WEATHER FORECAST METHODOLOGIES

Model Configuration

Regional forecasts are generated daily with the PC Linux version developed by AMI Environmental (AMI) of the advanced, three-dimensional mesoscale model MM5 (Version 3) . The MM5 mesoscale model is a limited-area, nonhydrostatic, terrain-following sigma-coordinate model designed to simulate or predict mesoscale and regional-scale atmospheric circulation. It has been developed at Penn State University and National Center of Atmospheric Research (NCAR) since the 1970s as a community mesoscale model. It is continuously being improved by contributions from several universities and government laboratories as well as private consulting firms around the world.

The MM5 forecasts use two modeling domains that cover Texas and parts of neighboring states and northern Mexico. The outer domain (domain D01) consists of 42 x 42 grid points, spaced 45 km apart with 25 vertical levels. The inner domain (domain D02) contains 76 x 76 grid points, spaced 15 km apart also with 25 levels.

• Maps of 45-km Domain: Modeling Domain , Terrain Elevations , Land Uses
• Maps of 15-km Domain: Modeling Domain , Terrain Elevations ,  Land Uses

• National Centers for Environmental Prediction (NCEP) MRF planetary boundary layer scheme
• Dudhia (1989) radiation scheme
• 5-layer soil model
• Grell (1993) cumulus parameterization
• Explicit moisture scheme (including simple ice physics but no mixed phase processes)

Initial conditions are generated by interpolating the NCEP ETA analyses to the MM5 grids. Lateral boundary conditions are derived from the ETA forecasts and updated every 6 hours throughout the MM5 forecasts.

Forecast

The initialization and forecasts for both grids are completed on an AMD SMP (symmetric multiprocessor) Linux PC . Two MM5 forecasts, valid for 48 hours, are available daily for 00 UTC (1800 CST ) and 12 UTC (0600 CST).

Applications

To further resolve the complex terrain features and other localized effects, nested grids of finer spacings (5 km and 1.67 km) are used in forecast applications for wind energy plants. The MM5 forecasts are also coupled with diagnostic wind models for fine scale modeling (1 km or less in spatial spacing). Predicted winds and other meteorological fields are used as inputs to air quality simulation models such as CALPUFF for inert pollutants, and UAM and Models3 for ozone and PM2.5.