LAM Simulations of Present-Day Climate with observed Boundary Conditions: Performance Analysis over the Northern Italy

The climatic version (CLAMBO) of the limited area model of the Regional Meteorological Service of Bologna (LAMBO), which is in use as an operational short range weather prediction model since 1993, has been applied to climate simulation over the European area. This version has an approximate horizontal resolution of 30 km by 30 km with 20 vertical levels, with a physical package including moist processes, interactive radiation, vertical diffusion and a description of the biosphere based on BATS. To establish the LAM ability to reproduce the present-day climate features of the surface parameters, as precipitation, air temperature and turbulent fluxes over the Northem Italy, CLAMBO has been integrated for 19 months (October 1989 - April 1991), using observed boundary conditions. Starting from the basic remark that CLAMBO is able to reproduce the observed monthly mean circulation (mslp and 500 hPa geopotential height), model-produced surface air temperature and precipitation were compared with observation gathered in 155 stations distributed over the Northem Italy. The bias of the monthly mean precipitation fields over the whole area is quite small during all the integration period, and, over the Northem Italy the base-model run reproduces well the seasonal cycle of precipitation. Only looking at the subareas we can find an overestimation of rainfall over North-West and Central Alps, while the Po Valley presents an underestimation. Regarding 2 meter temperature, our results show that overall surface air temperature over the Northem Italy is poorly simulated by the model, mostly during the winter periods. The disagreement appear to be related to the feedback between the air water content and the short wave radiation at the surface, as also observed in the precipitation bias spatial structure. However the strong winter bias doesn't seem to produce a strong effect on upper air fields. To assess the impact of biospheric scheme change on the quality of evaluated two meter temperature and rainfall, an additional run has been performed with the bare soil scheme. Because over Northem Italy there are not observations of all components of the surface energetic budget at the temporal scale of interest we supply to this data deficiency by using the outputs of stand-alone version of biospheric model as data sources. Then we simulated, in one point (San Pietro Capofiume synop station), a 19 months time series (control data set) using a stand-alone version of surface scheme BATS. Comparing in one point the two CLAMBO simulations against the control data set, it was possible to identify some deficiencies in the radiation and surface scheme that could explain the temperature bias.