Hydrologic effects of fire in a sub-alpine watershed: AgES outperforms previous PRMS simulations

Authors: Mankin, Kyle; Wells, Ryan; KIPKA, HOLM - Colorado State University; Green, Timothy; Barnard, David

Submitted to: Journal of the ASABE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/22/2022
Publication Date: 8/1/2022
Citation: Mankin, K.R., Wells, R.M., Kipka, H., Green, T.R., Barnard, D.M. 2022. Hydrologic effects of fire in a sub-alpine watershed: AgES outperforms previous PRMS simulations. Transactions of the ASABE. 65(4):751-762. https://doi.org/10.13031/ja.14881.
DOI: https://doi.org/10.13031/ja.14881

Interpretive Summary: Two watershed models, AgES and PRMS, both accurately simulated the dramatic change in streamflow response to precipitation after fire. Annual streamflow increased by 170% in the postfire period compared to the prefire period. Postfire AgES parameters were drastically reduced for snow and rain interception, canopy density, and ET. Decreased soil depression storage and field capacity also contributed to increased streamflow response in AgES. AgES shows promise for simulating hydrologic impacts of sub-alpine forest resource management and fire response.

Technical Abstract: Forest fires alter hydrologic responses to precipitation, and quantification of streamflow changes can be challenging to simulate. Streamflow data before and after wildfire in a 14 km2 mixed-conifer, sub-alpine watershed in south-central New Mexico were used to calibrate a distributed watershed model (AgES) for prefire and postfire conditions. Results and parameterizations from AgES and a prior PRMS modeling study in the same watershed were compared. Both AgES and PRMS accurately simulated the dramatic change in streamflow response to precipitation after fire, including a smaller precipitation threshold to induce streamflow, although AgES had substantially less bias and better performance in monthly calibrations. Only AgES was assessed at the daily scale. AgES closely simulated daily streamflow throughout the study period, with slight overestimation in the prefire period (Oct 2007 to Oct 2011, 1.2% bias, 0.90 Nash-Sutcliffe efficiency [NSE]) and in the postfire period (Oct 2013 to Oct 2017, 4.4% bias, 0.54 NSE). Although each 4-year model period had nearly identical cumulative precipitation, annual streamflow increased by 170% in the postfire period compared to the prefire period. Appropriate soil and vegetation parameters were modified by the stepwise calibration process to represent the effects of fire in AgES: interception storage was decreased for rain (-99.8%) and snow (-26%), two factors influencing ET were decreased, soil depression storage was reduced (-97%), and soil field capacity was reduced (-50%). AgES demonstrated improved simulation of streamflow compared with PRMS, calibrated parameter shifts that were more consistent to interpret, and particular skill in simulating daily streamflow response to fire.