Assessment of Post-Fire Ecosystem Resilience in Andean Moorland Using Remote Sensing and Statistical Modelling

• The January 2024 wildfire reduced REEA forest cover from 41% to 20% post-fire • Post-fire NBR is the primary predictor of restoration need (AUC = 0.97) • Logistic regression mapped restoration needs with 93.9% accuracy • NO₂ and anthropogenic pressure inhibit Andean moorland post-fire recovery • A spatial three-zone WLC framework prioritises post-fire restoration The Andean moorland plays a crucial role in ecosystem functions, including water regulation, carbon storage, and biodiversity conservation. These ecosystems, however, face growing threats from wildfires, notably the January 2024 wildfire in the El Ángel Ecological Reserve (REEA), which affected more than 1,400 hectares and compromised the ecosystem's resilience. This study aims to assess post-fire regeneration in the REEA in Ecuador through a multitemporal analysis of Sentinel satellite images, spectral indices, environmental variables, and multivariate statistical models, to support evidence-based planning of priority restoration activities. The methodology integrates remote sensing, land cover and land use assessment, and statistical modelling (Principal Component Analysis and Logistic Regression), incorporating topographic, climatic, and anthropogenic factors. Spectral indices reveal a decline in photosynthetic vigour during the fire and a heterogeneous post-fire recovery. The land cover analysis reveals a structural shift, characterised by a decrease in forest cover (from 41% to 20%), an expansion of grasslands, and an increase in agricultural areas. The predictive model achieves high performance (Accuracy = 93.9% and AUC ∼0.97), identifying key predictor variables, including the post-fire Normalised Burn Ratio (NBR) and Normalised Difference Vegetation Index (NDVI), alongside proximity to agricultural areas and human settlements. These findings demonstrate that post-fire vegetation recovery is strongly governed by fire severity and anthropogenic pressure, highlighting the need for restoration plans that incorporate passive regeneration in resilient areas and active reforestation in high-disturbance areas. This study presents a replicable methodological framework for ecological planning, decision-making, and adaptive management in fire-affected moorland ecosystems.