Skip to content

Geospatial analytics for crisis assessment


Recent advances in state-of-the-art sensing technology have produced an explosion of data generated by Earth observation missions. Using the criterion of spatial resolution (smallest size an object or detail can be represented in an image), satellite images can be categorized in large, intermediate, and high or extremely high-resolution. Out of these three types only the very high-resolution satellite imagery is the best suited to perform context and situational assessments due to its ability to capture objects as small as half a square meter. The price paid for the increased details are the inflated costs associated with producing, storing, and processing it. A second important aspect is that high resolution imagery is usually covering much smaller areas and thus it might be difficult to get a coverage for any given specific area at a particular time instance.

Spatial resolution of gridded datasets
Figure 1: Spatial resolution of gridded datasets


Nightime lights data

The main source of nightime lights imagery is the Payne Institute for Public Policy's Earth Observation Group(COLEOG) part of Colorado School of Mines, USA. The Suomi NPP VIIRS DNB imagery collect reflectance in 0.5-0.9 µm bandwidth and features a spatial resolution of 375m at the subsatellite point. The VIIRS DNB scans the Earth globe twice a day and the imagery is processed by the COLEOG into multiple products, respectively, nightly/daily, monthly and yearly mosaics. The mosaics cover the globe between latitudes of 75N and 65S using later on top logic in overlapping areas. At the same time COLEOG produces a cloud mask that can be used to filter out pixels that are covered by clouds.


UNDP GeoHub's team has developed a data processing pipeline to convert the COLEOG data to Cloud Optimized Geotiff format and host it as a static Spatio Temporal Asset Catalog in Azure. As a result the converted data is discoverable, standardized and can be easily consumed by various clients.

STAC catalog
Figure 2: STAC catalog


By comparing the nighttime lights at two consecutive time instances over a specific area, insights related to change in the intensity of human activities/presence can be derived under specific circumstances with high confidence. A downside is that again, this imagery might not be available at a particular time instance or specific geographic areas. The solution in this case is to use monthly, seasonal, or even yearly mosaics (composition of several satellite images) if available.

Nighttime lights image mosaic covering East Africa
Figure 3: Nighttime lights image mosaic covering East Africa


Rapid change assessment algorithm

GeoHub features the Rapid change assessment tool specially designed for the purpose of quick change assesment on the ground using nighttime lighs imagery. Put in a simplistic manner, the nighttime lights based change detection algorithm entails evaluating the sign and magnitude of the difference between the reflectances at two distinct calendar dates for any given location/pixel in the area covered by the imagery. An important aspect is to consider also the cloudiness of the imagery because clouds can affect the precision of computations in a negative way.

Derna use case

To illustrate the potential of nighttime lights data we are going to use a real life use case, concretely the Derna dam collapse that happened on 11 of September 2023. The dam collapses released an estimated 30 million cubic meters (39 million cubic yards) of water, causing flooding downstream as the Wadi Derna overflowed its banks.The floods partially destroyed the city of Derna. As of 18 September, estimates for the number of casualties range from 5,300 to 20,000 people according to Wikipedia.

Derna, Libya
Figure 4: Derna, Libya

--

Identifying changes in human activity using nighttime lights data. This example shows the input data and the result of the RCA algorithm aplied to the floods that accured in Derna, Libya in September 2023
Figure 5: Identifying changes in human activity using nighttime lights data. This example shows the input data and the result of the RCA algorithm aplied to the floods that accured in Derna, Libya in September 2023


RCA tool in Geohub

GeoHub has been enhanced with UI components that allow users to run the tool and conduct analytics seamlesly in the all area covered by nighttime lighs mosaics.

The tool can be accessed using folowing steps:

1. Load The GeoHub's tool page and click on RCA tile

Load The GeoHub's tool page and click on RCA tile
Figure 6: Load The GeoHub's tool page and click on RCA tile

2. Select Nightly nighttime lights mosaic dataset

Select Nightly nighttime lights mosaic dataset
Figure 7: Select Nightly nighttime lights mosaic dataset

3. Fill in the start (09092023) and end(13092023) dates and press Add to Map button to load the layer

Select dates and add data to map
Figure 8: Select dates and add data to map

4. Type Derna in the geolocator input, pres enter and select Derna Libya entry

The difference of RCA tool for Lybia floods
Figure 9: The difference of RCA tool for Lybia floods

Note

You can also access same RCA tool from datasets page here.

  1. Click Nighthly nighttime lights mosaics
  2. Click Preview tab
  3. Click Tools tab
  4. Click Rapid Change Assessment Tool

The Lybia floods dataset shows the overall level of destruction
Figure 10: The Lybia floods dataset shows the overall level of destruction

The original dataset of lybia floods is available from here

By overlaying this data with the RCA global layer before applying 50% transparency to it, it is possible to observe the RCA has succesfully detected the affected areas.

Zoomed area where the floods affected significantly
Figure 11: Zoomed area where the floods affected significantly


Demonstrating the ability of RCA tool to detect changes and variations in nighttime lights intensity as a proxy to identification of areas affected by disasters
Figure 12: Demonstrating the ability of RCA tool to detect changes and variations in nighttime lights intensity as a proxy to identification of areas affected by disasters

Success

This example showcases the efectivenes and viability of developing and applying geospatial analytics to support the decision making processes.