Global ecoregion-scale variations in wildfire seasonality revealed by daily satellite data

Abstract Background The increasing frequency and extent of wildfires since the twenty-first century have raised growing concerns about their ecological impacts, yet their spatial–temporal dynamics remain inadequately understood. Satellite remote sensing provides a unique opportunity to capture and quantify wildfire seasonality across ecosystems through its extensive and continuous observation. To this end, we utilized the 0.25 × 0.3125° gridded daily fire radiative power (FRP) data to analyze wildfire seasonality across ecoregions from 2001 to 2022. Results Globally, fire season durations typically ranged from 30 to 120 days. Longer seasons (> 120 days) primarily occurred in monsoon-influenced and desert-climate ecoregions, while shorter seasons (< 60 days) appeared at high latitudes or rainforest ecosystems. Although at least one fire season per year was observed in 76.5% of ecoregions, secondary fire seasons were found in 32.6% of ecoregions, predominantly in the Northern Hemisphere. Over the 22 years, wildfire seasonality exhibited diverse trends in timing and intensity. Start and peak dates became earlier in 63.7% and 52.9% of ecoregions, respectively. Season durations increased in 62.2% of ecoregions, with significant expansion ( p < 0.1) in 15.4% (average + 2.8 days year⁻ 1 ). Seasonal fire intensity also demonstrated notable variations. Seasonal sum FRP significantly increased in 14.6% of ecoregions (average + 1500 MW year⁻ 1 ), but it decreased in 11.5% of ecoregions (average −3700 MW year⁻ 1 ). Seasonal daily average FRP significantly intensified in 10.9% of ecoregions (average + 46 MW year⁻ 1 ), but it declined in 13.0% of ecoregions (average –45 MW year⁻ 1 ). Furthermore, three distinctive fire season dynamics were identified: intensified intensity without prolonged season duration (occurred in 16.0% of ecoregions), diminished intensity with prolonged fire season (happened in 4.0% of ecoregions), and intensified intensity with prolonged fire season (appeared in 3.4% of ecoregions). Conclusions Our study established a globally applicable framework for identifying and quantifying fire seasons from daily FRP time series at the ecoregion scale, and highlighted ecoregions with significant long-term trends in fire seasonality.