Inferring the past: a combined CNN-LSTM deep learning framework to fuse satellites for historical inundation mapping

TL;DR

This paper introduces a combined CNN-LSTM deep learning framework that fuses satellite data from Sentinel-1 and MODIS to accurately infer and map historical flood events over Bangladesh, enhancing flood risk assessment.

Contribution

The novel framework effectively integrates spatial and temporal satellite data, outperforming existing methods in historical flood mapping accuracy.

Findings

Outperforms CNN-only models in flood extent prediction

Accurately infers 20 years of inundation over Bangladesh

Provides more consistent and reliable flood extent estimates

Abstract

Mapping floods using satellite data is crucial for managing and mitigating flood risks. Satellite imagery enables rapid and accurate analysis of large areas, providing critical information for emergency response and disaster management. Historical flood data derived from satellite imagery can inform long-term planning, risk management strategies, and insurance-related decisions. The Sentinel-1 satellite is effective for flood detection, but for longer time series, other satellites such as MODIS can be used in combination with deep learning models to accurately identify and map past flood events. We here develop a combined CNN--LSTM deep learning framework to fuse Sentinel-1 derived fractional flooded area with MODIS data in order to infer historical floods over Bangladesh. The results show how our framework outperforms a CNN-only approach and takes advantage of not only space, but also time in order to predict the fractional inundated area. The model is applied to historical MODIS data to infer the past 20 years of inundation extents over Bangladesh and compared to a thresholding algorithm and a physical model. Our fusion model outperforms both models in consistency and capacity to predict peak inundation extents.