Ekta Aggarwal^a, Sanjeev Gupta^a, Alexander C. Whittaker^a, Philippa Mason^a, Kartikeya S. Sangwan^a, Fritz Schlunegger^b

^a Imperial College London, ^b University of Bern

Major flood events cause adverse impacts on human populations, infrastructure, and resources. The occurrence of flood events has led to much research on flood vulnerability, exposure, and post-disaster assessment. However, there has been a limited focus on studying the recovery of societies and how this varies among different socio-economic communities. Satellite datasets are an effective and cost-efficient way to analyze the spatial extent of flooding and its impact on anthropogenic activity in floodplains. Here we present an analysis of NASA Black Marble nighttime lights (NTL) data for the 2022 Indus River flood in Pakistan as it provides information on lit infrastructure which is a useful proxy for real-time monitoring of human presence. We investigate the impact of the flooding on the variance in NTL radiance during and after the event, to investigate human exposure and response to floods.

The Black Marble daily NTL VNP46A2 data product is the daily moonlight-adjusted NTL. Here, we utilize this dataset to assess the spatial impact of the 2022 Indus River flooding, which was one of the biggest flood events in recent history affecting 33 million people. We focus on the downstream floodplains of the Indus basin because frequent flooding over this area is historically documented. To determine the spatial extent of flooding we use ESA Sentinel SAR data to map the spatial and temporal evolution of flood extent between June-September 2022. In tandem, we analyze NTL data to explore the variation in NTL radiance values over a similar timescale. In particular, we use the NTL to estimate the variation in its radiance for areas of high flood exposure. We find that the NTL radiance in the flooded areas was affected for approximately 13 weeks. It took approximately 10 weeks post-flooding for the area’s radiance to recover up to pre-flooding level. The mean radiance from pre-flood to during the main flood period shows a decline by a factor of 5. Additionally, there is varying exposure and recovery for different socio-economic communities in the area. Our findings have the potential to improve our understanding of human response to floods and vulnerability at the lowest administrative level in a fast, safe, and cost-effective manner. This may also provide a framework for policymakers to assess flood vulnerability and impact at a basin scale.