"This study evaluates the skill of a newly developed fire-weather index called the Hot-Dry-Windy Index (HDW) using five meteorological analysis datasets for twenty-three historical fire events, and sensitivities of that skill to assumptions made in index development. The meteorological analysis datasets used in this study are the Climate Forecast System Reanalysis, the Global Forecast System, the North American Regional Reanalysis, the Rapid Update Cycle, and the North American Mesoscale Forecast System. These datasets were chosen because they are widely used in weather forecast and other meteorological applications and because their data archives cover the period of historical fire events used in this study. The twenty-three historical wildland fire events were chosen to provide geographic diversity to this evaluation. The results of this study suggest that the original HDW formulation is capable of identifying the largest fire spread day for between 56.5 to 78.3% of the twenty-three wildland fire events used in this study, and that alterations to the HDW formulation do not positively impact the skill of this index. The results also indicate that inclusion of the grid points surrounding the central-grid-point containing the latitude and longitude of each fire event in the evaluation of index skill provides higher skill than when only considering the central grid point. The skill of HDW computed by the five meteorological analysis datasets vary between fires and no one dataset consistently outperforms or underperforms the others."--Page ii.
Read
