Automated Fire Assessment — the next level of forest fire monitoring.
Forest fires, be it anthropogenic (human activity) or natural, degrades the ecosystem, human and animal life and infrastructure. 2019 has seen an exponential rise in fire events across the world. California fires, the Amazon forest fires, Bandipur forest in India, Australian fires (which are still active) are just a few of these. Forest fires destroy large areas of land in a relatively short time and are often difficult to control. The recent Australian fire looked something like this from space.
And to the OLI infrared the fire looked like this
It clearly shows the widespread effect that such can have over the planet and disrupt normal life over large geographies.
Location intelligence and satellite imagery has recently started playing an active role in observing, monitoring and planning control strategies for such events. The bottleneck still remains for rapid analysis procedures. The Forest Survey of India has developed a system called the FAST 3.0, a revolutionary approach for distributing data of active fires to regional forest departments. The system updates twice a day and can deliver near real-time data. This data can be further added value by extracting burn area vectors and understanding the extent of the burn area. This vector extraction can be done using satellite imagery, but needs to be done using algorithms operating on cloud computing systems and running without any human intervention. Such an automated process can help in creating near real time insights and even compare other areas, or even same geography and a different time.
A multi-temporal analysis of satellite images gives a precise understanding of the fire on both a near real-time basis and in monitoring the situation post-event as well. For example, in February 2019 a large scale fire broke out in the Bandipur National Park, India.
Vasundharaa has been actively working in this space creating block based processes to extract the burn area vectors, within minutes of data delivery. Our algorithms are hosted on our partner UP42 marketplace. In the current scenario, the UP42 marketplace offers the varied satellite datasets and hosts algorithms developed by Vasundharaa for rapid analytics.
Time is of critical importance when dealing with large-scale fires to prevent loss of life and reduce health and socio-economic impacts. In this regard, real time analytics performed over diverse datasets is the foundation to monitor and effectively manage the event. This cost-effective approach provides consistent, high-quality updates on the progression of the events while supporting complete disaster management cycle of readiness, mitigation, response and restoration.
If deployed effectively real time analytics and diverse data connection can be very effective on ground not just to support the disaster management cycle, but also help in mitigating effects on businesses running the area. Route planning, assessing damage to the supply chain, affects that fire events might have on prospective market are the kind of questions industries can plan for in advance. Improving on the current observation and analytics capabilities, we aim at using Synthetic Aperture Radar sensors and automated analysis for giving out day and night, all weather observation and data distribution. With this we believe, true potential of earth observation can be explored and timely, actionable intelligence can be delivered to the right end users.
