• Hourly progression of large U.S. wildfires was evaluated with GOES fire product. • Fast progression hotspots are in Central Great Plains and Northern California. • Increased fire progression rate was detected in the afternoon. • We identified temperature, humidity, and wind ranges linked to fast progression. Fast-growing wildfires pose severe threats to human communities and ecosystems, making it critical to understand their progression dynamics for effective suppression and emergency response. While satellite-based wildfire observations have been used to monitor daily fire growth, hourly wildfire spread patterns remain poorly characterized at the national scale. In this study, we developed an integrated dataset of hourly wildfire progression rates and associated environmental factors for 294 large wildfires across the Contiguous United States (CONUS) from June 2017 to December 2024, using GOES-16 (Geostationary Operational Environmental Satellite-16) fire detection and characterization data. Our results highlight the Northern California and the Central Great Plains as two critical regions experiencing fast progression (>10 km 2 /h) during large wildfires. The fastest recorded progression rate was 320 km 2 /h on February 27, 2024, during Texas “Smokehouse Creek” Fire. The longest duration of fast fire progression (184 h in total) occurred during California “Dixie” Fire, which reached its maximum rate of 137 km 2 /h on August 5, 2021. Diurnal patterns showed elevated progression rates in the afternoon, peaking around local 3 pm, corresponding to more fire-favorable weather conditions. Factors’ order based on association with progression rates ranked by partial Kendall correlation coefficient is wind speed, temperature, relative humidity, slope, and fraction of grass, forest, and shrub. The spatial and temporal patterns of fast hourly fire progression identified in this study offer critical insights for wildfire management and preparedness, and the dataset provides a valuable benchmark for evaluating and validating fire models in simulating hourly fire spread.
Fang et al. (Fri,) studied this question.