Integrating data from different sensors for damage assessment after a natural disaster in rural areas

Rapid damage mapping following a disaster event is critical to ensure that the emergency response in the affected area is prompt and efficient. Amongst major disasters, earthquakes are characterized as unpredictable and of high frequency of occurrence. Previous and current studies focus mainly on the mapping of damaged structures in urban areas after an event such as an earthquake disaster. Yet, research focusing on the damage level or its distribution in rural areas is absent. According to the UN, nearly half of the world's population lives in rural areas and is expected to rise. Furthermore, their resources and capabilities for disaster relief operations are limited. Therefore, there is a great importance to assess the damage following a disaster in these areas.

The primary aim of this study is to characterize and assess the damage (level and extent), temporally and spatially, following an earthquake event, in rural settlements. This will allow producing an algorithm suitable for rural area rapid mapping, which will contribute to our understanding and will provide insights of the damage extent and will allow a better response and access to the affected regions and remote population.

For this purpose, a damage assessment algorithm that will map the damage in both urban and rural environments is proposed. This algorithm makes use of combining SAR and optical data for rapid damage mapping.

As a case study we will demonstrate this algorithm using the areas affected by the Sulawesi earthquake and subsequent tsunami event in Indonesia that occurred on 28 September 2018. High-resolution COSMO-SkyMed images pre and post the event, alongside a Sentinel-2 image pre- event are used as inputs.

The affected areas were analyzed with the SAR data using interferometric SAR (InSAR) coherence map. To overcome the loss of coherence caused by changes in vegetation cover, a vegetation mask was applied by using the NDVI to identify (and remove) vegetated areas from the coherence map. Then, thresholds were determined for the co-event coherence map (≤ 0.5) and the NDVI (≥ 0.4) and the two layers were combined into one. Based on the combined map, a damage assessment map was generated by using GIS spatial statistic tools (Fishnet and Zonal statistics). This map provides a quantitative assessment of the nature and distribution of the damage in rural and urban environments, as well the differences of damage features between them. The preliminary results show that while in urban area many structures were damaged, still in the rural areas the damage is larger, since most of the structures were damaged or even destroyed.