Intercomparison between BATS and LSPM surface schemes, using point micrometeorologicaldata set

This work has been developed with the aim to create an archive of climatological values of sensible, latent and ground-atmosphere heat fluxes in the Po Valley (CLIPS experiment); due to the unavailability of climatological archives of turbulent fluxes at synoptic scale, we have used the outputs of stand-alone runnings of biospheric models; this archive could be used to check the parametrizations of large- and meso-scale models in the surface layer. We started to check the reliability of our proposal by testing the model outputs by a comparison with observed data. We selected a flat, rural area in the middle-east Po Valley (San Pietro Capofiume, Italy) and used the data gathered in the experimental campaign SPCFLUX93 carried out there. The models adopted for the intercomparison have been the Biosphere-Atmosphere Transfer Scheme (BATS) of Dickinson et al. (1986 version) and the Land Surface Process Model (LSPM) of Cassardo et al. (1996 version). An improved version of BATS has been implemented by us changing in a substantial way the soil thermal and hydrological subroutines. The upper boundary conditions used for all models were taken by interpolating the synoptic observations carried out at San Pietro Capofiume (Italy) station; the algorithm used for the interpolations was tested with the data achieved in a fortnight campaign (SPCFLUX93) carried out at the same location during June 1993, showing a good agreement between interpolated and observed variables. Two experiments have been carried out; in the first one, the vegetation parameter set used by BATS has been used to force all models, while in the second one a vegetation cover value closest to the observations in the site has been used. The main results of this paper can be summarized as follows: in spite of the simplicity of the relationships used to evaluate global and net radiation (starting from synoptic observations), the produced results were of good quality; the use of local characteristics of the vegetation allowed us to reproduce more accurately the climatic characteristics of the location for all models; the predictions of turbulent fluxes given by LSPM are closer to the observations in both experiments, while the improvements of the modified version of BATS become more evident if the original BATS data-set (for vegetation cover) is used; finally, all models seem to give a satisfactory representation of the propagation of thermal wave into the soil.