| Numerous experiments were conducted with a regional climate model (MM5) to determine its ability to simulate the climate of tropical South America and to investigate the modeled climate's response to deforestation. Six month-long test simulations were performed for the wet (January) and dry seasons (July). MM5 was able to simulate the differences in regional circulation between the two seasons, and also reproduced the regional climate features of each month quite well. An evaluation of MM5's land surface model (OSULSM) showed that OSULSM does not partition energy at the surface correctly. The modeled sensible heat flux was far too high in comparison with observations. In order to mitigate this effect, soil layer depth, rooting depth, leaf area index, and initial soil moisture were increased. This configuration compared more favorably with observed data. Three experiments with 100% (DEF), 50% (PD1), and 20% (PD2) deforestation were performed for January and July. The partial deforestation experiments were designed to present a realistic picture of deforestation. The experiments showed that the model climate was more sensitive to land cover changes during the dry season. In addition, the response of the model to the imposed land cover changes was of the same or lesser magnitude than the difference between the wet and dry months for each season. In the January experiments, the area-averaged precipitation response was small. The July simulations showed a stronger response to the land cover changes, as precipitation decreased in all three experiments. The reduction in precipitation for each experiment was not proportional to the amount of deforestation: the decrease in precipitation for PD1 and PD2 was less than amount of deforestation relative to DEF, indicating that the amount and distribution of deforestation does have an impact on the climate signal. These simulations are sensitivity studies, and should not be interpreted as a forecast of the climatic effects of deforestation. In addition, the negative impacts of deforestation on tropical ecosystems are not considered here. Future research should focus on the dry and transition seasons, as the land surface appears to have a greater influence on regional climate during those time periods. |
