Abstract
Forest fire is one of the main issues of forest ecosystems around the world which has resulted in loss of biodiversity, forest degradation, soil erosion, and greenhouse gas emission. Ironically, the information on the forest fire regime and its pattern are still lacking in the Himalayan region. In this study, Moderate-Resolution Imaging Spectroradiometer active fire data products from 2001 to 2020 have been analysed for understanding the forest fire trends and its hotspots patterns during the active fire season (February to June). About 1347 average fire counts/year were recorded in six natural vegetations with the highest number of fires observed during the year 2012 (n = 3096) and minimum in 2011 (n = 210). Mann-Kendall trends analysis for the spatial and temporal pattern of fires indicated that there is a significant increase of forest fires towards higher elevation. Forest fire hotspot analysis using fire radiative power, fire frequency, and fire density showed that Uttarakhand is the most forest fire-prone state as compared to other north-western Himalayan states. It is also revealed that the May month has a higher number of fire counts and the evergreen needle forests have higher fire frequencies amongst the forest types. The forest fires were found to be more influenced by land surface temperature as compared to rainfall. The outcomes in this study on the temporal and spatial patterns for forest fire can be used for forest fire modelling. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s11069-022-05530-5.