Evaluation of the Impact of Lateral Mixing in the New WRF 3D PBL Parameterization

Typical Planetary Boundary Layers (PBL) parameterizations are designed to describe vertical mixing relying on the assumption of horizontal homogeneity. However, this assumption is not valid when horizontal turbulent exchange processes are important, such as heterogeneous landscapes and mountainous terrain. Furthermore, high-resolution mesoscale simulations over heterogeneous land surface are becoming more common with the rapid increase in computational resources prompting the need to implement compatible PBL parameterizations. Based on the turbulence closure model described by Mellor and Yamada (1982) a new scheme was developed to include the lateral mixing as an integral part of the PBL parameterization in the dynamical core of the Weather and Forecasting Research (WRF) model. Specifically, the scheme is based on the level 2 Mellor-Yamada model and performs horizontal mixing based on the diagnosed horizontal turbulent fluxes in addition to vertical mixing. WRF simulations were conducted over a 10-day period on August 13-24, 2016, at 750 and 250 m resolution using a domain centered in the Columbia River upstream of the Columbia River gorge. The simulation period covers both clear-sky and overcast conditions. The results are compared with a commonly used vertical PBL parameterization, Mellor-Yamada-Nakanishi-Niino (MYNN), and evaluated against wind observations from the Wind Forecast Improvement Project 2 (WFIP2) including LIDAR, SODAR, radar and surface meteorological measurements, and against CloudSat profiles.