
On 6 February 2023 two major earthquakes (M7.8 and M7.5) ruptured the East Anatolian Fault around Kahramanmaraş, collapsing tens of thousands of buildings across a wide region of southern Türkiye. In the first 72 hours, responders need damage maps regardless of weather or daylight — and no single satellite revisits fast enough to provide them alone.
The answer is to task multiple constellations at once and compress time-to-first-usable-image. PassPrediction plans that fan-out neutrally: draw the affected region, and it surfaces every optical and SAR pass across all operators in your window, ranks them by how quickly a usable scene can be captured and delivered, and compares constellations so you can order the fastest feasible mix — VHR optical for building-level damage plus SAR for the night and cloud gaps.
Draw the AOI as the affected region — the swathe of towns and cities along the rupture — not a single epicentre, because damage is distributed and you want every pass whose swath touches any part of it. Set a tight window around the event: ideally a clean pre-event baseline plus the first feasible post-event passes, so change can be measured against a known-good state.
The metric that matters is time-to-first-usable-pass: the earliest overpass that is in geometry, not blocked by cloud, and within an acceptable look angle. Widening the satellite set to every operator and relaxing the off-nadir tolerance is what shrinks that from days to hours when lives depend on it.
No single constellation revisits a fixed region fast enough in the first day, so the winning move is breadth. VHR optical — WorldView Legion — gives building-level damage grading in daylight and clear skies, the detail responders need to prioritise collapse sites. SAR — ICEYE — fills the gaps by imaging through night and winter cloud, delivering flood, structural, and access information when optical is grounded.
Tasking both and taking whichever pass comes first is precisely the pattern that compresses time-to-first-image. PassPrediction is built to compare them: instead of committing to one operator and hoping its next pass is clear, you plan across the whole market and pick the fastest feasible acquisition, sensor-agnostic.
Run a pass search over the region for the response window and sort by coverage and earliest time so the fastest, most complete passes surface first. Filter SAR without the daylight constraint and optical with it, so each sensor is evaluated on its real strengths, and use the tasking view to confirm the swaths actually blanket the affected towns.
Add the latency estimate to see when each pass's data could reach the response cell — capture is only half the clock — and run the constellation comparison to rank the optical and SAR options together. The result is a ranked, neutral tasking plan you hand to the operators who can deliver first.
Multi-constellation — WorldView Legion + ICEYE (optical + SAR) — over the Kahramanmaraş, Türkiye Area of Interest.
14 feasible passes over the AOI in 3-day.
| Best pass | Start (UTC) | Coverage | Off-nadir |
|---|---|---|---|
| WorldView Legion 4 | 2026-07-15T03:28:05.754016+00:00 | 10.4% | 3.740849782301413° |
| Stage | Duration |
|---|---|
| Order ingest | 10 s |
| Uplink wait | 51 s |
| Execution | 1 h 14 m |
| Downlink wait | 26 m |
| Processing | 10 m |
| Delivery | 1 m |
| Total | 1 h 53 m |
Downlinked through AWS Ground Station Dubbo. The AWS Ground Station and KSAT antennas nearest the AOI keep downlink latency low so the first post-event passes are on the ground while they still drive rescue decisions.
| # | Constellation | Score |
|---|---|---|
| 1 | WorldView Legion | 0.63 |
| 2 | ICEYE | 0.57 |
| 3 | WorldView | 0.43 |
It depends on how many satellites you task and your look-angle tolerance. Fanning out across all optical and SAR operators with a generous off-nadir often yields a usable pass within hours; restricting to one sensor can push it to days.
No single constellation revisits fast enough alone. Optical gives building-level damage detail; SAR sees through night and cloud. Tasking both and taking the first usable pass minimises time-to-first-image.
For change-based damage mapping, yes — plan one clean pre-event scene so post-event imagery can be differenced against a known-good state.
No. PassPrediction does not sell imagery — it plans feasibility across all operators, then you order from the provider of your choice.
Draw your Area of Interest, set the window and look-angle limits, and PassPrediction ranks every feasible pass across all operators — neutrally, in your browser, free to start.
Open the planner →