Wildfire is an important disturbance affecting hydrological processes through alteration of vegetation cover and soil characteristics. The effects of fire on hydrological systems at the catchment scale are not well known, largely because site specific data from both before and after wildfire are rare. In this study a modelling approach was employed for change detection analyses of one such dataset to quantify effects of wildfire on catchment hydrology. Data from the Entiat Experimental Forest (Washington State, US) were used, a conceptual runoff model was applied for pre- and post-fire periods and changes were analyzed in three different ways: reconstruction of runoff series, comparison of model parameters and comparison of simulations using parameter sets calibrated to the two different periods. On average, observed post-fire peak flows were 120% higher than those modelled based on pre-fire conditions. For the post-fire period, parameter values for the snow routine indicated deeper snow packs and earlier and more rapid snowmelt. The net effect of the changes in all parameters was largely increased post-fire peak flows. Overall, the analyses show that change detection modelling provides a viable alternative to the paired-watershed approach for analyzing wildfire disturbance effects on runoff dynamics and supports discussions on changes in hydrological processes.
Keywords: conceptual runoff model, Entiat, land-cover change, runoff change detection, wildfire