ABSTRACT
Climate is one of the most important drivers of regional fire activity, and recent disastrous wildfires in several regions of the world may signal fire–weather tipping points. There is uncertainty regarding the degree to which surface fire–weather variables influence anomalous fire activity in India’s Western Ghats. We compiled a comprehensive daily time series of MODIS fire detections between 2004 and 2009 from the MODIS thermal anomalies dataset (Collection 6: MOD14A1, and MYD14A1) for the Nilgiris landscape. Simultaneously, we also obtained daily weather data from the Mulehole automatic weather station, also located in the Nilgiris landscape. We then determined the meteorological drivers of daily fire–weather anomalies and compared these with the number of MODIS fire detections in the Nilgiris landscape. The mean daily potential evapotranspiration (PET) was highest in 2004, 3.8 ± 0.9 mm, and lowest in 2008, 2 ± 0.6 mm. The mean daily MODIS fire detections were also highest in 2004, 2.5 ± 4.1, whereas the lowest mean daily MODIS fire detections were in 2006, 0.7 ± 2.1. Both the mean daily PET and the mean daily MODIS fire detections were significantly higher in 2004 compared to all other years (P < 0.01). The daily variability in wildfire incidence in 2004 is strongly associated with PET (Spearman’s correlation = 0.69, P < 0.01). Of the weather metrics, daily PET in 2004 demonstrated the strongest relationship with daily MODIS fire detections (R 2 = 0.33). Our findings establish that daily synoptic weather patterns, as characterized in hot and dry years, could be significant predictors of the anomalous fire activity at landscape scales in the Western Ghats, India.
