Simple Biosphere Model, Version 2.5 (SiB2.5)

The land-surface is an important source and sink of radiation, sensible heat, water, momentum, and trace gases to the atmosphere. A parameterized representation of these exchanges is an important component in all climate and weather models (Betts et al., 1996). Land-surface parameterizations have undergone tremendous development in the past decade, and now include a much greater degree of biophysical realism and self-consistency than was being used just a few years ago (Sellers et al., 1997). Biophysically-based land-surface parameterization has been shown to lead to better ability of numerical weather prediction (NWP) models to forecast weather under extreme climatic conditions such as droughts (Atlas et al., 1993) and floods (Beljaars et al., 1996). Land-surface parameterization is an important part of the numerical weather analysis and forecasting infrastructure at the European Center for Medium-Range Weather Forecasting (Viterbo et al., 1995). Land-surface predictions can also be used to evaluate the accuracy of and develop better parameterizations for assimilation of weather and climate data (Betts et al., 1998).

A new generation of land-surface parameterizations has emerged in recent years in which exchanges of water and heat at the vegetated land surface are linked to exchanges of CO2 (Sellers et al., 1997). This linkage is based on the fact that physiological control of evapotranspiration by plants is an evolved optimization mechanism that seeks to maximize carbon fixed by photosynthesis (by drawing CO2 into leaves through stomatal pores) and yet reduce water loss from the plant (by closing stomata). The representation of this linkage in land-surface parameterizations has been shown to improve the simulated diurnal cycle of temperature and humidity. It also allows key parameters controlling surface exchanges to be related to the spectral reflectance characteristics of vegetation (Sellers et al., 1996a, 1996b). This carbon-water linkage also opens the door for the models to predict the flux of CO2 in a self-consistent way with simulated surface energy exchanges and turbulent and convective transport in the atmosphere. By coupling photosynthesis and energy flux calculations with an atmospheric transport model, Denning et al. (1996a, 1996b) were able to achieve a greater degree of realism in simulated diurnal and seasonal variations of CO2 than had previously been possible.

The Simple Biosphere (SiB) Model, originally developed by Sellers et al. (1986), was substantially modified (Sellers et al., 1996a) and is referred to as SiB2.5. The number of biome-specific parameters was reduced, and most are now derived directly from processed satellite data (Sellers et al., 1996b,c) rather than prescribed from the literature. Another major change is in the parameterization of stomatal and canopy conductance used in the calculation of the surface energy budget over land. This parameterization involves the direct calculation of the rate of carbon assimilation by photosynthesis (Farquhar et al., 1980), making possible the calculation of CO2 exchange between the global atmosphere and the terrestrial biota on a time step of several minutes (Denning et al., 1996a,b; Zhang et al., 1996). Photosynthetic carbon assimilation is linked to stomatal conductance and thence to the surface energy budget and atmospheric climate by the Ball-Berry equation (Collatz et al., 1991, 1992), which is the basis for the ability of the model to calculate the climatic effects of physiological responses to elevated CO2 (Sellers et al., 1996c). Soil respiration is calculated from the temperature and moisture of each layer of soil, and is scaled to achieve carbon balance over an annual cycle (Denning et al., 1996a). Recent improvements include the introduction of a 6-layer soil temperature submodel based on the work of Bonan (1996, 1998), and a revised surface energy budget that includes prognostic temperature and moisture in the canopy air space reservoir.