| Abstract: | The goal of the Project for Intercomparison of Land-surface Parameterization Schemes (PILPS) is to improve the understanding of the interactions between the atmosphere and the continental surface in climate and weather forecast models. In PILPS Phase 4(b), selected schemes are coupled to the Limited Area Prediction System (LAPS) developed by the Australian Bureau of Meteorology. To facilitate the comparison of PILPS schemes' behavior within LAPS, a single mode of coupling is selected: explicit coupling. This type of coupling is more flexible and avoids most of the problems raised when interchanging the surface schemes. Exploratory tests are conducted. Initially, experiments are run in which the land-surface schemes use the same parameters as in their original host models. Then, in other runs, the most important surface parameters are set constant in an attempt to reduce the scatter amongst the schemes' results. In order to understand the impact of initialisation of soil moisture on the schemes' results some extreme cases (wet and dry) are performed. The partitioning between surface fluxes is studied as well as the soil moisture budget. Both regional and local results are analysed. Sensitivity between LSS is found in the precipitation field with rainfall over the Australian continent altering by about 20%, but no significant change is found in the net radiation. The scatter in the surface energy fluxes amongst the schemes is large (up to 300 W m−2 locally, during the daytime peak) but is seldom affected by the choice of surface parameters. The dynamical range of flux partitioning between extremely dry and wet initialisation varies strongly amongst the schemes. Some major shortcoming with the BUCKET approach are seen in the re-evaporation of convective precipitation over dry land, in the very large evaporation from wet surfaces and the diurnal cycle of surface temperature. |
