Distributed Sensors for Wildfire Early Warnings

Abstract

Wildfires cause immense damage and loss of life. They are exacerbated by climate change effects and will be getting worse each year for some time to come. One way to deal with them is to have better means of early detection. Fires produce significant changes in the physical and chemical makeup of the local atmosphere that, in practice, can be detected by air quality monitors. This paper describes a sensor array that measures trace levels of fire produced emissions that is coupled to computation and communication equipment that is low power and low cost. The emission arrays are calibrated and deployed in controlled fire detection situations. Evaluations illustrate some important characteristics of detecting fire emissions including: 1] sensing CO and Particulates together reduce ambiguity of signals; and 2] fire emission signatures for relatively close fires produce rapid spikes in concentrations of emissions. Most importantly, the details of this work indicate that an individual sensor node consisting of only a CO and particle detector can provide an early indication of a wildfire. Additionally, the low-cost CO and particle sensors used in this study show a correlation of greater than 0.9 R2 with FRM reference monitors. The results are encouraging that very low-cost arrays could substantially contribute to an early warning system for detection of wildland fires thereby improving response times for mitigation measures.