Digital Mapping Burn Severity in Agricultural and Forestry Land over a Half-Decade Using Sentinel Satellite Images on the Google Earth Engine Platform: A Case Study in Isparta Province

Abstract

This study utilizes remote sensing, Google Earth Engine (GEE), and Sentinel-2 imagery to map burn severity in Isparta province, Türkiye, in August from 2017 to 2022. Employing true color and normalized burn ratio (NBR) maps, the research assesses land conditions and vegetation dynamics. The delta normalized burn ratio (dNBR) reveals changing burn severity, particularly in agricultural and forested areas. The study granularly analyzes burn severity in agricultural areas and forested/semi-natural regions. Results show a significant decline in the 'Unburned' class, indicating altered land use and increased fire impact. 'High post-fire regrowth' expansion signifies vegetation recovery, while varied patterns in the 'Low post-fire regrowth' class suggest influences from environmental variables, fire management, and climate. Total burn severity escalates, with a 133.2% rise in agricultural zones and an 888.0% surge in forested/semi-natural regions from 2018 to 2022, highlighting the growing wildfire threat. In conclusion, this research provides crucial insights into Isparta's environmental dynamics, emphasizing the need for meticulous analysis in shaping effective fire management strategies and conducting comprehensive environmental assessments.

Keywords

• Remote sensing
• Burn severity mapping
• Google Earth Engine
• Sentinel-2 satellite imagery
• Ecological impact
• Wildfire risk