Investigating Simple Fire Weather Indices Using Reanalysis Data

A fire weather index is beneficial to the fire community if the index shows diagnostic and/or prognostic value. Whether complex or simple, an effective index must be able to “pick out” days or times in which large fire growth or erratic fire behavior is more likely to occur. An index that maximizes on days when anomalous fire behavior occurs, but remains relatively low before and after the anomalous day can be especially useful. While one should test indices in real-time fire situations using the best data available now, one can also look to prior events to determine if there is any predictive power for significant fires in history. Recent development of long-term reanalysis products provides the fire weather community with a consistent dataset that can be used to test the validity of a fire weather index over many events across a wide area.

In this study, we will use meteorological data from the NCEP Climate Forecast System Reanalysis (CFSR) to examine the effectiveness of hot-dry-windy-based fire weather indices for recent fires around the United States. The CFSR is output at 0.5° horizontal grid spacing with 6-h time intervals. While the CFSR's coarse resolution is insufficient to determine the impact of local terrain effects or meso-/microscale weather phenomena, the data can be used to find if there are larger-scale meteorological influences that are more likely to affect a given location. In this presentation, we will discuss the successes and failures of a simple hot-dry-windy index for a selection of major fires since 1979. Results show that certain formulations of a hot-dry-windy fire index show maximum values on historical extreme fire behavior days, which suggests that this simple index might provide useful information to the fire community.