From daily to hourly scale: improving fire weather index with high-resolution Meso-NH simulations

Abstract

Wildfire danger assessment relies heavily on meteorological data and is challenged by the mismatch between real-time needs and the coarse spatial resolution of weather prediction outputs. This limitation can hinder the representation of local atmospheric processes that strongly influence wildfire behavior. This study evaluates the benefits of increasing the spatiotemporal resolution of inputs for Fire Weather Index (FWI) calculations to improve wildfire risk characterization. Three periods with active fires in the Portugal-Spain regions were analyzed using the atmospheric model Meso-NH, configured with nested domains at 1.5 km and 0.5 km resolution. These outputs were used to compute an hourly version of the FWI, allowing the analysis of sub-daily danger variability. Results reveal a well-defined diurnal cycle in the FWI and its Fine Fuel Moisture Code (FFMC) component, showing high sensitivity to near-surface conditions. High-resolution simulations successfully captured mesoscale processes, such as sea breezes and orographic flows. These processes generated marked spatial gradients in FWI values that are not detectable in daily products derived from coarse-resolution meteorological datasets. Overall, this study highlights the potential for integrating high-resolution numerical weather prediction into the Portuguese operational fire danger assessment systems given that AROME model shares the same physical package as Meso-NH.