Antecedent Precipitation as a Predictor of Wildfire Burn Area in Southern California
Open Access
- Author:
- Marcial, David
- Area of Honors:
- Meteorology
- Degree:
- Bachelor of Science
- Document Type:
- Thesis
- Thesis Supervisors:
- Laifang Li, Thesis Supervisor
Raymond Gabriel Najjar, Jr., Thesis Honors Advisor - Keywords:
- Meteorology
Fire Weather
Precipitation
Climate
Seasonal to Subseasonal Prediction
GIS
Climatology
Southern California - Abstract:
- As favorable wildfire conditions become more frequent with anthropogenic climate change, an improved understanding of antecedent fire influencers is paramount to improving fire management and preparation and associated resource allocation in fire-prone regions such as the chaparral-dominated landscape of southern California. While many fire weather studies include a precipitation component, antecedent precipitation windows tend to be few in number and predetermined, thereby disallowing a more thorough exploration of burn area (BA) – precipitation connections. We overcome this gap by creating 18 three- and six-month sliding windows each of antecedent precipitation and characterizing the BA – precipitation correlations at each lead time over the spatial domain of southern California. These correlations are strongest and most positive two wet seasons prior to peak fire season, reaching significance with logarithmic BA as the response variable. The six-month windows generally outperform their three-month counterparts. The correlations weaken with decreasing lead time and transition in sign around one summer and autumn before peak fire season. The trend of correlations is itself significant and consistent with other studies assessing BA – precipitation connections in vegetation-limited regions of the world, implying a dynamic process of vegetation growth and drying that provides a foundation conducive to wildfire spread. These relationships fail to characterize two of southern California’s most severe fire seasons, which were preceded by at least two dry years, suggesting that longer-term mechanisms drive such extreme fire activity. Although statistically insignificant, we conduct a secondary analysis that finds a positive correlation between BA and the difference in precipitation between the two previous wet seasons (two years prior – one year prior).