THE USE OF VISIBLE AND INFRARED DAY-TIME AND NIGHT-TIME IMAGES FOR FOREST FIRES' MONITORING

Timely detection of forest fires and reliable assessment of their size, dynamics, and consequences is an important and crucial point. Solving this problem would allow organizing rescue services during a fire more effectively and quickly. Operability and reliability of information make the activities on mitigation their consequences more task-oriented and effective and implement prevention measures more surely. Here, we present the results of processing and analysis of multi-spectral satellite images of low spatial resolution to assess the dynamics and consequences of the forest fires that happened in October-December 2017 on the territory of California. Based on the ERS data of visible and IR range from terra, Aqua, Suomi NPP and Sentinel-3 satellites, the fire sources were revealed, the forest tire dynamics was investigated, the boundaries of burnt-out localities were determined, and the area of affected territories was calculated. A qualitative assessment of reliability of detecting the fire sources (based on smoke plumes and heat anomalies) and a quantitative assessment of the accuracy of determination of fire sources' locations were made by comparing the results of processing the ERS data from MODIS, VIIRS and OLCI imagers with the results of fire detecting on the same days in the same territory obtained using specialized web-services of global satellite fire detection and monitoring. A quantitative assessment of the accuracy of determination of boundaries and areas of burnt-out localities was made by comparing the results of processing the ERS data from low-resolution imagers (MODIS, VIIRS and OLCI) with the results of processing the charge-free medium-resolution images from Sentinel-2A/B and Landsat-7/8 satellites made in the same period in the selected monitored territory and available online. The feasibility and expediency of joint processing, analysis, and interpretation of multi-spectral day-time imaging data and panchromatic night-time imaging data are shown with the aim of increasing the reliability and timeliness of detecting fire sources and assessing the dynamics of fire propagation. To increase the accuracy of determination of areas and boundaries of burnt-out localities, it is necessary to use the medium-resolution ERS data from Landsat-7/8 and Sentinel-2A / B satellites and from "terra satellite (ASTER imager).