Some effects of soil and vegetation databases on spectra of limited-area mesoscale simulations

Results are presented from an analysis of variance as a function of horizontal scale. The normalized difference-field spectra of kinetic energy, temperature, vapor mixing ratio, and cloud-water mixing ratio were computed as a function of wavenumber at several model levels within and just above the planetary boundary layer (PBL). The analysis was performed on simulations from a three-dimensional (3D) hydrodynamic mesoscale model that contained a soil-vegetation canopy model. The analysis was intended to highlight (in terms of wave spectra) the impact of changes in lower-boundary forcing through horizontal variations in soil and plant type. Experiments showed that the use in the model of 1 degrees resolution databases of soil and vegetation type produced higher amounts of variance in the simulated fields at most wavelengths--often by more than 10%--when compared to a simulation that utilized a uniform distribution. Furthermore, the use of databases generated by random specification of soil and vegetation types resulted in yet higher amounts of variance at most wavelengths. The normalized difference-field spectra of energy, temperature, and water vapor mixing ratio generally displayed positive slope (largest values at highest wavenumber) at the lowest model level and tended toward negative slope at higher levels. The magnitudes of the spectra also diminished rapidly with height. The effect of the lateral boundary conditions was generally much greater in terms of the spectral magnitudes than that due to the soil-vegetation databases.