Retrieval of Andean grassland vegetation water content from satellite-based datasets as a tool for wildfire prevention

Zubieta R.; Sinticala J.; Ccanchi Y.; Laqui W.; Saavedra M.; Sulca J. · Remote Sensing Applications: Society and Environment (2026) · DOI: 10.1016/j.rsase.2026.102239

Resumen

Wildfire events in Peru have increased significantly over the past two decades, primarily affecting grasslands in the Andes, where the use of fire in agricultural and livestock activities is the main ignition source. Nonetheless, vegetation parameters that estimate the impact of dry periods on grasslands—factors that may indirectly contribute to the occurrence of severe wildfires—are not currently monitored. This study investigates conditions conducive to wildfires in the southern Peruvian Andes by analyzing the capability of satellite datasets to represent vegetation water content (VWC) for the 2024–2025 period. Ground-based and satellite-based observations (MODIS and VIIRS datasets), together with linear regression models, were used to represent VWC. The results indicate that reductions in grassland moisture are strongly linked to dry season dynamics and wet season onset. Our findings suggest that spectral indices such as the GVMI and NDII, derived from MODIS and VIIRS data, can effectively monitor the spatial and temporal distribution of VWC at the regional scale, achieving performance levels of approximately 84% and 71%, respectively. This accuracy is largely attributable to the important contribution of data from the SWIR spectral band, which adequately captures intra-seasonal variations in VWC. Ranges of VWC thresholds for fuel ignition during the wildfire season are proposed in this paper, these may be useful for regional wildfire early warning systems aimed at detecting severe increases in fire activity.

Instituciones: Universidad Nacional Agraria La Molina; Universidad Nacional de Ingeniería; Universidad Nacional del Altiplano