Progress in coupled meteo-oceanic predictions in the mediterranean region

This manuscript reports the results of the coupled atmosphere-ocean model MIAO (Model of Interacting Atmosphere and Ocean) implemented in the Mediterranean Sea.
The representation of the boundary layer in MIAO is based on the Monin-Obukhov theory, with an iterative solution on the equation relating the dynamical quantities at the lowest level of the atmospheric model to the Monin-Obukhov length and the SSR. The values adopted  for the SSR in MIAO are based mainly on observational studies  of the momentum flux (Large and Pond, 1981) and of the sensible heat and moisture fluxes (Large and Pond 1982),  and on theoretical works (Janssen, 1991 and Makin, 1999) that investigate the dependency of the SSR on the wave spectrum.
The observations give different SSRs for moisture,  z_0q, and for sensible heat, z_0t:
 

z0t=2.2*10-9	stable conditions	(1)
z0t=4.9*10-5	unstable conditions	(2)
z0q=9.5*10-5	unstable conditions	(3)

These values are approximately independent from the wind speed. An increase of the roughness for high wind values (U₁₀>25m/s) could not be established experimentally because of lack of data, but it is suggested by the theoretical study of Makin (1999). The variations of z_0m over sea are usually described by the Charnock relation. Some observations (Donelan et al, 1993) and theoretical studies (Janssen, 1991) suggest a dependence of z_0m on the wind-wave spectrum, also at moderate wind speeds. Therefore, according to the observations, z_0m, z_0q, and z_0t, have a different dependence on the wind speed and the wave spectrum. In the MIAO Model, the values of eqs.(1-3) are used and z_0q=z_0t in stable condition. The description of z_0m is based on the Janssen's theory (Janssen, 1991) that is included in the WAM model. The eventual dependence of z_0t and z_0q on the wind speed is not relevant for the cases discussed in this paper, because the maximum wind speed at the lowest model level turns out to be smaller than 25m/s.
 
 

The areas where  the models BOLAM, WAM and POM have been implemented. The areas of POM and WAM overlaps almost exactly. The dots show the grid ponits of BOLAM (a rotated grid is used).

SLP at 10 oct 12UT: results of the one-way coupling simulation (left panel, the contour lines interval is4hPa) and difference between one-way and tow-way coupling simulations (right panel, the contour lines interval is 1hPa). The shadowing indicates where the one-way coupling produces lower SLP values, i.e. tha absence of shadowing over most of the SLP field is associated to the deeper minimum produced by the two-way coupling.

Totale heat flux (positive upward) at 9 oct 21UT: one-way coupling simulation (left panel, the contour lines interval is 200W/m² )  and the difference  between coupled and uncoupled simulation (right panel, the contour lines interval is 50W/ m² ). In the left panel the shadowing indicates where the upward flux is larger than 200W/m² . In the right panel the shadowing  indicates where the upward flux in the one-way coupling simulation is more than 50W/m² . Note that the 0 level contour line, which has a very noisy behaviour, is ometted.

Time series of wind speed (top), SWH (center), sea surface elevation (bottom) at the CNR-platform 15Km offshore the Venetian coast. Only the metereological component of the sea surface elevation is shown, i.e. the astronomical tide has been subtracted from the plotted data. The x-axis reports the time in hours from the beginning of the simulation.