Saving forests using a fire detection system
An automatic camera network which can detect and report forest fires up to 25km away has helped save tens of thousands of hectares of woodland in South Africa from destruction.
ForestWatch is one of the most advanced automated forest fire detection systems in the world and has a success rate of up to 96%.
It utilizes advanced image processing techniques partly developed by Professor Mike Kosch from Lancaster University's Department of Physics.
Professor Kosch is now a non-executive Director of the South African company EnviroVision Solutions (EVS) which employs ForestWatch to detect, geo-locate and report forest wildfires.
He brings his expertise as a space scientist, with night sky imaging of the auroras and imaging processing algorithms and techniques from observational and analysis research experience in the Arctic and Antarctic.
He said, "Processing the images of auroras is the same as detecting smoke plumes in forest fires, these are also fuzzy objects that move, that are far away, so if you apply the same techniques to detecting smoke plumes in forests, you can detect a fire that has started much sooner."
"The technology has improved to the point where we can now detect a smoke fire as small as a barbecue at ranges up to 10 miles away in the day or the night under any weather conditions. The trick is not to have too many false alarms because the cameras are so sensitive."
Wildfires result in loss of life, property and crops, as well as billions in economic costs to prevent, mitigate and suppress the damage. Yet they are notoriously difficult to rapidly detect and accurately locate at the time of ignition, when counter measures are most effective.
Combining programmable high-definition cameras and image processing software, the computerized ForestWatch system uses a mathematical algorithm to analyze imagery of the landscape, automatically detecting changes that may be the first whiffs of smoke.
When the system identifies a smoke, it alerts the human operator at a control center.
EVS has deployed 340 camera systems so far around the world, mainly in South Africa and North America, with trials underway in Australia, Chile, China, Indonesia, Greece and Spain. These have substantially reduced commercial forest wildfire losses.
Satellite burn-scar data demonstrates that in South African regions where ForestWatch was deployed, approximately 120,000 hectares of forest have been saved between 2014-2021. This equates to a total commercial saving of approximately £1 billion in South Africa alone from 2014 to 2020.
This equates to a total commercial saving of approximately £1 billion and CO2 reduction of 27 million tons, valued at £1.2 billion, in South Africa alone in the period 2014 to 2020.
EVS U.S. has deployed 137 ForestWatch systems in North America, including Alberta, Saskatchewan, California and Oregon. In North America, it is estimated that ForestWatch systems have saved £1.6 billion in timber in timber and 38 million tons of CO2, valued at £1.9 billion in the period 2014 to 2020.
Provided by Lancaster University