Forest Productivity for Soft Calibration of Soil Parameters in Eco-hydrologic Modeling

Abstract:

Calibration of soil drainage parameters in hydrologic models is typically achieved using statistics based on streamflow. Models that couple hydrology with ecosystem carbon and nutrient cycling also calculate estimates of carbon and nutrient stores and fluxes. Particularly in water-limited environments, these estimates will be sensitive to soil drainage parameters. We investigate the use of estimates of annual net primary productivity (annNPP) as an additional data source for soil parameter calibration. We combine literature-based estimates of annNPP with streamflow statistics to calibrate for soil parameters in three Western U.S. watersheds using a coupled eco-hydrology model. We show that for all sites, estimates of annNPP vary significantly across soil parameters selected solely using streamflow calibration. In all watersheds streamflow metrics select soil parameters that yield a range of annNPP estimates that can exceed literature-derived bounds for annNPP by 58-77%. Only 1-10% of the original soil parameter sets met both annNPP and streamflow criteria – a substantial reduction when compared to the percentage of acceptable parameter sets selected using annNPP or streamflow separately. Similarly, streamflow performance varies substantially across soil parameters selected based solely on annNPP criteria. Results show that annNPP in combination with streamflow-based metrics can better constrain soil parameters, although the usefulness varies across watersheds.