Approximately 43% to 66% of freshwater resources in the contiguous U.S. are supplied by forest watersheds. The increase in burned area and severity and droughts associated with climate change has caused growing concerns about the effects of wildfire on U.S. forests and water supplies. Fire-induced vegetation losses are perceived to increase net precipitation, decrease transpiration, and damage soils, resulting in accelerated post-fire runoff and erosion. However, a comprehensive continental-wide assessment of impacts of wildland fires on streamflow is not available. Municipal water supply is directly affected by wildfires through changes in peak flows, base flows, and annual water yield.
Data used include historical records on wildland fire, climate, streamflow, vegetation dynamics based on remote sensing for over 100 watersheds across a large physiographical gradient. A framework incorporating traditional hydrometric analysis and modern statistical methods was developed to detect and quantify the effects of wildland fire on annual water yield. These analytical methods include change point detection models, double-mass analysis of precipitation and streamflow data, precipitation and flow duration analysis, climate elasticity models, and gradient boosting models.
Among the 162 watersheds examined, fire impacts were greatest in the Lower Colorado, Rio Grande and parts of the Texas-Gulf regions (i.e., >+100% increase in streamflow), especially in Arizona where burned watersheds had a high burned area to drainage area ratio (i.e., >50%). Our analysis showed that climate factors had the greatest influence on streamflow for all watersheds, but snow and monthly precipitation variance played a more dominant role in western regions. We also found that burned area and topography were important in the West, while land use and individual wildland fires were important in the East. We conclude that forest fire management strategies to enhance ecosystem resilience and to mitigate hydrological impacts from wildland fire must be developed to reflect local conditions.