Revisions during 1988

5 January 1988 Forecast model cycle 30. A revised vertical diffusion scheme was implemented. The turbulent diffusion is now limited to below the top of the boundary layer except when static instability is generated. This modification restricts the vertical mixing to the boundary layer. The reduction of dissipation and momentum and heat mixing in the free atmosphere has a positive impact on zonal mean temperatures and reduces the zonal wind errors. The eddy activity becomes stronger.

Modest modifications in the parametrisation of the surface processes were included. The revision of the numerical scheme affects the partitioning of the surface moisture flux in terms of water extraction from the various contributing reservoirs. The interaction between convective precipitation and surface hydrology was revised as well as the interaction between the radiation and both the canopy layer and the snow.

The new surface parameters are only marginal influenced by these changes except in the case when snow is melting. Now the surface temperatures are allowed to be positive even with snow on the ground.

26 January 1988 Divergent structure functions were included in wind correlations of the analysis. The divergent structure functions improve the analysis significantly especially in the Tropics but the improvements were found short-lived during the assimilation cycle.

1 March 1988 The revision of the MARS interpolation software affects especially the surface orographic field of the ECMWF/TOGA Level III Basic Data Set.

12 July 1988 To minimise the impact of bad data in the data assimilation system the quality control algorithm have been modified which includes a more efficient OI check of SATEMs in areas with sufficient non-SATEM data and a general tightening of first-guess and OI rejection limits,

The structure functions were modified, resulting in an increased effective horizontal and vertical analysis resolution.

22 November 1988 Forecast model cycle 31. A modification of the surface scheme was implemented in order to correct some of the deficiencies of 2m temperature forecast.

1. The root profile was adjusted. The values of the root percentage in each of the 2 soil layers are now 50% (15%) surface layer, 50% (70%) intermediate layer and a 0% (15%) in the climate layer (percentage values within the brackets are valid for the old scheme). In the absence of precipitation no root extraction is allowed from the climate reservoir.

2. The background vegetation cover in dry situations was changed. No plant transpiration is allowed if the soil wetness in the root zone is lower than a threshold value. The background vegetation cover is riot decreased linearly to 0 when the root soil wetness decreases to 0.

14 December 1988 A change was made to the analysis, to prevent uncontrolled growth of spurious vortices at the top level of the model.