FireWx Development and Application of an Hourly Wildfire Potential Index for Predicting Wildfire Activity

Extreme fire weather drives costly and dangerous fire disasters such as the Marshall Fire in Colorado, Lahaina Fire in Hawai’i, and Jasper Wildfire Complex in Alberta. Meteorological data now refreshes quickly enough to develop rapidly-updating fire weather indices. The NOAA Global Systems Laboratory has developed an index of hourly wildfire potential (HWP), which takes into account wind speeds, humidity, and soil moisture from an hourly cycled 3-km numerical weather prediction system to quantify fire danger on an hourly time scale.

We further develop the HWP by relating variation of meteorological fields with satellite-observed fire radiative power for a number of large US wildfires during 2018 and 2020. HWP is evaluated in terms of its ability to predict hourly fire radiative power. The HWP outperforms other existing indices particularly in the afternoon and early evening.

In addition, we explore the value of a realtime 10-h dead fuel moisture content (FMC) dataset, developed using a random forest model applied to numerical weather prediction output as well as polar-orbiting satellite observations, for enabling improved prediction of wildfire activity. We illustrate the evolution of the FMC dataset for several fires during the realtime demonstration period (since Oct 2023), and we explore the usefulness of incorporating the FMC dataset in the HWP to improve prediction fire activity.