“…Our selection of an accumulated fuel dryness index for prediction of fire occurrence seems to agree with studies that have observed higher correlations of fire activity with longer time lag fire weather indices (e.g., [35,98,99]), reinforcing the hypothesis that antecedent water balance and accumulated drought can influence fire activity (e.g., [3,32,33,100]). Interestingly, the selected period of 50 days, out of the candidate 10-90 days evaluated, corresponds directly with the time lag of some frequently used indices that have been found to be related to fire activity, such as Drought Code (DC) (with a time lag of 52 days) ( [35,99,101]), 1000 h dead fuel moisture ( [17,18]), or 2-month SPI [36]. Furthermore, some of the shorter time lag widely used fire danger indices that have been frequently related to fire activity such as FWI (e.g., [31,37,102]) or ERC (e.g., [4,8]) integrate those longer time lag codes into their weighted calculation (e.g., [2,15]).…”