The Colorado State University - Regional Atmospheric Modeling System (CSU-RAMS) is being used to examine linkages between the Mexican monsoon and precipitation in the Great Plains region of the United States. Currently, seasonal simulations have been performed for July through September of the 1993 flood year in the midwest US and the 1997 El Nino year. Results of monsoon season-long control simulations and the response to changes in soil moisture and sea surface temperature will be presented.

The model configuration consists of a 120km resolution coarse grid that covers a region from west of Hawaii to Bermuda and from south of the equator up into Canada. Two 40km resolution nested grids exist, with one covering the western two-thirds of the United States and Mexico and the other covering the Pacific ITCZ. The model is initialized with NCEP reanalysis data, surface observations, rawinsonde data, variable soil moisture, and weekly averaged SST's. RAMS is running with two-stream Harrington radiation, one moment microphysics, and Kuo cumulus parameterization.

The completed 1993 and 1997 seasonal simulations have been examined and verified again NCEP reanalysis data and high resolution precipitation data. Model results look promising when verified against the NCEP upper level fields, such that the model is able to capture the large scale dynamics. For the duration of both seasonal runs, RAMS successfully simulates the mid-to-upper level geopotential heights, the temperature, and winds. The large scale mid-level anti-cyclone over the US and Mexico is resolved, as well as the easterly flow over Mexico. Monsoon surge events in the Gulf of California have also been isolated and verified against reanalysis and precipitation data.

Within the 40km grids, the RAMS model has successfully resolved the low-level jet and local maximum in mixing ratio that persists over the Gulf of California. It has also captured the upslope flow over the Sierra Madre Occidental (SMO) that provides the necessary lifting and moisture for the development of intense convection and resulting large amounts of precipitation that occur along this mountain range. Examination of model-predicted low-level moisture transport reveals that moisture advected from the Gulf of California is the primary monsoon moisture source, rather than the Gulf of Mexico. Time averages of moisture transport, mixing ratio, winds, and precipitation reveal the prominent diurnal variations that exist due to radiative effects and land-sea interactions; the maximum in convection, precipitation rate, and moisture transport occurs around 00Z in the model. Time-lapse of moisture transport and mixing ratio also reveals that the downstream branch of gulf surges can extend well into the mountains of Colorado and the western high plains, thereby, acting as an additional moisture source for precipitation enhancement over the Great Plains and mid-west several days following a surge event.

Seasonal accumulated precipitation amounts in the model are successful in predicting the placement of precipitation and relative amounts for most of the 40km continental grid, but there is an overestimation of precipitation along the SMO and an underestimation in the US mid-west. During the 1993 flood summer, much of the mid-west US precipitation fell in association with mesoscale convective systems that are not explicitly resolved by the model.