The devastating impact of war is a harsh reminder of the ongoing conflicts worldwide, with the destruction of civilian infrastructure being a tragic consequence. In a remarkable effort, researchers from Ludwig-Maximilians-Universität München (LMU) and the Technical University of Munich have devised an innovative method to automatically detect building destruction in war-torn areas.
The beauty of this method lies in its independence from commercial satellite imagery or training data. Instead, it utilizes synthetic aperture radar (SAR) data from the Sentinel-1 mission, which is readily available at 12-day intervals. By employing the InSAR interferometric technique, the team compares repeated images of the same region, calculating a coherence measure that indicates the similarity of backscattered radar signals. A sudden drop in coherence often signifies structural changes in buildings, such as damage or complete destruction.
To ensure accuracy, the team employs statistical assessment, estimating a 'normal' pattern of variation over time for each pixel and quantifying deviations using p-value probabilities. By integrating this data with building footprints from OpenStreetMap, the results can be aggregated at the building level, providing a measure of uncertainty.
Dr Daniel Racek, the first author of the study and a former doctoral researcher at LMU's Institute of Statistics, emphasizes the power of freely accessible data: "We can now track the evolution of destruction across space and time almost in real-time."
The method was successfully tested using case studies from the Beirut port explosion (2020), the destruction of Mariupol following the Russian invasion (2022), and the ongoing war in Gaza (2023 onwards). It accurately reconstructed both the spatial patterns and timing of building destruction.
The researchers believe this approach can serve as a swift and cost-effective tool for humanitarian situation assessments, academic research, and post-conflict reconstruction planning.
This research, funded by the Munich School for Data Science, has been published in PNAS Nexus, offering a glimmer of hope amidst the devastation of war.
