Comparing Fire Weather Index (FWI) and Vapor Pressure Deficit (VPD) as Predictors to Forecast Subseasonal Wildfire Emissions

Abstract

Wildfires are becoming increasingly common, and although several studies have examined how Canadian Forest Fire Weather Index (FWI) can be used to predict Fire Radiative Power (FRP), the radiative energy a fire emits, most of them only focus on Canadian wildfires. This study uses FWI and Vapor Pressure Deficit (VPD) to predict FRP in both the USA in 2020 as well as the 2020 Washington Labor Day Fires. VPD is the difference between the amount of moisture the air can hold at its saturation point and the amount of moisture currently in the air and is now considered more accurate than Relative Humidity (RH). We create cumulative distribution functions for FWI and VPD to predict FRP up to 7 days in the future. The accuracies of these FRP predictions are compared to each other as well as static methods, like persistent FRP and constantly decreasing FRP to further measure their precision. Results reveal that VPD predictions of FRP are more accurate than predictions with FWI and static FRP prediction methods. VPD predictions of 7 day FRP were 19.23% more accurate than constantly decreasing FRP and 76.03% more accurate than persistent FRP. More accurate predictions with VPD can help mitigate the risk of destructive wildfires, emphasizing the need for accurate and reliable wildfire prediction infrastructure to protect communities that are at risk.