In many climate model simulations, surface temperatures increase, whereas precipitation and évapotranspiration decrease in the areas affected by deforestation (Zhang et al., 1996; Lean et al., 1996; Henderson-Sellers et al., 1993). Evapotranspiration diminishes because of reduced surface roughness and an increase in albedo, as rainforest is altered to pasture. These changes contribute to the calculated decrease in rainfall following deforestation (Lean et ai, 1996). In Amazonia, model results and ground observations suggest that nearly two-thirds of the precipitation is recycled via rainforest évapotranspiration and soil evaporation, the balance going into surface and groundwater runoff (Zhang et al., 1996; Salati and Nobre, 1991).
Generally, the reduction in calculated annual precipitation exceeds the reduction in évapotranspiration, leading to decreases in runoff after forest clearance (see summary reviews in Henderson-Sellers et al., 1993; Lean et al., 1996). (One study found substantial increases in runoff, resulting from larger decreases in évapotranspiration than in precipitation; Lean et al., 1996.) These model results, however, generally run counter to field observations, which record increases in runoff (e.g., Bruijnzeel, 1996,1993). Changes in soil structure, such as loss of topsoil organic matter, compaction due to overgrazing and mechanical disturbances, and soil erosion, would combine to sharply reduce soil infiltration rates, thus increasing surface runoff, especially during wet seasons, and decreasing soil moisture in the dry seasons.
"Tracer" techniques employ climate models and variations in trace isotope concentrations to track sources and movement of atmospheric moisture (Salati et al., 1979; Koster et al., 1986; Druyan and Koster, 1989). These techniques, in conjuction with improved hydrologie models, could eventually help address the specific questions posed here—to what extent do large-scale anthropogenic transformations of the hydrologie cycle, such as impoundment of water in large reservoirs, irrigation, or deforestation alter the net balance between evaporation, precipitation, and/or advection of moisture to adjoining regions? Changes in this balance will determine (1) whether dam-building or irrigation would increase atmospheric moisture storage over land (especially in arid regions), which would reduce water transport to the ocean, and thus limit sea level rise, or (2) whether deforestation would decrease local evaporation/ évapotranspiration and precipitation, increase runoff, and hence augment rates of sea level rise. More advanced surface and groundwater hydrology models embedded within GCMs may be able to analyze additional feedbacks, such as the partitioning of water from groundwater mining among runoff, infiltration, evaporation, and the amount of water infiltrated beneath reservoirs and irrigated fields that contributes to underground flow.
Was this article helpful?