Hourly fire-weather model exposes local danger missed by daily maps
A peer-reviewed modelling study used high-resolution Meso-NH simulations to revisit three past fire episodes across the Portuguese–Spanish border. Calculating the Fire Weather Index hourly exposed a daily cycle and local gradients linked to sea breezes and terrain flows that coarse daily products missed. The work tests historical cases; it does not establish that an operational forecast will improve or predict where a fire will ignite.
Science··Evening
Hourly calculations reveal a daily cycle
A peer-reviewed Scientific Reports paper published on September 27 revisited three past fire episodes across the Portuguese–Spanish border region. Researchers used the Meso-NH atmospheric model on two nested grids, with cells 1.5 km and 0.5 km wide, to calculate the Fire Weather Index, or FWI, each hour. The conventional index combines temperature, humidity, wind and precipitation to rate meteorological conditions linked to fire behavior, commonly from daily inputs near local noon. In the simulations, the hourly index and its fine-fuel moisture component rose and fell through the day. A single daily value can miss that change, particularly when near-surface weather shifts quickly.[1]
Sea breezes and terrain sharpen local contrasts
The high-resolution runs represented sea breezes and flows shaped by hills and valleys. These local processes changed temperature, humidity and wind near the ground, producing marked differences in calculated fire-weather danger over relatively short distances. Coarser daily meteorological products did not resolve the same gradients in the cases examined. The model therefore supplied a more detailed picture of when and where atmospheric conditions became conducive to fire spread. The index describes weather-related danger, not a record of every ignition, and the paper did not claim that its maps located the cause of a particular fire.[1]
Operational forecasts still need testing
The work compares simulations for three historical fire periods, rather than testing a forecast issued before new fires. The authors point to a possible route into Portuguese fire-danger systems because the operational AROME weather model shares a physical package with Meso-NH. That connection makes implementation plausible, but it is not a measured gain in warning accuracy or lead time. Forecasters would still need to test the hourly approach on unseen periods and judge whether the extra spatial detail changes decisions. For now, the demonstrated finding is that these simulations exposed local and within-day weather contrasts that the coarse daily products missed.[1]
