| Home > Newsevents > Training > Rcourse_notes > PARAMETRIZATION > SURFACE_ASSIMILATION > |
Land surface assimilation
March 2001
By Jean-François Mahfouf and Pedro Viterbo
European Centre for Medium-Range Weather
Forecasts
Table of contents
Next Section
1 Introduction
The importance of land surface processes has been recognised for a long time by the climate modelling community, in order to describe in a consistent way all the components of the water and energy cycle over long periods of time. As a consequence a variety of schemes have been devised ranging from the simple bucket model of Manabe (1969) to the complex soil-vegetation SiB (Simple Biosphere) model of Sellers et al. (1986).
Although a number of sensitivity studies have shown that land surface processes can also affect medium range weather forecasts (Rind 1982; Rowntree and Bolton 1983; Yeh et al. 1984; Rowell and Blondin 1990; Beljaars et al. 1996), the parametrization of the interactions between continental surfaces and the lower atmosphere is still rather crude in most numerical weather prediction models.
Recent field experiments, such as HAPEX-MOBILHY 86, FIFE 87 or BOREAS 94, have enabled the development and the validation of land surface schemes describing the most important processes governing the water, heat and momentum exchanges while remaining simple enough to be included in operational weather forecasts models (Noilhan and Planton 1989; Viterbo and Beljaars 1995). Such parametrizations can improve significantly the quality of the forecasts of weather elements (Bougeault et al. 1991; Lanzinger 1995). Initialisation of the prognostic soil variables at global scale is an important issue given the very different time scales of evolution between the atmospheric and the soil systems. Realistic schemes appear to be sensitive to the specification of initial soil temperatures and water contents (Jacquemin and Noilhan 1990; Bouttier et al. 1993a, 1993b). The increase in realism of other physical parametrizations (clouds, radiation, convection) in numerical weather prediction models has also made errors in the surface representation easier to identify. The importance of positive feedbacks between the surface and the atmosphere must be underlined : realistic mechanisms should be represented (which is not possible when surface boundary conditions are fixed), but spurious ones, resulting from systematic errors in the representation of some components of the energy and water cycles, should be removed with an appropriate initialisation procedure.
In the following, we will start in Section 2 by describing the basic features common to most surface schemes. Special features of the land surface assimilation are identified in Section 3, in terms of data availability and its relation to model variables. Section 4 reviews the problems of simple surface initialization procedures. Methods for the initialisation of soil water in numerical models using near-surface temperature and humidity observations are described in Section 5, while Section 6 explains the drawbacks of on-site surface observations of soil moisture and discusses the need for remote sensing. Finally, Section 7 reviews the initialisation of other slowly varying land surface variables, such as snow mass, deep soil temperatures and vegetation characteristics.
Training Course Notes Front Page
Table of contents
Next Section
 |
12.06.2002 |
 |
© ECMWF |
