Quantile‐Based Fire Weather Index Better Informs Detection and Variability of Wildfire Risks in China

Abstract The Fire Weather Index (FWI) system, developed in Canada, has now been widely used to monitor and forecast wildfires worldwide. However, it shows a marked mismatch to wildfire occurrences and burned areas when applied to the domain of China, thus hindering the understanding on the response of fire risks to climate change in the populated area. Here, we designed quantile‐based FWI (QFWI) that proves to be more accurate in indicating wildfire occurrences across the country. Based on the correspondence of fire incidence and burned areas to QFWI, we proposed a classification scheme indicative of low‐, medium‐, and high‐fire dangers in China, and further examined the historical variability of wildfire dangers accordingly. The results show that China's wildfire risks did not increase monotonically with warming, but exhibited an inter‐decadal variability modulated synergistically by changes in multiple climate variables. From 1961 to 1990, the high‐risk days decreased at −0.98 ± 0.71 days yr −1 primarily driven by the slowdown of daily maximum wind speed (WS max ). After 1990, broad swathes of the country experienced significant increases (0.64 ± 0.54 days yr −1 ) in high‐risk days, particularly in southwest China, contributed mainly by warming of daily maximum air temperature ( T max ) and drying of daily minimum relative humidity (RH min ). The increasingly favorable weather conditions in combination with continuous fuel accumulation from massive afforestation efforts alarmingly signals hightened wildfire risks into future decades, even in humid southern China.