3.2 Sea breeze simulation

**Figure 4:** Results of a sea breeze simulation: The island (reaching from X=50 km to X=200 km) in the center of the domain is much more heated by solar radiation than the water surfaces at the borders. A local low pressure system therefore is produced which leads to a sea breeze front with strong positive vertical wind speeds. The front travels from the coast to the center of the island during the day. The conditions at 3 p.m. are shown as an example.
$\begin{figure} \begin{center} \leavevmode \epsfysize=19cm \epsffile{island.eps} \end{center} \end{figure}$

**Figure 5:** Results of Xian and Pielke (1991) for the same situation as in Figure 4.
$\begin{figure} \begin{center} \leavevmode \epsfxsize=\textwidth \epsffile{island-orig.eps} \end{center} \end{figure}$

See breeze simulations are an ideal method to test the soundness of a model implementation, because they show the behavior and the coupling between fluid dynamics, turbulence, the soil module, and the radiation parameterization. A straight analytical solution to the problem does not exist. However, several studies have been carried out (Neumann and Mahrer; 1974; Pielke; 1984; Thunis; 1995; Xian and Pielke; 1991) and therefore it is known how the correct solution should look like. To test Metphomod a two-dimensional sea breeze situation with an island of 150 km extent was simulated. We tried to reproduce the same situation as the corresponding example in Xian and Pielke (1991). This was not fully possible since Metphomod uses different parameterizations for soil atmosphere interactions and for radiation than RAMS. The sea surface temperature, was set to 283 K, the potential land surface temperature and the potential air temperature just above the ground were set to 300 K. A potential temperature gradient of 4.5 K $\cdot$ km ^{- 1}, and an initial wind speed of 0 ms ^{- 1} were set as initial conditions. Soil parameters were set up to produce a similar heating effect as found by Xian and Pielke (1991). The simulation was started for 17 June, 6 a.m. Figure 4 shows vertical and horizontal wind speeds, and potential temperature at 3 p.m. The sea breeze front (SBF) can be clearly seen in the figure. It travels from to the coast to the center of the island according to theory (e.g. Simpson et al.; 1977). Although Metphomod uses different parameterizations than RAMS and was applied in non-hydrostatic mode, the results are similar to those of Xian and Pielke (1991) (Figure 5). The most important differences are: (1) the see breeze front has a smaller horizontal extent, (2) a region with subsiding air is produced in the middle of the island. This is due to the fact that Metphomod's turbulence parameterization scheme leads to a less efficient mixing in regions with low wind speeds which leads to a local high pressure area.