Close Window

Deforestation and Climate of Amazon: simulations using BRAMS coupled to GEMTM

Josivan da Cruz Beltrão, Universidade Federal do Pará, josivan@ufpa.br (Presenting)
Júlia Paiva Cohen, Universidade Federal do Pará, jcpcohen@ufpa.br
Adilson Wagner Gandú, Universidade do Estado de São Paulo, adwgandu@model.iag.usp.br

In order to understand the climate changes caused by deforestation that took over the Amazon in the last decades, we used the Brazilian Regional Atmospheric Modeling System (BRAMS) coupled to a dynamic vegetation model "Energy and Mass Transport Model" (GEMTM) [Chen & Coughenour, 1994]. Two land-cover scenarios were used as surface boundary condition based on the DINAMICA model [Soares-Filho et al, 2005]. The land-cover scenario for the year 2002 was used for a control simulation, while the scenario for the year 2050 was used to estimate the impact of the land-cover change on the climate. Preliminary analysis show that in the control run the accumulated precipitation was over-estimated as compared to observations, however, the spatial distribution was well simulated as compared to observed maps. The replacement of forest by pasture shows a decrease of rainfall in some areas and an increase over other areas. At the state of Pará there is reduction of about 25 % during the rainy period and about 15% for the dry period. However, in the state of Amazonas, there is a stronger decrease and over larger areas for the dry period. There is also an increase in the temperature over most of the region. However, during the dry season the higher increase of temperature takes place over the most deforested areas at the east of the state of the Amazonas and the state of Acre. The major conclusion is that deforestation produces a drier and hotter Amazon, but the results show a lighter effect as compared with the Global Circulation Models (GCMs). This is probably due to the use of the dynamical vegetation model that produces structural changes on its characteristics controlled by the climate conditions, and by the use of more realistic land-cover scenarios.

Science Theme:  HY (Hydrometeorology)

Presentation Type:  Poster

Abstract ID: 61

Close Window