The purpose of this research is to model and provide better understanding of atmospheric flows in deep valleys. The final aim is to simulate a peculiar wind, the Ora del Garda, caused by the interaction between a lake breeze and a valley wind in the Garda Valley (in Northern Italy). In addition, we want to make the shading subroutine described below available for VTMS simulations in Salt Lake Valley. In order to achieve this, it was decided to use the Advanced Regional Prediction System (ARPS), an operational, three-dimensional, compressible, and non-hydrostatic modeling tool, developed by the Center for Analysis and Prediction of Storms at the University of Oklahoma.

One of the core needs, when modeling valley winds, is to simulate accurately inversion-layer breakups. These needs were fully documented by Whiteman's field data and Anquetin's numerical simulations.

When effects of synoptic-scale gradients are neglected, solar radiation is the only factor that drives inversion-layer breakups in steep valleys. Destruction of this stable layer generally occurs before 10:00 AM. An accurate model of incoming solar radiation is therefore crucial in the morning. ARPS provides an extensive radiation package that includes the incidence angle of sunrays in its computations. For instance, non-exposed mountain slopes do not receive incoming solar radiation. However, neither ARPS 5.0 nor, as far as we are aware, other major atmospheric simulation codes take into account the shade induced by topography. This is particularly important when dealing with real topographies, which induce delays in the sunrise of one hour or more, as observed in field measurements by Whiteman.

To account for shade and correct the radiation model, we created a new subroutine. For each grid point, this subroutine checks whether or not a virtual line, in the direction of the sun, hits an obstacle. If it does, incoming solar radiation is set to zero at the original grid point. Under the assumption that the topographic data is precise enough for this type of analysis, the topographic details are found by a linear interpolation on grid cell boundaries.

The paper describes the shade technique methodology in more detail, presents numerical simulation results, and highlights the topographies and times in which the shade routine is needed. We use simulations for different idealized valleys to demonstrate how topography influences inversion-layer breakups. The simulations show that the time-scale of inversion-layer destruction for idealized valleys is consistent with field data gathered by Whiteman for similar valleys. These results allow us to believe that this new ARPS subroutine will significantly enhance the simulation of inversion-layer breakups in steep valleys.

The support of AC by TotalFinaElf and of TdeC and RS by the VTMX Program of DoE is appreciated.