Variability and persistence of hillslope initial conditions: A continuous perspective on subsurface flow response to rain events

Abstract

► HYDRUS simulations examine event response to hillslope-scale initial conditions. ► Hillslope storage exhibits preferred ranges for wet and dry seasons. ► Hillslope field capacity separates wet and dry season preferred storage ranges. ► Subsurface stormflow requires an initial storage threshold near field capacity. ► Total subsurface stormflow is a nonlinear function of initial storage. Runoff response to rain events depends on the initial moisture conditions in the subsurface. This study explores subsurface stormflow response to initial conditions within the context of a continuous hillslope water balance. A hypothetical hillslope with three-dimensional variably saturated subsurface flow is developed using the HYDRUS model forced with a year-long sequence of hourly precipitation and transpiration from Seattle, WA, USA. Using six different soil hydraulic parameter sets, test simulations examine (1) variability of hillslope initial conditions prior to rain events, (2) persistence of initial conditions in a continuous simulation, and (3) effects of initial conditions on subsurface stormflow during rain events. Results show that hillslope initial conditions vary seasonally, producing bimodal distributions of storage values with preferred storage ranges for wet and dry seasons. Preferred storage ranges differ by soil texture and by hydraulic conductivity. Wet season initial conditions are most frequently at storage values above hillslope field capacity, with higher preferred ranges of storage for scenarios with lower saturated hydraulic conductivity. Dry season hillslope storage values converge to minimum values below field capacity, and these minimum values vary with soil texture but not with saturated hydraulic conductivity. Tests of initial condition convergence show that scenarios with different initial storages can eventually converge under either persistent wetting or persistent drying until hillslope storage reaches the wet or dry preferred states. Dry preferred states are unlikely to produce subsurface stormflow, as test rain events generate hillslope outflow only when initial storage is at or above an event and parameter-specific threshold near the hillslope field capacity. Above this flow threshold, all simulations show a nonlinear increase in subsurface stormflow with increasing initial storage, with the steepest rates of increase for scenarios with the highest values of saturated hydraulic conductivity. Simulation experiments present a quantitative approach for deriving functional relationships between event response, initial storage, and hillslope water retention.