Assessing Burn Severity and Vegetation Impact Using Sentinel-2 Satellite Imagery and Geospatial Analysis in Mae Ka Subdistrict, Mueang District, Phayao, Thailand

Wildfires are natural disasters that severely impact ecosystems, economies, and societies. Rapid and accurate detection and assessment of wildfire-affected areas are crucial for effective management and restoration. This research utilizes Sentinel-2 satellite imagery to investigate wildfire damage by analyzing pre- and post-fire satellite data to 1) analyze vegetation index changes before and after burning; 2) generate burn severity maps of the affected areas in Mae Ka subdistrict, Mueang district, Phayao, Thailand; and 3) examine the relationship between vegetation indices and burn severity levels. The study employs various spectral indices, including Normalized Burn Ratio (NBR), delta Normalized Burn Ratio (dNBR), Normalized Difference Water Index (NDWI), Relative Burn Ratio (RBR), and Normalized Difference Vegetation Index (NDVI). Results indicate that the monthly RBR index demonstrates high efficiency in detecting burned areas, with an overall accuracy of 90.6% and a Kappa coefficient of 0.76. Furthermore, the delta Normalized Difference Vegetation Index (dNDVI) effectively assesses wildfire impacts on vegetation. Areas experiencing severe burn impacts exhibit low dNDVI values, indicating significant vegetation damage. This study highlights the efficacy of satellite imagery and various spectral indices in evaluating wildfire damage. The Relativized Burn Ratio (RBR) performs better in burned area detection, while the dNDVI provides valuable insights into vegetation damage assessment. In conclusion, this research highlights the significant potential of integrating Sentinel-2 imagery with spectral indices in forest management. The methodology presented here offers a powerful tool for enhancing our capacity to assess, monitor, and respond to wildfire impacts, ultimately contributing to an effective and sustainable forest ecosystem management practices.