Traversing the Gray Zone with Scale-aware Turbulence Closures

University Of California-Berkeley, Berkeley CA

Investigators

Abstract

With advances in computing power, atmospheric models for weather prediction are being used at higher and higher resolutions. Weather prediction models were originally designed to operate at about 10-km grid resolution. At about 1-km grid resolution, atmospheric models enter the so-called “gray zone,” where traditional physics parameterizations no longer hold, leading to errors in weather prediction. This project will create a new scale-aware framework to improve these parameterizations in the gray zone, with broad impacts for weather and regional climate forecasts, which increasingly use models in the gray zone for practical applications. The researchers will collaborate with scientists at NCAR and NOAA to guide this process and to make sure that the scale-aware approach is thoroughly tested and ready to implement in operational codes. Ultimately, this will further enable a vast array of applications at gray-zone and finer resolutions, such as weather prediction, wind energy forecasting, wildfire smoke modeling, and air quality modeling in urban environments. This project will also provide connections with the environmental justice community of concern in Stockton, California, where two teachers will be provided with summer research opportunities to develop a hands-on air pollution curriculum which will reach 700 high school students in the first two years. The overall goal of this research is to extend scale-aware turbulence parameterizations into the gray zone to allow a consistent representation of turbulent and resolved motions as grid resolutions increase. The scale-aware turbulence framework will be tested in the WRF model with a variety of physically and mathematically consistent closure models, with a focus on the dynamic reconstruction model (DRM) and simpler “mixed models”. Mixed models combine a scale-similarity component with an eddy-viscosity model; mixed models are scale aware and have been shown to allow mesoscale models to better operate in the gray zone. This project will extend and thoroughly evaluate mixed models across the gray zone. New strategies for grid nesting will also be tested, including adjusting grid nesting ratios and using the FastEddy GPU code with very large domains to explore skipping the gray zone. Simulations will be performed over the Southern Great Plains, an area of interest for wind resources, and over Northern California, including inland locations with poor air quality and offshore regions being considered for wind energy development. Long-term simulations will be performed to assess improvement in model skill. The project will also take initial steps to include these turbulence closure innovations in operational forecasting. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.