Quantifying the Impact of Land Surface Initialization on Southern Great Plains Low-Level Jet Forecast Skill

Numerical Weather Prediction (NWP) is the backbone of current decision support for weather-sensitive activities. Over the years, marginal gains in forecast skill have been achieved through the utilization of more frequent and more accurate atmospheric observations. Comparatively less attention has been focused on the role of soil moisture on forecast skill due to an inadequate ground observational network. This study investigates how incorporating NASA high-resolution (9km) Soil Moisture Active-Passive (SMAP) satellite-based soil moisture retrievals will refine short-range forecast accuracy of the onset, maintenance, and decay of the southern Great Plains low-level jet (LLJ). The LLJ is a regular, periodic, and dominant feature over the region that critically supports agricultural and wind energy production by conveying both water vapor and wind north from the Gulf of Mexico. Our NASA Unified Weather Research and Forecasting (NU-WRF) model analyses span the warm-season months (May-September) of a recent 5-year period (2012-2016) and focus on a subset of particularly strong LLJ events incident on the Sweetwater, Texas wind farm. For each event, we quantify incremental forecast improvement (or degradation) due to prescribed or assimilated SMAP soil moisture relative to the default coupled model forecast. Additional model realizations are generated to separately isolate the contribution of real-time vegetation parameterization to forecast skill. Overall focus is placed on evaluating the forecasts of LLJ strength, nose height, vertical sheer, diurnal cycle, duration, and position. Improving predictability of the LLJ has huge potential scientific and economic benefits.