Catchment-scale hydrology simulations using inter-variable multi-parameter terrain descriptions

Abstract Global simulations of hydrology and hydraulics combine network-type routing with hydrological simulations on large Hydrological Response Units (HRUs). In order to efficiently simulate the hydrological response of HRUs, empirical equations are generally applied to averaged physical parameters or empirical constants. The usage of physically-based equations on a spatial description of the HRU could provide benefits in accuracy and scenario assessment. In this research we develop a multi-parameter inter-variable terrain description to simulate HRU hydrology and flow. Water is distributed over these classifications and used for application of physically-based equations. A mass-balanced scheme is presented that uses parameter inter-variability matrices to update the distributions of water. Furthermore, the method is used to simulate processes using physically-based equations that result in fluxes from and to other classes within a distribution, resulting in a semi-spatial process description within an HRU. Finally, physically-based equations are thus applied to a dynamic semi-spatial HRU description using multiple inter-variable physical parameters. A variety of study cases on hypothetical and real events are performed to highlight the behavior and sensitivity of the model, as well as comparison to existing methods. Un-calibrated comparison with full spatial flow simulations show reasonable agreement with a Nash-Sutcliffe coefficient of 0.47. Comparison to simulations of real events on catchments in Northern China and Central Spain show good comparability with an average Nash-Sutcliffe coefficient of 0.56. The performance of the model indicates good applicability when considering the efficiency of the method. Simulation time is in the ranges from fractions of seconds on consumer hardware without dependency of the area of the catchment. Additionally, the model uses physical parameters as input and physically-based descriptions of processes. This results in interactions between infiltration and flow that alter flow dynamics and hydrograph shapes in similar manners to full spatial hydrology and flow models. Finally, the method provides the possibility of linked additional processes such as groundwater and erosion, and could be applied on a global scale.