Bringing science and development together through news and analysis

  • Remote sensing for natural disasters: Facts and figures

Artist's rendering of NASA's Landsat 7 satellite
Image credit: Wikimedia Commons

Sian Lewis explains how remote sensing can be used to help manage natural disasters and highlights ongoing efforts and obstacles.

Severe geo-physical or climatic events, including earthquakes, volcanic eruptions, landslides, droughts, floods, cyclones and fire that threaten people or property, are termed natural hazards.

When they destroy people's lives and livelihoods, they become natural disasters. Since the turn of the century, the Emergency Events Database (EM-DAT) has recorded an average of 397 disasters each year.

Developing countries suffer more than 95 per cent of all deaths caused by natural disasters (see Figure 1). Their high population densities and poor infrastructure, coupled with unstable landforms and exposure to severe weather events, makes them particularly vulnerable.  

Remote sensing — the science of acquiring information about the Earth using remote instruments, such as satellites — is inherently useful for disaster management. Satellites offer accurate, frequent and almost instantaneous data over large areas anywhere in the world. When a disaster strikes, remote sensing is often the only way to view what is happening on the ground.

Figure 1. Number of victims of natural disasters per 100,000 inhabitants 1976-2005 source: EM-DAT

Just last year (2008), natural disasters affected 214 million people, killed more than 235,000 and cost more than US$190 billion. [1]

Natural events can't be prevented, but potential disasters can be 'managed' to minimise loss of life through a four-part cycle of mitigation, preparedness, response and recovery (see Box 1).

Box 1: The four-part disaster cycle

  • Mitigation. Long-term efforts to prevent hazards from becoming disasters or make them less damaging. These include structural measures such as creating flood levees or reinforcing buildings, as well as non-structural measures such as risk assessment and land-use planning.
  • Preparedness. Planning for when disaster strikes, including developing communication strategies, early warning systems, and stockpiling supplies.
  • Response. Implementing plans after a disaster. This includes mobilising emergency services, coordinating search and rescue, and mapping the extent of the damage.
  • Recovery. Restoring an area, often through rebuilding and rehabilitation, then returning to mitigation measures.
Figure 2: The disaster management cycle

Roles for remote sensing

Remote sensing has many uses in disaster management, from risk modelling and vulnerability analysis, to early warning, to damage assessment (see Table 1).

Disaster

Mitigation

Preparedness

Response

Recovery

Cyclone

Risk modelling;

vulnerability analysis.

Early warning;

long-range climate modelling.

Identifying escape routes;

crisis mapping;

impact assessment;

cyclone monitoring;

storm surge predictions.

Damage assessment;

spatial planning.

Drought

Risk modelling;

vulnerability analysis;

land and water management planning.

Weather forecasting;

vegetation monitoring;

crop water requirement mapping;

early warning.

Monitoring vegetation;

damage assessment.

Informing drought mitigation.

Earthquake

Building stock assessment;

hazard mapping.

Measuring strain accumulation.

Planning routes for search and rescue;

damage assessment;

evacuation planning;

deformation mapping.

Damage assessment;

identifying sites for rehabilitation.

Fire

Mapping fire-prone areas;

monitoring fuel load;

risk modelling.

Fire detection;

predicting spread/direction of fire;

early warning.

Coordinating fire fighting efforts.

Damage assessment.

Flood

Mapping flood-prone areas;

delineating flood-plains;

land-use mapping.

Flood detection;

early warning;

rainfall mapping.

Flood mapping;

evacuation planning;

damage assessment.

Damage assessment;

spatial planning.

Landslide

Risk modelling;

hazard mapping;

digital elevation models.

Monitoring rainfall and slope stability.

Mapping affected areas;

Damage assessment;

spatial planning;

suggesting management practices.

Volcano

Risk modelling;

hazard mapping;

digital elevation models.

Emissions monitoring;

thermal alerts.

Mapping lava flows;

evacuation planning.

Damage assessment;

spatial planning.

Table 1: Ways remote sensing can help disaster management

Many types of satellites are used for earth observation but the area they see, and the frequency of observations, varies. Two complementary types are particularly relevant to disaster management. Polar-orbiting satellites fly in a relatively low orbit (often at around 1000km above the ground), providing relatively high spatial resolution. But they only collect data over the same point once every few days.

Geostationary satellites are positioned at a much higher altitude (about 36,000km). They orbit the Earth at the same speed as the Earth rotates on its axis, in effect remaining stationary above the ground and viewing the whole earth disk below. Their spatial data is much coarser, but is collected at the same point every 15 minutes.

Each satellite carries one or more sensors on board that take measurements in different wavelengths. Many are useful for disaster monitoring — thermal sensors spot active fires, infrared sensors can pick up floods, and microwave sensors (that penetrate clouds and smoke) can be used to measure earth deformations before and during earthquakes or volcanic eruptions (see Table 2). 

Wavelength

Waveband

Useful for

Example sensors

Visible

0.4-0.7mm

Vegetation mapping

SPOT; Landsat TM

Building stock assessment

AVHRR; MODIS; IKONOS

Population density

IKONOS; MODIS

Digital elevation model

ASTER; PRISM

Near infrared

0.7-1.0mm

Vegetation mapping

SPOT; Landsat TM; AVHRR; MODIS

Flood mapping

MODIS

Shortwave infrared

0.7-3.0mm

Water vapour

AIRS

Thermal infrared

3.0-14mm

Active fire detection

MODIS

Burn scar mapping

MODIS

Hotspots

MODIS; AVHRR

Volcanic activity

Hyperion

Microwave (radar)

0.1-100cm

Earth deformation and ground movement

Radarsat SAR; PALSAR

Rainfall

Meteosat; Microwave Imager (aboard TRMM)

River discharge and volume

AMSR-E

Flood mapping and forecasting

AMSR-E

Surface winds

QuikScat radar

3D storm structure

Precipitation radar (aboard TRMM)

Table 2: Applications of different wavebands for disaster management

Acronyms: Satellite Pour l'Observation de la Terre (SPOT); Thematic Mapper (TM); Advanced Very High Resolution Radiometer (AVHRR); Moderate Resolution Imaging Spectroradiometer (MODIS); Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER); Panchromatic Remote-sensing Instrument for Stereo Mapping (PRISM); Synthetic Aperture Radar (SAR); Phased Array type L-band SAR (PALSAR); Tropical Rainfall Measuring Mission (TRMM); Global Precipitation Measurement (GPM); Advanced Microwave Scanning Radiometer (AMSR-E); Atmospheric Infrared Sounder (AIRS)

Forecasting famine

Drought — and the famine it can cause — is a major disaster in the developing world, particularly Africa. Unlike many natural disasters, famine sets in slowly and can often be predicted months in advance.

Long-term climate forecasts, derived from satellite observations, can help build various scenarios before or during the early crop-growing season. Throughout the season, satellite rainfall data can help monitor growing conditions and predict soil moisture. And at the end of the season, satellite-observed vegetation can be used to check on likely crop production and yield. [2]

The Famine Early Warning Systems Network (FEWS NET), funded by USAID, monitors food security via satellites. It uses vegetation indices, calculated from sensors including AVHRR, MODIS and those aboard SPOT, to monitor vegetation vigour and density and spot problems as they develop.

It estimates rainfall using Meteosat infrared data, combined with rain gauge reports and microwave satellite observations, so as to model hydrological systems and how weather patterns might affect agriculture. FEWS NET also compares rainfall trends over time.

By combining satellite data with regional analyses of prices, grain stores, political conditions and agricultural inputs, FEWS NET provides effective early warnings for when drought might bring food shortages.

And FEWS NET is not the only initiative using satellite data for drought and famine prediction and monitoring. The Global Monitoring for Food Security, funded by the European Space Agency (ESA), similarly provides early warning of famine in sub-Saharan Africa.

The Agricultural, Hydrological and Meteorological Programme (AGRHYMET) in West Africa, the IGAD Climate Prediction and Applications Centre (ICPAC) run by East Africa's Intergovernmental Authority on Development, and the Southern African Development Community's Regional Remote Sensing Unit in South Africa also do similar jobs.

Response is the problem in this part of the world — getting the information into the hands of decision-makers and then implementing a plan for relief — not so much prediction.

Flood monitoring

Satellites can also warn of floods, and inform response and recovery efforts.

Satellites such as the Tropical Rainfall Monitoring Mission can measure and map rainfall, helping forecast heavy rains and floods. Sentinel Asia — a team of 51 organisations from 18 countries — delivers remote sensing data via the Internet as easy-to-interpret information for both early warning and flood damage assessment across Asia.

It uses the Dartmouth Flood Observatory's (DFO's) River Watch flood detection and measurement system, based on AMSR-E data, to map flood hazards and warn disaster managers and residents in flood-prone areas when rivers are likely to burst their banks.

NASA also uses DFO analyses for river basins across the world in its flood sensorWeb. The sensorWeb's role is to automatically alert disaster managers and government agencies to impending floods.

It detects anomalies in river discharges and volumes from the DFO's Active Atlas of Large Floods. This triggers requests to satellites such as MODIS for high-resolution data over the area of interest. These are then immediately processed and forwarded to scientists and local interested partners. [3]

SensorWebs are equally relevant to other disasters, including volcanoes, fires and dust storms — they just need different satellite data, depending on what variables are monitored.

The volcano sensorWeb for example, uses MODIS and AVHRR to detect volcanic activity based on thermal alerts. It looks for locations that are hot and different to the surrounding area (but not bright). Alerts then trigger observations from NASA's Hyperion satellite sensor, which is highly sensitive in thermal infrared. 

Fire mapping

Thermal alerts from MODIS also feed in to a fire sensorWeb. The MODIS Rapid Response System provides daily satellite images in near real time (within a few hours of data collection). These identify hotspots and trigger requests to other satellites to collect additional information on the active fire.

And MODIS produces global fire maps that show active fires over the past ten days (see Figure 3). This active fire mapping system is used by a wide array of fire monitoring programmes, including Sentinel Asia, the Global Fire Monitoring Center and the regional visualisation and monitoring system SERVIR that covers Latin America and the Caribbean.

Figure 3: MODIS global fire map 9–18 August, 2009

MODIS is also a key component of South Africa's Advanced Fire Information System (AFIS), where it is used to detect hotspots. Data are combined with wind vector information to calculate the trajectory of active fires. AFIS uses the information to drive an automatic alerting system to warn people when there is a fire coming their way (see Fires: Spotted from satellites, warned by phone).

Earthquake response

At present, earthquakes are hard to predict. But remote sensing could improve forecasts using Interferometric Synthetic Aperture Radar (InSAR). This technique combines two or more sequential radar images to measure ground motion between them very accurately — on the scale of a few centimetres (or even millimetres). InSAR instruments, such as PALSAR, are already routinely used after earthquakes to assess damage and the extent of ground movement and deformation.

Remote sensing really comes into play in facilitating emergency relief and assessing damage after an earthquake. High-resolution visible imagery from any number of satellites can help search and rescue teams navigate round cities, as well as improve estimates of economic losses.

The World Agency of Planetary Monitoring and Earthquake Risk Reduction (WAPMERR) uses remote sensing to improve knowledge of building stocks — for example the number and height of buildings. High resolution imagery can also help hazard mapping to guide building codes and disaster preparedness strategies.

BOX 2: Sichuan Earthquake

In May 2008 an earthquake measuring 7.9 on the Richter scale — the most powerful earthquake since 1976 — struck the Sichuan province of China. It killed more than 87,000 people and affected around 45 million across ten provinces. Around 12.5 million animals died and more than 26 million buildings were damaged (around 5 million suffered total collapse). Economic losses have been estimated at US$85 billion. [1]

The enormity of the disaster, heavy rainfall, the region's inaccessibility and the risk of aftershocks and landslides complicated rapid response efforts. Remote sensing data became indispensable in this crisis.

The National Disaster Reduction Center of China (NDRCC) took the lead in using satellite data to support emergency relief. Within half an hour of the earthquake it produced the first map of the damage. Over the next several weeks, the NDRCC received and processed nearly 1300 images from 22 sensors to monitor and evaluate the area. Several Chinese government departments and mapping experts across the world supported the work.

Remote sensing helped guide relief workers and identify and mitigate additional threats. Landslides formed over 30 natural dams in rivers, increasing the risk of flooding and debris flow — the dam that formed at Tangjiashan trapped water that threatened 1.3 million people. [4] Satellite imagery helped monitor these dams and direct evacuation efforts.

Data from after the quake have also been used for longer-term studies aimed at helping understand seismic cycles and how faults behave. A Euro-Chinese team under the ESA-funded Dragon 2 programme, for example, have used InSAR data to measure deformation and map ground displacement during and after the earthquake (see Figure 4). [5]

Figure 4: InSAR image of deformation during and after the May 2008 Sichuan earthquake. Rainbow fringes show ground displacement during and after earthquake. Credit: Jianbao Sun, IGCEA, Seismology and Geology, No. 3, 2008

 

Coping with cyclones

Meteorologists have used satellite images to monitor storms for decades. For example, the World Meteorological Organization's Tropical Cyclone Programme uses satellite observations, together with meteorological measurements and modelling, to produce cyclone warnings.

These estimate the storm's position, direction and speed, maximum wind speeds, areas likely to be affected, and likely storm surges. The programme issues these to government officials, river port authorities, the general public, coast guard, non-governmental organisations and cyclone preparedness programmes across the world.

The Bay of Bengal is particularly vulnerable to cyclones (see Cyclones in the Indian Ocean: Facts and figures), and India uses satellite data across a broad programme of preparation and response.

The country's Kalpana-1 and INSAT-3A satellites carry sensors that collect meteorological data in visible, near infrared and short-wave infrared. These provide data on cloud motion, sea surface temperature and rainfall. [6] A network of cyclone warning centres analyse the data and then issue timely warnings of impending cyclones. The warnings give information on the cyclone itself as well as likely damage and suggested action.

When a cyclone threatens, the centres issue bulletins on its position, wind speed, pressure and development characteristics every hour.

Equal access?

The UN Economic Commission for Africa (UN ECA) argues that having timely access to remote sensing data is a powerful tool for regional sustainable development. [7] And in principle, remote sensing offers developed and developing countries the same quality and frequency of data. But cost is still a barrier. [7]

Several initiatives are working to overcome this. For example, the International Charter on Space and Major Disasters — established in 1999 and now signed by nearly 20 space agencies and organisations — provides free data to any country suffering from a natural disaster. As soon as a disaster strikes, any authorised user, which includes civil protection agencies, rescue services and defence departments, can call a single number and request participating satellites to acquire imagery over the affected area.

The International Charter helped with floods in Senegal on 2 September and those in Burkina Faso on 17 September this year. Both emergency requests received near-immediate data from RADARSAT and SPOT.

But the International Charter can only be activated after a disaster has struck, so does little to help developing countries acquire data for mitigation, planning and preparedness. 

It's not the only force for better access to satellite data though. The Global Earth Observation System of Systems (GEOSS), managed by the intergovernmental Group on Earth Observations (GEO), supports satellite access at all stages of the disaster management cycle. GEOSS promotes common technical standards so that data from the thousands of different instruments can be combined into coherent datasets.

GEOSS is also responsible for GEONETCast — a global network of communication satellites and alternative Web dissemination channels that get environmental data to disaster managers (and others). It provides data from various satellites including Meteosat, Geostationary Operational Environmental Satellite (GOES), Terra and SPOT to regional centres in Europe, Africa and Asia via a small receiving station. These centres then disseminate the data to local stakeholders using digital video broadcast.

Sentinel Asia and SERVIR are other major components of GEOSS. And GEO has done much to convince individual space agencies to release their data for free. Two years ago, it announced that the China-Brazil Earth Resources Satellite (CBERS) would distribute its images without charge.

Emerging from a GEO ministerial summit in Cape Town late last year, NASA announced that it would make the full archive, and future data, from the Landsat satellites free. The decision opened up remote sensing data to thousands of users across the world. In the first month after the announcement, Landsat distributed more than 200,000 scenes — about ten times more than the pre-announcement yearly average.

And in June 2009 Japan's Ministry of Economy, Trade and Industry and NASA contributed a 30m resolution global digital elevation model derived from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) to GEOSS, making access to it free. 

Of course, it's all very well to provide the data for free, via the Internet, but for regions such as Africa, that still suffer from low-speed networks and low bandwidth, access remains a problem.

Home-grown satellites

Some developing countries have invested heavily in both Earth observation and communication technology, and launched their own satellite — or constellation of satellites — to monitor and respond to natural hazards and disasters.

CBERS is a partnership between Brazil and China that began in 1988. Together, they have built two satellites carrying instruments to monitor Earth resources and have committed to building another two. CBERS is widely heralded as a major success model for South-South cooperation in space technology and, among other things, helps control fire in the Amazon region.

Algeria, China and Nigeria all own and control their own satellites, as part of the Disaster Monitoring Constellation (DMC) — a network of seven multispectral satellites (comparable to Landsat) built by UK-based Surrey Satellite Technology. They are equally spaced around the Earth to provide daily imaging capability. Their data produce maps and information to help disaster relief efforts. And DMC partners are signed up to the International Charter.

India also has an extensive space research programme, run by the Indian Space Research Organisation (ISRO). It includes a suite of remote sensing satellites — the Indian Remote Sensing satellite system (IRS) — the first of which was launched in 1988. India's National Database for Emergency Management uses IRS imagery to provide flood maps, relief support maps, embankment breach impact maps and flood frequency maps. IRS data can also track cyclones, predict their landfall, and give early warnings of tsunamis.

India's Disaster Management Support programme, also launched by the Department of Space, responds to all natural disasters in the country. Using IRS and other data, it delivers products such as hazard maps, early warnings, and vulnerability indices, but also focuses on building relationships between policymakers, international organisations and emergency operations agencies to cope with disasters.

Establishing such home grown solutions requires serious investment in both technology and capacity — investment that is still lacking in many developing regions. In Africa, for example, the remote sensing infrastructure in most countries (outside South Africa) suffers severe shortages of financial resources, technical expertise and political commitment.

Few countries on the continent have active space programmes, and many decision-makers simply do not recognise remote sensing as a useful tool for development.

Some international organisations are working to improve the situation. The UN Platform for Space-based Information for Disaster Management and Emergency Response (UN-SPIDER) runs regional workshops and provides technical advice to individual countries. Last year (2008) they sent a technical team to Burkina Faso to advise the government on how to include space-based technology in their national plans.

Regional organisations are playing their part too. Over the past decade, the African Association of Remote Sensing of the Environment has been promoting training. And more recent initiatives such as the University Network for Disaster Reduction in Africa (UNEDRA) have targeted universities as centres to improve remote sensing research and collaboration.

The time is ripe for engaging developing country researchers and policymakers in remote sensing for disaster management. Data and software costs are plummeting, information communication technology is developing quickly, and tools such as Google Earth are starting to get policymakers enthused about satellite imagery.

Sian Lewis is commissioning editor for SciDev.Net and holds a PhD from University College London in remote sensing.

References

[1] Rodriguez, J., Vos, F., Below, R. et al Annual disaster statistical review 2008: The numbers and trends Centre for Research on the Epidemiology of Disasters (2009)

[2] Ross, K. W., Brown, M. E., Verdin, J. P. et al Review of FEWS NET biophysical monitoring requirements Environmental Research Letters 4 (2009)

[3] Chien, S., Davies, A., Tran, D. et al Using automated planning for sensorweb response Jet Propulsion Laboratory, NASA (2004)

[4] Balz, T., Li, D. The Sichuan earthquake GIM International 22:10 (2008)

[5] Crustal deformation in China associated with the seismic cycle of major faults or related to lakes loading on the lithosphere: Measurement by SAR interferometry ESA

[6] Tropical cyclone operational plan for the Bay of Bengal and the Arabian Sea. Tropical Cyclone Programme Report No. TCP21 (2008)

[7] Rochon, G. L., Quansah, J. E., Mohamed, M. A. et al Applicability of Near-Real-Time Satellite Data Acquisition and Analysis & Distribution of Geoinformation in Support of African Development UN ECA (2005)