Remote Detection of Overwintering Fires

This report presents research conducted by Dr. Christopher Manuel Selman over July 18th, 2022, to July 18th, 2023, for his Karle's Fellowship in Detecting Overwintering Arctic Fires. Overwintering fires are fires that span two years; in the first year, there is an incipient fire. This fire burns deep into the peat of boreal forests and smolders over the winter, into the next thaw. After the ground has thawed, the fire re-emerges. Currently, remote detection or even simple identification of overwinters is difficult. For detection, persistent cloud cover, snow cover and frozen soil present insurmountable odds for any one wavelength to overcome. A multispectral approach, utilizing low earth orbiting visible, infrared and microwave instruments may mitigate these weaknesses. However, before any analysis can begin, case studies must first be identified. In this study, I discuss the development of a parsing tool for Alaskan Wildland Fire Information data using a series of Python routines. One candidate fire of suitable size identified from this utility and another fire from the literature are chosen for analysis. Data from the Visible Infrared Imaging Radiometer Suite (VIIRS) and Soil Moisture Active Passive (SMAP) instruments are then ingested into newly developed capabilities of the Navy's Geolocated Information Processing Suite (GeoIPS.) These capabilities include a normalized burn ratio and VIIRS Day Land Cloud Fire RGB, as well as microwave brightness temperature plots from the SMAP passive microwave radiometer and synthetic aperture radar.