Climatic Impact of Amazon Deforestation---A Mechanistic Model Study

Ning Zeng, Robert E. Dickinson and Xubin Zeng
Institute of Atmo spheric Physics
The University of Arizona
Tucson, Arizona 85721

Recent General Circulation Model (GCM) experiments suggest a drastic change in the regional climate, especially the hydrological cycle, after hypothesized Amazon basin-wide deforestation. To facilitate the theoretical understanding of such a change, we develop an intermediate-level model for tropical climatology, including atmosphere-land-ocean inter action. The model consists of linearized steady-state primitive equations with simplified thermodynamics. A simple hydrological cycle is also included. Special attention has been paid to land-surface processes. In comparison with previous simple modeling work on tropical climatology or anomalies, the present model is more sophisticated in predicting, with little input, most of the important meteorological variables; nevertheless, it is computationally simple. It generally better simulates tropical climatology and the ENSO anomaly than do many of the previous simple models.

The climatic impact of Amazon deforestation is studied in the context of this model. Model results show a much weakened Atlantic Walker/Hadley circulation as a result of the existence of a strong positive feedback loop in the atmospheric circulation system and the hydrological cycle. The regional climate is highly sensitive to albedo change and sensitive to evapotranspiration change. The pure dynamical effect of surface roughness length on convergence is small, but the surface flow anomaly displays intriguing features . Analysis of the thermodynamic equation reveals that the balance between convective heating, adiabatic cooling, and radiation largely determines the deforestation response. Studies of the consequences of hypothetical continuous deforestation suggest that the replacement of forest by desert may be able to sustain a dry climate. Scaling analysis mo tivated by our modeling efforts also helps to interpret the common results of many GCM simulations.

When a simple mixed-layer ocean model is coupled with the atmospheric model, the results suggest a 1$^\circ$C decrease in SST gradient across the equatorial Atlantic ocean in response to Amazon deforestation. The magnitude depends on the coupling strength.