Understand why we did not recognize a fire event or its footprint

• Cloud cover can obscure fire detection: A dense cloud cover can hinder satellite-based fire detection systems. This is because a thick layer of clouds can make it difficult for the satellite to detect heat anomalies.

• Small fires may not be detected: Small fires are not always detected by satellites or fire detection algorithms. This is because they may not produce enough heat to be seen by the satellite. As of now, LEO Satellites integrated into Wildfire Solution can detect heat sources as small as 10*10 meters

• Detection delays: Satellite-based fire detection systems may experience delays as satellites need time to cover areas and downlink data at specific terrestrial or spatial stations. In our Wildfire Solution, fire detection takes under 3 minutes after data availability. However, GEO Satellites provide updates every 30 minutes, while LEO Satellites do so every 60-90 minutes. These variations affect the timeliness of fire detection info.

• False alarms: Satellite-based fire detection systems can also generate false alarms. For example, a hotspot on the ground that is not a fire but a constant heat source, such as an industry, a volcano,  or a reflective surface, can trigger an alert.

• The fire was extinguished before any satellite could detect it: Forest fires that happen to occur during the so-called afternoon gap (A time of the day when there are no LEO Satellites overpassing the Monitored Area), have almost no chance of being detected if people react fast and extinguish it before a LEO satellite can perform the detection for it. 

What about the burnt areas?

Burnt areas are calculated from the input of previous cluster detections associated with them.

For the calculation of burnt areas, we utilize the images provided by the high-resolution LEO Satellites integrated into the Wildfire Solution, Sentinel 2-A & B, and Landsat 8 & 9.  The calculation is based on the comparison between before and after images from the affected areas and to find differences between them after a cluster detection.

The logic for burnt areas calculations is as follows:

• Cluster Confidence: Min. value of 0.4

• New calculations/updates are available  every 2-3 days 

At OroraTech we’re constantly working to enhance the available technology's capabilities to overcome the current challenges for satellite-based fire detection. To find out more about our progress in this matter, please visit this link.