Predicting Spatially Explicit Composite Burn Index (CBI) from Different Spectral Indices Derived from Sentinel 2A: A Case of Study in Tunisia

Fire severity, which quantifies the degree of organic matter consumption, is an important component fire regime. High-severity fires have major ecological implications, affecting carbon uptake, storage and emissions, soil nutrients, plant regeneration, among other ecosystem services. Accordingly, spatially explicit maps severity are required to develop improved tools manage restore most damaged areas. The aim this study field-based (composite burn index—CBI) from different spectral indices derived Sentinel 2A images using several regression models. areas two recent large that occurred in Tunisia summer 2021. We employed normalized ratio (NBR): differenced NBR (dNBR), relative (RdNBR), relativized Ratio (RBR). In addition, we calculated burned area index for 2 (BAIS2) thermal anomaly (TAI). Different tree decision models (i.e., recursive partitioning method [RPART], bagging trees [Bagging], boosted [BRT]), as well a generalized additive model [GAM]), were applied predict CBI. main results indicated RBR, followed by dNBR, predicting Moreover, BRT was best model, explaining 92% CBI variance training set points 88% when validation set. These suggested adequacy RBR assessing mapping forest Mediterranean forests. could help improve post-fire recovery restoration efforts.