Experimental forecasts of mountain waves for the Terrain-Induced Rotor Experiment (T-REX)

The field phase of the Terrain-induced Rotor Experiment (T-REX) took place in the Spring 2006 to the lee of the Sierra Nevada. A large number of special observations were taken in the Owens Valley, California, and adjacent regions to document the characteristics of mountain waves, rotors and their environment. T-REX also involved a numerical modeling component: several models were run in realtime and retrospective mode to forecast mountain wave development.

At NOAA Earth System Research Laboratory, two configurations of the Weather Research and Forecasting (WRF) model were run as part of T-REX. The two configurations used identical initial and boundary conditions, similar domain and grid spacing (2 km), and the same physics packages, but differed in dynamic core. One setup employed the Advanced Research WRF (ARW) core, developed by the National Center for Atmospheric Research, while the other used the Non-hydrostatic Mesoscale Model (NMM) core, developed by NOAA National Centers for Environmental Prediction.

Subjective analysis of the results indicate that the two cores of the WRF model yield significantly different mountain wave forecasts. Moreover, sensitivity tests indicate that the simulated mountain waves are sensitive to the timestep used and to the formulations for upper boundary condition and damping used. Results from the ARW and NMM will be presented and their forecasts will be intercompared. Additionally, a preliminary comparison between model forecasts and observations, with emphasis on aircraft data, will be presented.