Combining precipitation forecasts and vegetation health to predict fire risk at subseasonal timescale in the Amazon

Abstract Current forecast systems provide reliable deterministic forecasts at the scale of weather (1–7 days) and probabilistic outcomes seasons (1–9 months). Only in recent years research has begun transitioning to operational settings numerical predictions for a lead time 2–4 weeks, timescale known as subseasonal. The Subseasonal Experiment (SubX) multi-model ensemble mean precipitation (2017–2021) days 8–14 (week-2 forecast) is used covariate logistic regression models predict fire risk Amazon. In complementary experiment, vegetation health index (VHI) added SubX predictor fires. We find that can be skillfully assessed most Amazon where fires occur regularly. some sectors, week-2 alone risk, but addition VHI results both (a) larger portion domain with skillful and; (b) higher skill sectors. By comparing two sectors Amazon, we information provided by relevant mosaic land covers includes savannas grassland, whereas sole areas cover dominated forests. Our illustrate potential using model forecasts, subseasonal timescale, combination satellite remote sensing obtain operationalization methods presented this study could allow better preparedness reduction greater than week.