Wildfires that start in backcountry areas sometimes burn for hours before being detected and reported. Satellites offer a vantage point from which infrared sensors can detect fires. Individual satellites in low Earth orbit (LEO) offer infrequent overpasses, making the delay from ignition to detection unacceptably long. Geostationary satellites offer a platform from which to maintain a round-the-clock vigil, but lack geographic precision, and cannot detect a rather small fire within a large pixel definitively above noise.

The Global Fire Detection Constellation’s 4-μm camera point design incorporates a high-resolution focal plane array and onboard FPGA processing.
The solution is to place a number of small sensors aboard satellites distributed in a LEO constellation, such as Iridium-Next. The sensors are close enough to Earth to provide adequate spatial resolution for geographic specificity, and can be distributed in sufficient number to provide coverage of every point in the United States every 30 minutes or less. Sub-pixel detection is imperative to keep sensor size small.

A non-optimized system concept has been developed for deployment aboard Iridium-Next or a subsequent satellite constellation offering detection of new fires reaching a size threshold of 15×15 meters within 30 minutes, and downlinking the geographic location to within 500 m to the appropriate emergency service provider within the jurisdiction. The innovation is a distribution of small infrared framing cameras aboard a satellite constellation. They could process the images onboard as part of the instrument in near real time to reduce the false-positive rate, and then determine the geographic coordinates of new detections so that only the geographic coordinates and minimal intensity and size parameters need be downlinked and routed via a Web tool, and e-mailed to the appropriate emergency service.

Future value could accrue if the system is implemented by reducing the time from wildfire ignition to first application of firefighting resources by U.S. federal and state agencies responsible for fire suppression. Early firefighting can minimize the size of a fire, in some cases saving many millions of dollars in firefighting costs, and reducing property loss and loss of life, expense of evacuations, etc. While fighting a large fire, such a system can also dynamically map the full fire perimeter every half hour of all large fires in progress, enabling improved short-term forecasting for deployment of firefighting resources, and forecasting hazards to firefighting crews on the ground. Such mapping is presently done on and off for a few of the largest wildfires, but is limited by cost and availability of airborne platforms and instrumentation, especially when different large wildfires are burning simultaneously in different parts of the United States and abroad. Foreign policy benefits could accrue by providing early detection and monitoring services for other countries with wildfire problems.

This work was done by Robert L. Staehle, David R. Thompson, Edward Blazejewski, William R. Johnson, Michael B. Mercury, and Paula J. Pingree of Caltech for NASA’s Jet Propulsion Laboratory. NPO-48478