Numerical Simulations of the Influence of Land Surface Heterogeneity during the Land-Atmosphere Feedback Experiment (LAFE)

Land surface characteristics can play a large role in the surface energy budget, modulating the sensible and latent heat fluxes that drive boundary layer (BL) turbulence and circulations. Due to this importance, the Land Atmosphere Feedback Experiment (LAFE) was conducted at the Atmospheric Radiation Measurement (ARM) site in northern Oklahoma in August 2017. One goal of the experiment was to determine how differences in land characteristics influence BL structure. To accomplish this goal, a suite of instruments were deployed over a roughly 10 km x 10 km region, and included micrometeorological towers, aircraft, soundings, and remote sensing platforms such as lidars and radiometers.

In order to gain some insight on the scale and variability of surface fluxes that influence turbulent kinetic energy (TKE) and BL structure, large eddy simulations (LES) were conducted. The sensitivity of BL development and structure to heterogeneities in land surface characteristics were evaluated. Of particular interest is to determine how different scales in surface heterogeneity and magnitudes of gradients between neighboring land surface types affect the development of nonclassical mesoscale circulations (NCMCs). These circulations are important as they can be tied to the development and location of BL cumuli and convection initiation (CI) by creating sustained mesoscale updrafts and by redistributing low-level water vapor and heat. Previous studies have examined how certain heterogeneities can affect the development of NCMCs under idealized conditions. LAFE was successful, as numerous intensive observation periods (IOPs) were conducted and used to initialize and verify the simulations under varying atmospheric conditions. Also, given the duration of the experiment (~ 1 month), the simulations can provide longer-term statistics such as the evolution of vertical velocity variance and TKE to compare to the observations. Such statistics are important to improve BL parameterizations.