An optical satellite-based analysis of phenology and post-fire vegetation recovery in UK upland moorlands

Abstract

Vegetation fuel dynamics in the UK's upland moorlands are important in determining landscape susceptibility to wildfire. Changes in fuel availability are influenced by phenology, land management activities or disturbances such as wildfires. Monitoring such changes is therefore essential to assess wildfire risks and impacts. This study used vegetation indices (VIs) derived from Sentinel-2 time series (2017–2023) and harmonic modelling to capture the phenology of key fuel properties and monitor post-fire vegetation recovery in four upland land cover types (acid grassland, heather, heather grassland, bog). We identified periods of high flammability, assessed the impact of wildfires on the spectral signal, and determined the time for spectral recovery as well as potential drivers of recovery times. Results showed the period of highest flammability from mid-February to early May in acid grassland, extending to early June in heather and heather grassland, and late June in bog. Summer fires caused more pronounced changes in fuel properties than spring fires, particularly in moisture-related VIs. Graminoid-dominated areas recovered rapidly (under a year), matching field observations, while dwarf shrub-dominated areas required up to three years, consistent with measurements of vegetation cover on burned areas but not with height. Spectral recovery times were primarily explained by land cover class, burn severity, season, and winter snow cover (R2 = 0.66). Field data highlighted pre-fire stand age's role in heather recovery and grasses' impact on spectral signals. This study improves understanding of fuel dynamics in upland moorlands through satellite monitoring, providing critical insights for more effective wildfire risk assessments and management strategies.