The basic elements of land-atmosphere interaction (LA-I) are the exchanges of moisture and energy between these two systems. Historically, many of the important aspects of this interaction have been treated in the areas related to micrometeorology, agriculture and forest meteorology, planetary boundary layer, and hydrology. More recently, LA-I has also become recognized as important for studies of biogeochemical cycling, climate, mesoscale meteorology, and numerical weather prediction. Initial recognition of the importance of LA-I in these latter areas extends back at least two decades, but major advances have occurred over the period covered by this review. Land has been recognized as, in principle, a major element of the climate system since the initiation of the World Climate Research Programme (WCRP), and land was included in simple form in even the earliest General Circulation Model (GCM) climate studies.
The present review emphasizes advances in understanding of land interactions as part of the
climate system, but also touches on related questions. For a contemporary view of Climate
System Modeling, cf. Trenberth [1992]. Also, Prinn [1994]
reviews atmospheric chemistry as part of the climate system including land interactions, and
Mooney and Koch [1994] address the issue of effects of rising concentrations of CO
on
terrestrial vegetation.
Fluxes of moisture and heat from the land surface determine the overlying distributions of atmospheric temperature, water vapor, precipitation, cloud properties, and hence the downward radiative fluxes at the surface. For example, Koster and Suarez [1994] find in decadal GCM simulations that inclusion of an interactive land surface leads to substantially greater interannual variability of precipitation over both tropical and mid-latitude land than does observed ocean temperature variations. How the coupling between land and atmosphere depends on the formulation of the land and atmospheric processes as represented in models is now being explored. Some of the general aspects of this question are common to study of ocean-atmosphere interactions.
What especially distinguishes the land question is the differences between land and ocean surfaces. A much wider range of surface moisture and temperature conditions is realized over land in proceeding from arid to moist climatic zones and from tropical to polar climates. The relatively low heat capacity of the land surface and its limited capacity for water storage imply much stronger diurnal variations in surface conditions over land than over the ocean, and more direct responses to changing atmospheric inputs of energy and moisture as cloud properties and precipitation change. First GCM climate models were averaged over the diurnal cycle, but now most include a diurnally varying sun [e.g., Randall et al., 1991].
The limited capacities for heat and water storage over land, combined with the heterogeneous nature of underlying soils, vegetation, and slope, imply potentially large heterogeneities in sensible and latent fluxes. Whether these fluxes, on the average, differ drastically from fluxes from an assumed homogeneous surface, what mesoscale circulations they might drive, and how these might affect large-scale conditions are crucial questions now being explored.