. 1 shows a comparison of an InSAR-derived
quake research and risk mitigation. However, the deformation time series with GPS network time series
nature of significant seismic events–with greater and data, providing insight into the spatial and temporal
more widespread occurrence than volcanic hazards relationship between continuous GPS data and InSAR.
and resultant loss impact that is much higher than that The availability of both temporal and spatial deforma-
of even more widespread landslide and subsidence tion data allows for a greater level of understanding of
hazards (IGOS Geohazards Theme Report, 2004)– the dynamics of the area. InSAR techniques are an
presents extant research needs. Research requirements integral component of an Earth observation capability
in earthquake science that will contribute to better for seismic risk mitigation and response (CEOS, 2003).
seismic risk management and forecasting on a global Delineation of seismic source zones requires
basis include documenting the location, slip rates and understanding the geology, tectonics, and paleoseis-
earthquake history of dangerous faults; understanding mic and neotectonic features of the subject region.
the kinematics and dynamics of active fault systems Investigations of surface deformation, plate-boundary
on interseismic time scales and applying this under- interactions, frictional properties of faults, and
188 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198

Fig. 1. (Left) InSAR time series inversion map of the radar line-of-sight deformation average velocity, overlying the multi-look SAR amplitude
image (gray scale). Small black squares mark Southern California Integrated GPS Network (SCIGN) GPS site locations. (Right) InSAR time
series (black triangles) at selected points. Plots compare the InSAR time series to SCIGN GPS (red*) time series for indicated sites (produced
from European Space Agency remote sensing data, ERS-1 and ERS-2. Figure courtesy P. Lundgren, NASA/JPL; see also Lanari et al., 2004).

mechanical properties of the Earth’s crust and litho- the resultant patterns of damage (see Lohman et al.,
sphere help to determine what controls the spatial and 2002). This is critical for effective risk management.
temporal characteristics of earthquakes. For example, While there are in operation notably successful
surface displacements due to the 2003 Bam, Iran, dense GPS geodetic networks in regions prone to
earthquake were mapped using ENVISAT radar data potentially catastrophic seismic and/or volcanic
to reveal that over 2 m of slip occurred at depth on a events, such as southern California (Hudnut et al.,
blind strike-slip fault, where no morphological fea- 2002) and particularly Japan (Shimada and Bock,
tures were present (Talebian et al., 2004). Space-based 1992), economic constraints limit the widespread
observations of the entire earthquake cycle, including global deployment of these networks. Furthermore, a
the aseismic accumulation of strain between events lack of standard formats and established archives, plus
(Fielding et al., 2004) are critical for learning about limited accessibility for the different kinds of defor-
the phenomenology and for forecasting potentially mation data are major challenges for the integration of
hazardous earthquakes. Remote and in situ data that local GPS data globally, and the integration of GPS
support attendant scientific and engineering models data with older, heritage deformation data sets (IGOS
are necessary in order to understand the source- Geohazards Theme Report, 2004). The ability of
rupture process, fault plane geometry and thus infer geodetic data to resolve variations in slip patterns also
 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198 189

diminishes greatly with slip depth. Prediction of high, particularly near plate boundaries (Bowin,
strong motion velocities from geodetic data alone 1991), are poorly sampled by short period waves
offers limited spectral response, thus being a poor and global gravity data from abovementioned satel-
replacement for actual strong motion recordings that lite systems are spatially continuous and uniquely
are critical to earthquake engineering. In either case, complementary.
the spectrum of ground displacement signatures Finally, observations of land cover, land use and of
measurable by GPS and seismic networks is sampled the built environment, structures and lifelines specif-
discretely. ically are a critical component of risk management
Satellite remote sensing systems not only offer and loss estimation methodologies. Integration of
spatially continuous information of the tectonic land- high-resolution satellite remote sensing with InSAR
scape but also contribute to the understanding of and airborne Light Detection and Ranging (LiDAR) is
specific fault systems. Combined with ground net- being researched as a means to image, classify and
work data, remote sensing enables a better under- inventory the built environment through the extraction
standing of displacements, and validation of slip of land cover and digital terrain models (Gamba and
models that are cast in a regional setting of tectonic Houshmand, 2002). This contributes to vulnerability
strain (Cakir et al., 2003) and help constrain source assessments and to rapid post-disaster damage assess-
characterization (e.g. Lundgren and Stramondo, ment (Rejaie and Shinozuka, 2004), through integra-
2002). Satellite remote sensing observations are tion with demographic data, infrastructure and
providing insights into how stress is transferred building stock databases in a geographic information
between fault systems from depth and to the surface, system (GIS). Rapid damage assessment is critical for
how much energy is released by earthquakes and effective allocation of disaster response and relief
other modes of deformation (Argus et al., in press) resources, including federal insurance assistance.
and how faults fail mechanically. Available satellite remote sensing systems, from civil
In certain cases, earthquakes can produce global space agencies and commercial imaging sources such
gravity perturbations that are detectable through as IKONOS, OrbView and QuickBird, are witnessing
analysis of satellite data from missions dedicated to increased utilization in disaster management research
gravity field determination, such as the Challenging and operational domains.
Mini-satellite Payload (CHAMP), Gravity Recovery
and Climate Experiment (GRACE) and the Gravity
Field and Steady-State Ocean Circulation Explorer 3. Volcanic hazards
(GOCE). Coseismic gravity and geoid changes differ
from other, larger and more coherent high-frequency Subaerial volcanic eruptions occur often after long
variations such as Earth tides. The coseismic effects intervals of dormancy and thus opportunities for direct
of great earthquakes such as the 1960 Chile, 1964 geophysical measurements are intermittent and spora-
and 2002 Alaska, and 2003 Hokkaido events cause dic. While there are numerous indicators of subaerial
global gravitational field changes that are sufficiently volcanic activity, in addition to surface deformation
large to be detected by GRACE, for example, based and seismicity–such as thermal emissions, and
on degree variance analysis using spherical harmonic changes in gravity, emission of gasses plus ash and
representation of dislocation theory (Sun and Okubo, clastic eruptions–little is known about the global
2004) and a normal mode technique comparing the levels of these activities and how these phenomena
degree amplitude spectra of select earthquakes with are related. Furthermore, the physical mechanisms
GRACE sensitivity (Gross and Chao, 2001). Such that cause surface deformation and those that control
gravity data also can provide important constraints the rates and styles of eruptions are poorly under-
on the interpretation of seismological data (Tondi et stood. The ability to predict or otherwise forecast the
al., 2003), as for testing the shorter wavelength timing, magnitude, and style of volcanic eruptions on
features of three-dimensional tomographic models the Earth’s land surface is an important yet generally
based on the inversion of short period seismic waves. unmet objective in volcanic hazards assessment and
Many parts of the lithosphere, where heterogeneity is mitigation planning (SESWG Report, 2002).
190 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198

An observation strategy that incorporates seismic kilometers (Glaze and Baloga, 2003). The accuracy of
and in situ monitoring, geodetic measurements, high- these models depends on the accuracy of the available
resolution topography [i.e. Digital Terrain Elevation topographic data; measurements from the Shuttle
Data (DTED) Level 1) and hyperspectral imaging can Radar Topography Mission (SRTM) provide invalu-
contribute significantly to volcanic hazard risk assess- able data, particularly in remote regions where high-
ment, mitigation and response. Observational require- resolution topographic data are unavailable (e.g.
ments for volcanic hazards include the three- Stevens et al., 2003). A monitoring strategy that
dimensional spatial distribution of seismicity; the includes the use of such predictive flow models and
characteristic deformation of the volcanic edifice risk zonation could support land use planning,
using geodetic and gravimetric techniques that include particularly in developing countries where the impact
identification of faults, fractures, landslides and flank of volcanic hazards to populations and lifelines is
instabilities, rift systems and calderas; characterization continually increasing.
of gas and ash emissions by species (SO2, CO2) and Remote sensing indeed is defining a new paradigm
flux; and characterizing and monitoring of thermal for volcanological observations (Pieri and Abrams,
features, their nature, location, temperature and 2004). Imaging spectroscopy (or hyperspectral imag-
possible heat flux (CEOS, 2003). ing) in both the solar-reflected (0.4–2.5 Am) and
Fundamental to the understanding of eruptive thermal portions (3–5 Am and 8–12 Am) of the
systems are the identification and characterization of spectrum, permits the identification, separation, and
active volcanoes—namely a comprehensive global measurement of subtle variations reflecting the over-
inventory. This would comprise not only geodetic lapping molecular absorption and constituent scatter-
observations, with InSAR for example (Pritchard and ing signatures of materials present on the Earth’s
Simons, 2004), but spectroscopic observations of surface. Measurements of surface deposits and com-
debris flows and land surface, as with Hyperion–the position, surface temperature, topography and surface
first spaceborne imaging spectrometer–onboard the deformation, SO2 and ash detection and tracking (Fig.
NASA Earth Observing-1 (EO-1) satellite (Crowley et 2), and modeling are needed to better characterize,
al., 2003; Wright and Flynn, 2003), and ash and understand and predict the volcanic hazards environ-
emissions (see Fig. 2), as with the NOAA Geosta- ment. Measurements made by the Airborne Visible/
tionary Operational Environmental Satellite (GOES) Infrared Imaging Spectrometer (AVIRIS) and by
(Ellrod et al., 2003) and the Advanced Spaceborne Hyperion have been used to map subtle changes in
Thermal Emission and Reflection Radiometer near-surface rock chemistry and, thereby, identify
(ASTER) (Watanabe and Matsuo, 2003; Pieri and zones of volcanic-debris-flow susceptibility on the
Abrams, 2004). Such measurements and new obser- basis of rock strength inferred from specific minera-
vational tools are enabling a rapid growth in the logical indicators of hydrothermal alteration (Crowley
understanding of volcanic hazards worldwide. et al., 2003).
Geodetic observations of volcanoes with GPS and Global monitoring, including of remote areas, at
InSAR are yielding high-resolution digital elevation weekly time intervals with spaceborne systems would
models (DEMs) (Lu et al., 2003) and full vector enable the requisite sensitivity to low-level but more
deformation rate maps that complement traditional nearly continuous processes for assessment of risks in
ground-based geodetic techniques. These high-reso- short-term early warning systems. In the event of an
lution measurements are required in order to reduce eruption, shorter time intervals are desired, with
ambiguities in inferences of magma chamber geom- updating several times per day. However, in these
etry from outward structural changes (e.g. Lanari et cases, only a spotlight view of a targeted area of the
al., 2003). Computer models of a variety of flows are globe is needed, for example to provide volumetric
increasingly being use in volcanic hazard assessment estimation of eruptive lava outflow and source
to predict potential areas of devastation (Stevens et al., modeling (Lundgren and Rosen, 2003), distinguishing
2003). Furthermore, DEMs are being used to predict between surface deformation caused by magma
lava flow and lahar paths on remote volcanoes, with a movement or fluid pressure build-up. A similar
promising level of accuracy over distances of tens of rationale holds for the timing of spectroscopic
 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198 191

Fig. 2. Mt. Etna is one of the world’s most active volcanoes and has been studied for centuries from the ground. On November 3, 2002 Mt.
Etna’s ash-laden plume was imaged by ASTER. The plume is seen blowing towards the south–southeast, over the city and airport of Catania,
Sicily. The previous day, the plume was blowing towards the northwest, and posed no hazard to Catania. The eruption of Mt. Etna, Europe’s
most active volcano, began on October 27. The image covers an area of 50.876.5 km [image courtesy M. Abrams, NASA/JPL].

measurements that provide sensitivity to heat flux and Resolution Imaging Spectroradiometer (MODIS)
gas emissions (e.g. SO2 and CO2) (Prata et al., onboard the NASA Terra and Aura spacecraft are
submitted for publication). Proper temporal resolu- providing imagery of subaerial volcanoes on Earth
tion, temperature change sensitivity of the order of 0.5 every 2 days (Watson et al., 2004). There is now an
K and accurate measurements of gas emissions, along online archive of eruptions going back nearly 5 years,
with surface deformation maps, may allow the using MODIS data for global monitoring through
forecasting of eruptions. For example, the Moderate detection of thermal signatures (Wright et al., 2002,
192 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198

2004). Retrieval algorithms are able to quantify monitor post-slide motion and characterize debris size
volcanic ash, ice, sulfates and SO2 using thermal and distribution (Singhroy, 1995).
infrared (8–12 Am). Identification of outgassed Contributions of spaceborne remote sensing to
species near vents and craters provides information flood warning, disaster assessment and hazard reduc-
on subsurface activity and processes and may tion will rely on a broad-based program of remotely
ultimately assist in forecasting eruptions. Thermal sensed and in situ measurements of rainfall, river
measurements of land surface temperature, together heights, soil moisture, with vegetation change provid-
with simultaneous measurements of the changing ing critical indices for flood and landslide hazards.
emissivity, provide additional constraints on mag- Integration of remote sensing and in situ measurements
matic processes and volcanic activity. Even the is needed, along with hydrologic models benefiting
contribution of subaerial active volcanoes to the from improvements in multi-scale observations of
Earth’s energy budget can be estimated (Wright and climate and weather, from global to synoptic and
Flynn, 2004). mesoscale to storm scales. For example, the SeaWinds
radar aboard QuikSCAT and MODIS optical data are
processed and combined with a GIS for monitoring
4. Flood and landslide hazards flood propensity and developing weekly surface water
anomaly maps (Fig. 3) that emphasize the sustained
Floods are among the most devastating natural excess moisture receipts most likely to cause river
hazards in the world, claiming the largest amount of flooding (see www.dartmouth.edu/~floods). Occa-
lives and property damage (CEOS, 2003). Remotely sional (5–10 years) quantification of soil composition
sensed data play an integral role in reconstructing the and thickness would suffice in areas governed by
recent history of the land surface and in predicting gradual processes, but more frequent measurements
hazard events such as floods and landslides, subsidence will be needed in areas affected by such dynamic
events and other ground instabilities. Reconstruction of events as floods or landslides.
past erosion, deformation, and deposition and quanti- Diverse observational requirements are imposed by
fication of tectonic, climatic, and biologic inputs– the need for surface, subsurface and hydrologic
including human-induced changes–to the evolving characterization, including the delineation of flood
landscape underpin the ability to develop a process- and landslide zones for risk assessments (see Carrasco
based understanding of the Earth’s dynamic surface. et al., 2003) and mitigation planning, and zones prone
Since land-surface properties change through time, to subsidence due to groundwater interactions (see
remote sensing of such changes yields critical Buckley et al., 2003). The types of measurements that
temporal control on landscape evolution. The need are needed to quantify, model, and predict flood
for higher spatial and temporal resolution data is hazards include 1-m DEMs with 5 cm accuracy for
pinpointed by recognizing that destructive floods or catchment geometry and hill-slope angles used for
landslides can be launched by intense, short-lived water routing and landslide threshold assessments;
storm cells a few kilometers in extent. The height and hourly measurements of rainfall intensity and duration
width of rivers, as well as rainfall intensity and with 1–2 mm accuracy; and 12-hourly measurements
amounts, need to be measured hourly during storms. of soil moisture to assess infiltration and runoff
Hossain and Anagnostou (2004) give an assessment potential. Seasonal measurements of vegetation cover
of the current state of passive microwave and infrared- and canopy structure provide for water interception
based satellite systems for flood prediction. Few data and soil strength assessment, while 5-m resolution
exist on soil moisture, thickness, and strength, or on geologic mapping provides an overview of rock
vegetation cover, fire history, or detailed topography. strength, permeability and erosion potential. In con-
Synthetic aperture radar (SAR) and Landsat Thematic junction, multi-channel and multi-sensor data from
Mapper (TM) data have been integrated to provide meteorological satellites are assimilated into numer-
information on land cover and the geomorphology of ical weather prediction models to estimate precipita-
slopes, to inventory and characterize landslide poten- tion intensity, amount and coverage, winds and other
tial in high relief areas (Singhroy et al., 1998), and to factors that impact the severity of flood hazards.
 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198 193

Fig. 3. Wet soil conditions of the Tisza River Basin in Eastern Europe from analysis of QuickScat and MODIS data [image courtesy S.V.
Nghiem, NASA/JPL and R. Brakenridge, Dartmouth University].

The principal contribution of Earth observation information on flow emplacement parameters (i.e.
data is to provide the morphological, land use and rate, velocity and rheology), and factors such as
geological detail to help determine how a landslide lithology, location of faults, slope, vegetation and land
failed and the cause of failure (CEOS, 2003). GIS is use. The remote sensors that increasingly will provide
being used increasingly for regional risk assessment, flood and landslide hazard monitoring data include
including the integration of inventory mapping, InSAR, GPS (Malet et al., 2002; Gili et al., 2000),
location of surface structures and roughness providing visible and near infrared/ thermal infrared (VNIR/
194 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198

TIR) imaging, multi-parameter SAR, laser altimetry, world’s 15 largest cities lie along the coast or on
and microwave imaging. SAR data can provide an all- estuaries. About 53% of the US population, for
weather flood mapping capability and can be useful example, lives near the coast (Small et al., 2000).
for the estimation of hydrological parameters such as Any short-term or long-term sea-level change relative
soil moisture (soil surface wetness), wet snow to vertical ground motion is of great socioeconomic
mapping, and the monitoring of wetlands and flood concern, yet no accurate estimate of the vulnerability
extent delineation (CEOS, 2003). InSAR offers the of coastal communities exists.
capability for measuring displacements and providing The effects of sea level rise are spatially non-
very high accuracy topographic mapping. However, uniform due to local coastal variables, such as
even with the integration of in situ measurements, the interactions between lithology, geomorphology, and
ability for a predictive capability for the occurrence wave climate, currents and storm frequencies (Gor-
and extent of landslide impact falls behind that for nitz, 1991). Paleo-environmental and historical data
mitigation planning (see CEOS, 2003). A thorough clearly have indicated the occurrence of such changes
evaluation of erosion hazards in the United States in the past, and the potential impact under enhanced
related to coastal processes and flooding is provided greenhouse conditions (Mcinnes et al., 2003). Sea
by the Heinz Center (2000), ranging from risk level rise itself is not globally uniform, as Fig. 4
assessment, economic impact and insurance programs shows. The TOPEX/POSEIDON altimetric satellite
to management and policy. Satellite-based observa- now in its 12th year of operation has established this
tions will need to be augmented with extensive land- remarkable record with the most accurate measure-
based measurements and data from existing and ments of sea surface topography to date. Its successor
future, integrated hydrologic and geodetic arrays. Jason-1, and the planned NASA Ocean Surface
Topography Mission (each a collaborative mission
between the US and French space agencies), as well
5. Coastal inundation as ERS-1 and -2, and recently ENVISAT, of the
European Space Agency, all contribute to monitoring
Atmospheric and oceanic processes have a signifi- this crucial quantity.
cant effect on coastal geomorphology. Sea level rise as Sea level rise does not just passively inundate low-
a consequence of global climate change represents an lying coastal regions. Sea level rise, as a symptom of
enormous risk to coastal populations. Eleven of the climate change, and changes in storm frequency or

Fig. 4. Trends in sea level derived from TOPEX/POSEIDON data for the period 1993–2003. Inverse barometer (IB) correction applied [see
http://www.sealevel.colorado.edu and Leuliette et al. (2004)].
 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198 195

intensity lead to greater erosion of the coasts and shows the effect of a 5-m sea level rise on the Florida
accelerate the process whereby the ocean inundates coast of the US. Flood damages would increase
what was previously land surface. The Intergovern- accordingly and contribute to higher sediment depos-
mental Panel on Climate Change (IPCC) has outlined its at inlets, further exacerbating the inundation hazard
major impacts of rising sea level on coastal commun- risk.
ities (IPCC, 2001), which include beach erosion, An example of an integrated observation strategy
inundation of land and increased flood and storm consists of space measurements of ocean vector winds
damage. The IPCC reports that 1 cm rise in sea level (SeaWinds on QuickSCAT and ADEOS-2; ERS-1 and
erodes beaches about 1 m horizontally; a 50-cm rise in -2 and now Meteosat) to assess the strength of storms
sea level will inundate 8500–19000 km2 of dry land; at sea from their surface wind vectors; in addition,
and a higher sea level will provide a higher base for Tropical Rainfall Mapping Mission (TRMM) meas-
storm surges. A 1-m rise in sea level would enable a urements of the precipitation associated with storms,
15-year storm to flood areas that today are only with NASA Atmospheric Infrared Sounder (AIRS),
flooded by 100-year storms (IPCC, 2001). Fig. 5 MODIS and other instruments observing the cloud

Fig. 5. The potential rise in sea level caused by melting of the Greenland ice sheets seriously jeopardizes low-lying areas such as the Florida
coast. Red shows where land would be submerged for an estimated 5-m sea-level rise [image courtesy M. Kobrick, NASA/JPL].
196 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198

patterns above the storm, and passive radiometers often lead to disasters. Satellite remote sensing data
measuring the water vapor around the storm. By and derived geospatial products increasingly are
studying changes in storm frequency and intensity, for complementing ground-based network data, and in
example, better estimates can be made of what future situ and field observations for disaster assessment
climate change will bring to coastal regions. Inte- and response. Many advances in satellite remote
grated SAR and TM imagery have been used to sensing have been and will continue to be made as
monitor changes in coastal geomorphology and land various resources are secured for technology and
cover, flood and erosional damage, and to facilitate infrastructure development, in efforts towards bridg-
planning and maintenance of mitigations (Singhroy, ing the transition from natural hazards research to
1995). Advances to-date in the science and technol- enhanced operational capability in disaster manage-
ogy of shoreline change mapping and projection of ment. The Advanced Synthetic Aperture Radar
future shoreline positions are elements of erosion (ASAR) operating at C-band, onboard ENVISAT,
hazard zone identification under the FEMA National launched by the European Space Agency in March
Flood Insurance Program (Leatherman, 2003). The 2002, is an enhanced capabilities continuation of the
U.S. Environmental Protection Agency has studied ERS-1/2 that has as one of its mission objectives the
the environmental impact and economic costs asso- monitoring of earthquake and volcanic hazards. The
ciated with coastal inundation due to the greenhouse Japanese Advanced Land Observing Satellite
effect and sea level rise (Titus et al., 1991). These (ALOS) includes a panchromatic stereo imager for
assessments have direct bearing on coastal land-use digital elevation mapping, a visible and near I/R
planning for risk mitigation. radiometer and phased array L-band SAR (PAL-
In addition to the continual effects of atmospheric SAR). The PALSAR system is designed to provide
and oceanic processes on coastal geomorphology, continuous images of land deformation. Moreover,
earthquakes underneath or near the ocean (particu- tools to produce information products through
larly at deep ocean trenches and island arcs) and less integration, assimilation, modeling and realistic com-
commonly submarine landslides and volcanic erup- putational simulations must continue to be devel-
tions can generate tsunamis with the potential to oped, addressing issues of data access continuity,
inundate the coast. The magnitude 9.0 Indonesian completeness, interoperability and validation.
earthquake of December 26, 2004 off the west coast The solid Earth science research community must
of northern Sumatra is a recent and indeed historic continue to demonstrate the potential of these remote
example of extreme coastal inundation resulting from sensing systems and derived products for operational
generation of tsunamis. The devastating losses to life decision-making that impacts the ability to reduce
and property from this great event are a reminder of losses to life and property. Assimilation of science,
the potential of natural hazards to change the land- model outputs and satellite data into decision support
scape and calls attention to the need for integrated tools and systems through applications, validation and
monitoring systems, including in situ and orbiting performance benchmarking is a critical step. Policy-
sensors, real-time communications, extremely fast and decision-makers, emergency managers and res-
assessment, and prearranged communication lines to ponders, in turn, will use the enhanced decision
those in the best position to warn populations support systems, geospatial information products,
throughout the globe, where the socioeconomic model-based forecasts and visualizations in long-term
consequences of catastrophic disaster events are planning of emergency services and lifelines, com-
widespread. prehensive disaster assessment and response prioriti-
zation. Cost-effective approaches will be necessary,
with the participation of the commercial sector in
6. Conclusions distinct elements of an overall observational architec-
ture, as the resources available are limited. Interna-
Integrated satellite-based observations and numer- tional partnerships and cross-agency relationships can
ical modeling are leading to new levels of under- be expected to increasingly enable civil space
standing of the complex solid Earth processes that agencies to develop a broad range of observations
 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198 197

and shared data sets that respond to the needs of the Donnellan, A., Rundle, J., Ries, J., Fox, G., Pierce, M., Parker, J.,
operational and policy communities in disaster man- Crippen, R., DeJong, E., Chao, B., Kuang, W., McLeod, D.,
Matu’ura, M., Bloxham, J., 2004. Illuminating Earth’s interior
agement while addressing key questions in the solid through advanced computing. Comput. Sci. Eng. 6 (1), 36 – 44.
Earth research community. Ellrod, G.P., Connell, B.H., Hillger, D.W., 2003. Improved
detection of airborne volcanic ash using multispectral infrared
satellite data. J. Geophys. Res., [Atmos.] 108 (D12) (Art. No.
Acknowledgements 4356).
Fielding, E.J., Wright, T.J., Muller, J., Parsons, B.E., Walker, R.,
This review paper was written at the Jet Propulsion 2004. A seismic deformation of a fold-and-thrust belt imaged by
SAR interferometry near Shahdad, SE Iran. Geology 32 (7),
Laboratory, California Institute of Technology, under 577 – 580.
contract to the National Aeronautics and Space Gamba, P., Houshmand, B., 2002. Joint analysis of SAR, LIDAR
Administration, and was guided in part by the NASA and aerial imagery for simultaneous extraction of land cover,
Solid Earth Science Working Group (SESWG) DTM and 3D shape of buildings. Int. J. Remote Sens. 23 (20),
Report, bLiving on a Restless Planet,Q 2002; the 4439 – 4450.
Gili, J.A., Corominas, J., Rius, J., 2000. Using global positioning
CEOS hazard reports and the CENR Interim Report of system techniques in landslide monitoring. Eng. Geol. 55 (3),
the Subcommittee on Disaster Reduction. We thank 167 – 192.
the reviewers of the initial manuscript for their Glaze, L.S., Baloga, S.M., 2003. DEM flow path prediction
comments and recommendations; and Mike Abrams, algorithm for geologic mass movements. Environ. Eng. Geosci.
Paul Lundgren and Son Nghiem for sharing some data 9 (3), 225 – 240.
Gornitz, V., 1991. Global coastal hazards from future sea level rise.
analyses and discussion. Glob. Planet. Change 89 (4), 379 – 398.
Gross, R.S., Chao, B.F., 2001. The gravitational signature of
earthquakes. In: Sideris, M.G. (Ed.), Proc. IAG International
References Symposium Gravity, Geoid and Geodynamics (GGG2000),
Banff, Alberta, Canada, July 31–August 4, 2000, vol. 123.
Argus, D., Heflin, M., Peltzer, G., Crampe, F., Webb, F., in press. Springer-Verlag, New York, pp. 205 – 210.
Interseismic strain accumulation and anthropogenic motion in Heinz Center, 2000. Evaluation of Erosion Hazards. Report to the
metropolitan Los Angeles, J. Geophys. Res. Federal Emergency Management Agency, The H. John Heinz III
Bowin, C., 1991. The Earth’s gravity-field and plate-tectonics. Center for Science, Economics and the Environment, Wash-
Tectonophysics 187 (1–3), 69 – 89. ington D.C. http://www.heinzcenter.org (accessed 22 Dec.
Buckley, S.M., Rosen, P.A., Hensley, S., Tapley, B.D., 2003. Land 2004).
subsidence in Houston, Texas, measured by radar interferometry Hossain, F., Anagnostou, E.N., 2004. Assessment of current
and constrained by extensiometers. J. Geophys. Res. 108 (B11) passive-microwave and infrared-based satellite rainfall remote
(Art. No. 2542). sensing for flood prediction. J. Geophys. Res. 109 (D7) (Art.
Cakir, Z., Chabalier, J.B., Armijo, R., Meyer, B., Barka, A., No. D07102).
Peltzer, G., 2003. Coseismic and early post-seismic slip Hudnut, K.W., Bock, Y., Galetzka, J.E., Webb, F.H., Young,
associated with the 1999 Izmit earthquake (Turkey), from W.H., 2002. The Southern California Integrated GPS Net-
SAR interferometry and tectonic field observations. Geophys. J. work (SCIGN). In: Fujinawa, Y., Yoshida, A. (Eds.), Seismotec-
Int. 155 (1), 93 – 110. tonics in Convergent Plate Boundary. TERRAPUB, Tokyo,
Carrasco, R.M., Pedraza, J., Martin-Duque, J.F., Sanz, M.A., Japan, pp. 167 – 189.
Bodoque, J.M., 2003. Hazard zoning for landslides connected IGOS (Integrated Global Observing Strategy) Geohazards Theme
to torrential floods in the Jerte Valley (Spain) by using GIS Report, 2004. http://www.dup.esrin.esa.it/igos-geohazards/pdf/
techniques. Nat. Hazards 30 (3), 361 – 381. igos_report.zip (accessed 22 Dec. 2004).
CENR, 2003. Reducing Disaster Vulnerability Through Science and IPCC, 2001. Climate change 2001-impacts, adaptation and vulner-
Technology. Interim Report of the National Science and ability. Contribution of working group II to the third assessment
Technology Council, Subcommittee on Disaster Reduction, report of the intergovernmental panel on climate change (IPCC).
July. In: McCarthy, James J., Canziani, Osvaldo F., Leary, Neil A.,
CEOS, 2003. The Use of Earth Observing Satellites for Hazard Dokken, David J., White, Kasey S. (Eds.), Cambridge Uni-
Support: Assessments and Scenarios. Final Report of the CEOS versity Press, UK. (1000 p. Available at http://www.grida.no/
Disaster Management Support Group (DMSG), November. climate/ipcc_tar/wg2/ (accessed 22 Dec. 2004)).
Crowley, J.K., Hubbard, B.E., Mars, J.C., 2003. Analysis of Lanari, R., Berardino, P., Borgstrom, S., del Gaudio, C., de
potential debris flow source areas on Mount Shasta, California, Martino, P., Fornaro, G., Guarino, S., Ricciardi, P., Sansosti, E.,
by using airborne and satellite remote sensing data. Remote Lundgren, P., 2003. The use of IFSAR and classical geodetic
Sens. Environ. 87 (2–3), 345 – 358. techniques for caldera unrest studies: application to the Campi
198 D.M. Tralli et al. / ISPRS Journal of Photogrammetry & Remote Sensing 59 (2005) 185–198

Flegrei uplift event of 2000. J. Volcanol. Geotherm. Res. 133 Shimada, S., Bock, Y., 1992. Crustal deformation measurements in
(1–4), 247 – 260. central Japan determined by a Global Positioning System fixed-
Lanari, R., Lundgren, P., Manzo, M., Casu, F., 2004. Satellite radar point network. J. Geophys. Res. 97 (B9), 12437 – 12455.
interferometry time series analysis of surface deformation for Singhroy, V., 1995. SAR integrated techniques for geohazard
Los Angeles, California. Geophys. Res. Lett. 31 (23) (Art. No. assessment, Natural Hazards: monitoring and assessment
L23613). using remote sensing techniques. Adv. Space Res. 33 (3),
Leatherman, S.P., 2003. Shoreline change mapping and management 205 – 290.
along the US East Coast. J. Coast. Res., 5 – 13. (Sp. Iss. 38). Singhroy, V., Mattar, K.E., Gray, A.L., 1998. Landslide character-
Leuliette, E.W., Nerem, R.S., Mitchum, G.T., 2004. Calibration of ization in Canada using interferometric SAR and combined
TOPEX/Poseidon and Jason altimeter data to construct a SAR and TM images. Adv. Space Res. 21 (3), 465 – 476.
continuous record of mean sea level change. Mar. Geod. 27 Small, C., Gornitz, V., Cohen, J.E., 2000. Coastal hazards and the
(1–2), 79 – 94. global distribution of population. Environ. Geosci. 7 (1), 3 – 12.
Lohman, R.B., Simons, M., Savage, B., 2002. Location and Stevens, N.F., Manville, V., Heron, D.W., 2003. The sensitivity of a
mechanism of the Little Skull Mountain earthquake as con- volcanic flow model to digital elevation model accuracy: experi-
strained by satellite radar interferometry and seismic waveform ments with digitized map contours and interferomegtric SAR at
modeling. J. Geophys. Res., [Solid Earth] 107 (B6) (Art. 2118). Ruapehu and Taranaki volcanoes, New Zealand. J. Volcanol.
Lu, Z., Patrick, M., Fielding, E.J., Trautwein, C., 2003. Lava Geotherm. Res. 119 (1–4), 89 – 105.
volume from the 1997 eruption of Okmok volcano, Alaska, Sun, W., Okubo, S., 2004. Coseismic deformations detectable by
estimated from spaceborne and airborne interferometric syn- satellite gravity missions: a case study of Alaska (1964, 2002) and
thetic aperture radar. IEEE Trans. Geosci. Remote Sens. 41 (6), Hokkaido (2003) earthquakes in the spectral domain. J. Geophys.
1428 – 1436. Res., [Solid Earth] 109 (B4) (Art. No. B04405).
Lundgren, P., Rosen, P.A., 2003. Source model for the 2001 flank Talebian, M., Fiedling, E.J., Funning, G.J., Ghorashi, M., Jackson,
eruption of Mt. Etna volcano. Geophys. Res. Lett. 30 (7), 1388. J., Nazari, H., Parsons, B., Priestley, K., Rosen, P.A., Walker, R.,
Lundgren, P., Stramondo, S., 2002. Slip distribution of the 1997 Wright, T.J., 2004. The 2003 Bam (Iran) earthquake: rupture of
Umbria-Marche earthquake sequence: joint inversion of GPS a blind strike-slip fault. Geophys. Res. Lett. 31 (11) (Art. No.
and synthetic aperture radar interferometry data. J. Geophys. L11611).
Res., [Solid Earth] 107 (B11) (Art. No. 2316). Titus, J.G., Park, R.A., Leatherman, S.P., Weggel, J.R., Greene,
Malet, J.P., Maquaire, O., Calais, E., 2002. The use of Global M.S., Mausel, P.W., Brown, S., Gaunt, G., Trehan, M., Yohe,
Positioning System techniques for the continuous monitoring of G., 1991. Greenhouse effect and sea level rise: the cost of
landslides: application to the Super-Sauze Earthflow (Alpes-de- holding back the sea. Coast. Manage. 19 (2), 171 – 204.
Haute-Provence, France). Geomorphology 43 (1–2), 33 – 54. Tondi, R., De Franco, R., Biella, G., 2003. Deep and shallow solid-
Mcinnes, K.L., Walsh, K.J.E., Hubbert, G.D., Beer, T., 2003. Impact earth structures reconstructed with sequential integrated inver-
of sea-level rise and storm surge on a coastal community. Nat. sion (SII) of seismic and gravity data. Space Sci. Rev. 108 (1–2),
Hazards 30 (2), 187 – 207. 115 – 130.
NRC (National Research Council), 2003. Living on an active earth: Watanabe, H., Matsuo, K., 2003. Rock type classification by multi-
perspectives on earthquake science. Committee on the Science band TIR of ASTER. Geosci. J. 7 (4), 347 – 358.
of Earthquakes. National Academies Press, Washington, D.C. Watson, I.M., Realmuto, V.J., Rose, W.I., Prata, A.J., Bluth, G.J.S.,
432 pp. Gu, Y., Bader, C.E., Yu, T., 2004. Thermal infrared remote
Pieri, D., Abrams, M., 2004. ASTER watches the world’s sensing of volcanic emissions using the moderate resolution
volcanoes: a new paradigm for volcanological observations imaging spectroradiometer. J. Volcanol. Geotherm. Res. 135 (1–
from orbit. J. Volcanol. Geotherm. Res. 135 (1–2), 13 – 28. 2), 75 – 89.
Prata, A.J., Watson, I.M., Rose, W.I., Realmuto, V., Bluth, G.J.S., Wright, R., Flynn, L.P., 2003. On the retrieval of lava-flow surface
Sevranckx, R., submitted for publication. Volcanic sulfur temperatures from infrared satellite data. Geology 31 (10),
dioxide measurements derived from infrared satellite measure- 396 – 893.
ments. Geophys. Res. Lett. Wright, R., Flynn, L.P., 2004. Space-based estimate of the volcanic
Pritchard, M.E., Simons, M., 2004. An InSar-based survey of heat flux into the atmosphere duing 2001 and 2002. Geology 32
volcanic deformation in the central Andes. Geochem. Geophys. (3), 189 – 192.
Geosyst. 5 (Art. No. Q02002). Wright, R., Flynn, L.P., Garbeil, H., Harris, A., Pilger, E., 2002.
Rejaie, A., Shinozuka, M., 2004. Reconnaissance of Golcuk 1999 Automated volcanic eruption detection using MODIS. Remote
earthquake damage using satellite images. J. Aerosp. Eng. 17 Sens. Environ. 82 (1), 135 – 155.
(1), 20 – 25. Wright, R., Flynn, L.P., Garbeil, H., Harris, A., Pilger, E., 2004.
SESWG (Solid Earth Science Working Group) Report, 2002. Living MODVOLC: near-real-time thermal monitoring of global
on a Restless Planet. S. C. Solomon, Chair, NASA Jet volcanism. J. Volcanol. Geotherm. Res. 135 (1–2), 29 – 49.
Propulsion Laboratory, Report JPL 400-1040, November.