WRF Inversion Base Height Ensembles for Simulating California Summertime Marine Boundary Layer Clouds

It is well known that the marine boundary layer (MBL) stratocumulus clouds play an important role in climate and weather in California, especially during the summertime. Solar energy generation forecasting is particularly impacted by MBL clouds as distributed solar power systems are abundant in coastal areas. Inland penetration and lifetime of MBL clouds are often believed to be largely determined by the inversion base height (IBH) and inversion strength. However, the MBL clouds are poorly modeled in numerical weather prediction (NWP) due to the difficulty of accurately simulating turbulence, microphysics, and radiation in the boundary layer and the interaction with the inversion and MBL clouds. We found that the North American Model (NAM) as well as Weather Research and Forecasting model (WRF) initialized with NAM simulated an IBH that was biased low compared with measurements from operational radio soundings. The WRF simulates insufficient cloud cover before sunrise and tends to have shorter lifetime post-sunrise. We hypothesize that the erroneous IBH in the NAM initial conditions as well as the WRF Planetary Boundary Layer (PBL) schemes caused the shallower PBL leading to underforecast of cloud extent and cloud liquid water. First, a simple one dimensional mixed layer model is used to understand how the cloud liquid water tendency changes as a function of IBH. Then, an analysis of WRF initialization time revealed that WRF initialized at 6 UTC with 0 UTC NAM provides the most accurate solar forecast. WRF IBH ensembles are developed by increasing the IBH while holding the inversion strength unchanged. Validation against Global Horizontal Irradiance (GHI) measurements from San Diego Gas & Electric (SDG&E) Weather Stations and Clean Power Research SolarAnywhere dataset showed that the WRF IBH ensembles reduce GHI biases and further the inland penetration of MBL clouds.