Cenários de emissões de gases de efeito estufa no Estado de Roraima, Brasil (2000 a 2050)

Abstract

The increased of greenhouse gas (GHG) emission is causing global climate change. Mitigation and adaptation of climate effects require accurate GHG emissions estimates in a local application scale. Uncertainties related to the current estimates of emissions from land use change are mainly due to the exclusion of understory forest fires emissions, the spatial scale used in the measurements of deforested area and the carbon storage in forest biomass. This study has a proposal improve estimates of GHG emissions to the atmosphere in Roraima, northern Brazilian Amazon, increasing the spatial resolution for the detection of deforestation and forest fires, as well as adjustments in the calculation of biomass/carbon stocks in forest ecosystems. All of these variables were integrated into scenarios of GHG emissions assuming values observed in the period 2000 to 2010 to simulate four scenarios over the period of 2011 to 2050 (predicted values). The geographical area that was focus of this study in Roraima was inclusionary for all forest systems without any environmental protection, and exclusionary to non-forest systems (savannas and oligothrophic grasslands), water courses, and protected areas (conservation units - CU and indigenous lands - IL). All deforestation and understory forest fires observed in the focal area were year-to-year mapped (resolution ≥ 1 ha) using as reference Landsat TM / ETM images between the years 2000-2010. Occurrences of deforestation and forest fires were distinguished by phytoclimatic zones, defined here as regions containing forest groups established in different climatic types occurring in Roraima. Statistical tests were performed to the evaluation of the effects of phytoclimatic zones in the magnitude, annual frequency of deforested and forest fires areas, and biomass/carbon stocks. Total deforestation area was not distinguished by phytoclimatic zones, but forest fires were significantly higher in the zone characterized by the presence of seasonal forests and ecotones distributed in areas of higher water deficit. In the modeling the spatial distribution of the stock of original forest biomass (live + dead, above + below ground) was used a base of forest inventories of timber volume available for the state. The biomass was quantified at the level of forest types (forest groups) and land use, and spatialized using geostatistical techniques. Biomass (Mg ha-1) was different among forest groups and land use, where IL and CU may play an important role in mitigating the effects of global warming due to the large biomass stock within its boundaries. Estimates of GHG emissions (CO2, CH4 and N2O) to the atmosphere were performed using the approach of the annual balance. In the initial period (2000-2010) was observed that the highest emissions occurred in El Niño years (droughts), triggering the occurrence of large forest fires. In the emissions estimated for the following period (2011-2050), the scenario of governance (MT-GOV) simulated the control of deforestation committed by Brazil in 2009 in Copenhagen (reducing deforestation rates by 80% from baseline 2005). GHG emissions were also reduced in the same proportion. The "business as usual- intermediate" scenario (MT-BAU-i or Business As Usual-i), simulated the increase in deforestation and GHG emissions due to increased migration to the state because of the reconstruction and paving of BR-319 (Porto Velho-RO to Manaus-AM). On the other hand, the scenarios RD-GOV - governance and RD-BAU - business as usual, simulated the continued growth of deforestation and GHG emissions due to the growth of farming on forest areas. The RD-GOV scenario used only 50% of the growth rates of livestock and agriculture used by RD-BAU. The RD-BAU was the largest GHG emissions scenary when compared to all other, simulating uncontrolled growth of farming. Taking into account the results of 2000-2010 interval and the results of emission scenarios from 2011 to 2050, it is concluded that the higher GHG emissions are related to phytoclimatic zone characterized by water deficit and presence of seasonal forests and ecotones. This area is closest to Boa Vista (capital of Roraima) and concentrates the majority of the state population. The scenarios also indicated that phytoclimatic zone with lower water deficit, dominated by Ombrophilous forest groups, will depend on the decisions taken by the state government for implementation of public policies of containment and control of deforestation, with the intention of rearrangement of agricultural activity in the state. The rearrangement of the state's agricultural policy is directly linked to the impact of the reopening of the BR-319 to the south of Roraima, which should cause a substantial increase in population and consequent increase in demand for land. The refinement of the work scale for detection of deforested areas, including forest fires and adjustments in the biomass/carbon stocks calculation were important to improve GHG emissions estimates for the four scenarios proposed in this study. The results presented here provide a better basis for making decisions that should be made by local and federal government.