Wireless Sensor Network Based Fire Detection and Growth Model for Management of Wildfire

Abstract

Forest fires are inevitable, and at the same time, it is not possible to fully recover from the damage caused by these disasters. In the forest ecosystem, though the fires are a natural occurrence, yet the fire seasons are nowadays turn out to be more severe and extensive. Approximately 67 Mha of forest area have been burnt annually between 2003 and 2012. In 2015 alone, the total forest area affected by fires was 98 million hectares. Due to its widespread impact, it is essential to control or at least minimise the effect of forest fires. However, there are various factors that affect forest fire. Few of these factors cannot be ascertained in real-time. The possible danger could be reduced to a certain degree by deploying early fire detection systems as the forest fires detected in the nascent stage are easy to control and manage. But all the fires cannot be controlled easily and therefore understanding of the growth pattern for a fire can help the authorities in planning fire suppression strategies during an outbreak. These fire growth models take into consideration the environmental factors that directly affect the behaviour of the fire in the forest and give a spatiotemporal estimate of the fire location and approach. newline