Influence of soil moisture gradient on convective cloud development during HI-SCALE

Description

The impact of soil moisture gradients in convective cloud development is studied using numerical simulations with a 300 m grid spacing over a 300 km wide domain, centered at Department of Energy’s Atmospheric Measurement Southern Great Plains (SGP) site. Simulations are run for August 30, 2016, during the Holistic Interactions of Shallow Clouds, Aerosols, and Land-Ecosystems (HI-SCALE) field campaign. Each of five simulation starts from an identical atmospheric initial condition but with different soil moisture gradients, ranging from a spatially uniform distribution to those with a strong spatial gradient. One simulation, initialized with the observed soil moisture pattern, shows a good agreement with observed surface fluxes and boundary-layer state across the SGP site. Different soil conditions result in varying evaporative fraction (EF) and distinctly different distributions of clouds and precipitation. Key processes linking the changes in EF and changes in clouds are examined by dividing the whole domain into 30-km wide subdomains, similar to the grid cell sizes of global climate models. Applying the soil moisture-cloud feedback framework proposed by Findell and Eltahir to the 30-km subdomains, we found their framework qualitatively depicts how soil moisture differences affect the simulated cloud development and precipitation. The morning profiles over most of the domain are close to moist adiabatic in the lower troposphere, making higher EF more conducive to convection (wet soil advantage) by increasing the moist static energy of the boundary-layer (BL) air. In these areas, we see strong increase in precipitation and cloud top height being associated with increased EF. A small fraction of the domain is suggested as "dry soil advantage", and show reduced precipitation with increased EF. However, the relationship between EF and convective clouds is strongly non-linear, and as the soil moisture gradient increases, we see less qualitative agreement with this frame work, such that areas with reduced EF also produce greater precipitation and taller clouds. The disagreement is attributed to advection of BL air properties and developing clouds, as well as disturbances by cold pools that occur within the model domain. Importance of these processes at the 30-km scale are assessed through budget equations and the joint probability density functions of the boundary-layer moisture and temperature as implemented in some cumulus parameterizations.