Representing frozen soil and its effect on infiltration in Earth System Models

Abstract:

Presence of ice in soil would reduce infiltration and increase surface runoff. Field studies in local-scale open areas showed that soil infiltration capacity is normally reduced due to the presence of ice. However, many filed studies showed that there are weak or no clear effects of frozen soil on infiltration and runoff generation especially in forested areas at large scales. Many land surface models (LSMs) for use in large scale Earth System Models show that explicitly including soil ice in LSMs degrades the simulation of runoff in cold regions.

In this talk, I will present the current status of representations of soil water phase change and its thermal effects in LSMs and more specifically on a parameterization scheme for infiltration under frozen soil conditions for large-scale applications guided by runoff observations. This scheme computes vertical water fluxes by introducing the concept of a fractional permeable area (of a model grid), resulting in schemes for computing matric potential and saturated hydraulic conductivity under presence of ice. The scheme produced favorable simulations of runoff and terrestrial water storages (TWS) when compared with observed discharge and monthly TWS derived from gravity signals of the Gravity Recovery and Climate Experiment (GRACE) satellites. This scheme is now adopted by some major LSMs for weather predictions and climate projections.