Context
After a significant earthquake, responders and insurers face the same question at different timescales: where is the damage concentrated? Field assessment teams cannot be everywhere at once, and claims teams may receive thousands of notifications from across a wide area.
Challenge
Provide a spatially consistent, independent view of where the strongest surface change occurred — within days of the event — to support prioritisation of inspections, emergency response, and claims triage.
Data
SAR acquisitions from before and after the event (Sentinel-1 and, where tasked, higher-resolution commercial missions). Pre-event archive imagery enables both coseismic deformation mapping and coherence-based change detection.
Method
Two complementary analyses: interferometric processing to map coseismic ground deformation across the affected region, and coherence change detection comparing pre- and post-event image pairs to highlight areas where the radar signature of the urban surface changed abruptly — a signal that correlates with concentrated structural change.
Findings
Research applications of this approach have shown that coherence loss and deformation patterns can identify districts with concentrated surface change and help distinguish them from largely unaffected areas. This research experience informs how we would structure a post-event assessment for a client. (Based on the founder's research experience with earthquake-affected urban areas; specific published figures available on request.)
Business relevance
For insurers: an independent, event-wide evidence layer to support claims triage and resource allocation. For authorities and operators: a same-week prioritisation map when field information is still fragmentary.
Limitations
Coherence change indicates that the surface changed — it does not measure damage severity, distinguish structural failure from debris or repairs, or replace safety inspection. Deformation maps show ground movement, which is not identical to building damage. Field validation remains essential.