The massive 7.8 earthquake that hit Nepal on April 25 was the worst since 1934. More than 7,300 died, and 14,000 were injured. Earth observation satellite data have provided valuable information needed for rescue and recovery following such disasters, but mapping and using the data needed for ground relief in Nepal proved challenging.
Fifteen government space agencies from around the world and many private satellite constellations from the United States and Europe gave fast access to critical satellite imagery in the earthquake-stricken regions of Nepal. An international charter, “Space and Major Disasters,” was activated immediately after the earthquake. The charter’s goal is to acquire and provide satellite data free to those affected by major disasters in the world.
The U.S. Geological Survey provides access to these images through its website. Other agencies and private companies such as Copernicus, Digital Globe and Planet Labs distributed the data from their websites, too. To release the data as fast as possible, the quality was often compromised.
Interpreting satellite imagery acquired after an earthquake allows us to identify damaged infrastructure – buildings, roads, bridges – and landscape. By comparing these images with pre-quake images generated from archived satellite data, we can identify changes. These comparisons are useful for pinpointing damaged villages and towns, assessing passable routes and roads for rescue workers and identifying areas suitable for setting up aid camps where goods and services can be distributed.
Acquiring pre-earthquake data was easy; several agencies have collected and archived the data. Post-earthquake data are a bit more challenging, due to cloud cover during the time of satellite passes. Multiple satellites or satellite passes increase the possibility of cloud-free images. Even then, getting a cloud-free mosaic of the entire damaged area in Nepal immediately after the earthquake was not possible.
Radar satellites can penetrate clouds and acquire data, even on cloudy days, but their spatial resolution is not comparable with the very high resolution satellite data such as WorldView, GeoEye, Planet Labs, SkyBox and Astrium. Nonetheless, radar data are useful for identifying landslides and mapping surface deformations caused by earthquakes. For example, sentinel-1 data from the European Space Agency measured lifting of Kathmandu by 3 feet.
Several agencies provided satellite data in different platforms including file-transfer protocol, web download and cloud space. For example, several cloud-mapping platforms – including National Geospatial Agency’s public site, DigitalGlob’s Tomnod, Google’s crisis map and StreetMaps – allowed volunteers to help identify damaged buildings and roads using nothing but an Internet connection.
Thousands of volunteers and experts from around the world mobilized
to find cloud-free images and to map damaged infrastructure and landscape. Participation of thousands of volunteers was encouraging, but the amateur mappers made it difficult to produce consistent and high-quality products.
A centralized coordination was lacking. No organizations or groups consolidated maps produced by different groups, which meant some areas were mapped multiple times and other areas were left out.
Mapping is just one part of the story. Use of the data products for rescue and recovery is another. What mapping and information were needed on the ground and what was being mapped often did not match. Several organizations and groups published their data online, but rescue teams in many parts of the country did not have access to the Internet. Providing a printed map to rescue teams would have been more useful.
It was reported that Nepal had better images immediately after the earthquake, as well as a greater number of volunteers, compared with previous natural disasters. However, it was evident that rescue teams needed prior training to fully understand the importance and use of maps.
Satellite mapping during the first two weeks after the Nepal earthquake was not easy: It was marred by clouds, release of low-quality data, inexperienced mappers, a lack of centralized coordination and a limited use of data products for rescue and recovery efforts on the ground. These lessons should prove valuable for future rescue and recovery efforts.
Dr. Chandra Giri, a native of Nepal, is a research physical scientist of the U. S. Geological Survey and an adjunct professor at Duke University.