Análisis de la severidad de un incendio en un relicto de bosque seco tropical en Colombia

Abstract

Given that fire is part of the natural dynamics of certain ecosystems, and a way of studying the physical damage that occurs in the ecosystem is through severity, in this study the main objective was to determine the severity of a forest fire occurred in a dry forest of the Armero Guayabal municipality, north of the Tolima department in August 2016. Through the spectral indexes NDVI, NBR and dNBR derived from images of the Sentinel 2A sensor, the use of this sensor in fine scales analysis was validated. Through the NDVI index, 4 different coverages were discriminated, being forest the largest representation. After the incendiary event, the average reflectance of the NDVI decreased to 0.53 for which "forest" decreased by 52 ha and "without vegetation" decreased by 47.96 ha. The area affected by the fire was 96.15 ha, of which 36.69 ha were typified with moderate burning severity and 33.9 ha with moderate-high burning severity.