The Gap Left by Commercial Satellites

As the Iran conflict spreads across the Middle East, the flow of visual information from the ground has been severed. Iran's internet blackout limits the trickle of photos and videos reaching social media. More critically, commercial satellite imagery — long a backstop when ground reporting is impossible — is facing increasing restrictions. Even well-funded institutions with premium access are finding themselves without reliable visual data.

Into this gap, open-source intelligence outlet Bellingcat has released a free damage detection tool: the Iran Conflict Damage Proxy Map. It relies on Synthetic Aperture Radar (SAR) data from the Sentinel-1 satellite, operated as part of the European Space Agency's Copernicus mission.

How Radar Traces Destruction

Unlike optical cameras, SAR emits microwave pulses toward the earth's surface and interprets their echoes to capture structural information — regardless of cloud cover or time of day. The tool's core algorithm, the Pixel-Wise T-Test (PWTT), was developed by UCL lecturer and Bellingcat contributor Ollie Ballinger.

The algorithm establishes a baseline using one year of pre-conflict SAR data, defining a normal range within which 99% of backscatter values fall for each pixel. Post-conflict imagery is then compared against this baseline. When a building is damaged, its backscatter values consistently fall outside their historical norm — a signal the tool flags for further investigation.

In one documented case in southwest Tehran, a cluster of warehouses showed consistently anomalous backscatter from March 14 onward, which was later confirmed via optical satellite imagery.

From Gaza to Iran: The Tool's Evolution

The predecessor to this tool was launched shortly after the Gaza war began in 2023, when digital controls and access restrictions made it difficult to assess destruction. The same SAR methodology was validated across Syria, Iraq, and Ukraine before being adapted for the current Iran crisis.

The restriction of commercial satellite imagery is not new. As conflicts intensify, governments increasingly pressure private providers, and companies self-censor for geopolitical or commercial reasons. For journalists and open-source researchers, this trend represents a structural threat to information access.

Sentinel-1 data, by contrast, is fully public and free — insulated from the decisions of commercial providers.

Beating AI's Blind Spots with Statistics

The PWTT approach also sidesteps a key failure mode of AI-based damage detection: overfitting. A model trained on one region often fails when deployed in an unseen area. Because PWTT compares each pixel only to its own historical baseline, it avoids this generalization problem and reduces false positives across diverse geographies.

What Comes Next [AI Analysis]

The trend toward commercial imagery restriction is likely to accelerate. As conflict parties increasingly treat information control as a strategic asset, pressure on private satellite operators is likely to grow. In this environment, publicly operated open satellite programs like ESA's Sentinel series are likely to gain even greater strategic relevance.

Bellingcat's Iran map may represent a template: open-source methods compensating for the twin barriers of digital blackouts and imagery restrictions. Similar methodologies may increasingly become standard tools for international organizations, human rights groups, and investigative outlets facing comparable information vacuums.

The tool's limitations are also clear: SAR-based analysis produces probabilistic estimates, not confirmed damage assessments. All flagged locations require verification through optical imagery or on-the-ground reporting before conclusions can be drawn.