Abstract
Decades of wildfire suppression combined with increasingly long and dry fire seasons have led to more frequent and severe wildfires, posing risks to municipal water supplies due to elevated runoff and erosion. Land managers increasingly rely on decision-support tools like the Water Erosion Prediction Project (WEPP) model to identify areas critical for hydrologic recovery. However, WEPP traditionally adjusts only infiltration and erodibility parameters and does not account for dynamic post-fire soil changes.This study aimed to improve WEPP’s predictive ability by quantifying fire-induced changes in soil properties and assessing their effects on streamflow and sediment yield. An in-situ burn experiment was conducted on loam and sandy loam soils in northeastern Washington, examining relationships between soil burn severity and soil response. Observed soil changes were then integrated into a WEPP sensitivity analysis across three catchments (22–100 ha) at the Hinkle Creek Paired Watershed Study near Roseburg, Oregon.
The field experiments revealed that soil texture displays a fining trend with soil heating, with a decrease in sand and an increase in silt, related to temperature. Bulk density initially decreased before increasing above unburned conditions 1-year after burning. Water repellency was measured on all plots, but little correlation was found between litter consumed and maximum temperature. Plant available water increased in 4 out of 5 burned samples, driven by both increases in field capacity (0-10 kPa) and decreases in wilting point (1500+ kPa). Total phosphorus remained stable across all treatments, while water-soluble P showed significant (p=0.002) declines in the burned plots, particularly in coarser soils.
The WEPP model adequately simulated pre-fire streamflow conditions (NSE: 0.41–0.51; KGE: 0.50–0.73). Applying these observed soil changes collectively to a 22-ha watershed at Hinkle Creek increased streamflow production, especially in early season peaks. These changes were not accompanied by increases in sediment, indicating that they were likely linked to increasing lateral flow rather than runoff. It is recommended that similar analyses be done in areas that experience higher intensity rainfall and are susceptible to infiltration-excess runoff to better understand how incorporating dynamic soil properties influences the predictability of post-fire decision support tools.