Examination of different water balance models for use with tree-ring proxy records to extend hydroclimatic records

Abstract:

The use of a mechanistic hydrologic approach with tree-ring data to extend hydroclimatic records provides information on past variability, thus improving assessments of future scenarios while also enabling the examination of factors that may alter streamflow independently of climate, such as changes in land use or land cover. In this study, we examined two seasonal water balance models as part of a mechanistic approach to reconstruct streamflow with proxy inputs of precipitation and air temperature: a Thornthwaite water balance model modified to operate on a seasonal time step, and a simple water balance model with a snow component. Both models were fit using PRISM and proxy seasonal temperature and precipitation, and their ability to reconstruct streamflow was evaluated for the upper reaches of the West Walker River basin at Coleville, CA. In addition to calibrating both models with a shuffled complex evolution approach, sensitivity analysis was performed to estimate parameter uncertainty. Because the modified Thornthwaite model performed best during calibration (R² values of 0.96 and 0.80 using PRISM and proxy inputs, respectively), it was then used to reconstruct streamflow during AD 1500-1980 for the target basin. In addition to producing similar wet and dry episodes as other regression-based reconstructions for the Great Basin, our mechanistic approach allowed for estimating actual evapotranspiration and April 1 snow water equivalence in addition to streamflow. This study shows that mechanistic models that can operate with limited input requirements provide options for proxy-based extensions of hydroclimate records in areas with limited instrumental records, and help characterize non-climatic reasons for streamflow variability during the reconstruction period.