In its initial phase, this LBA research program did not address the need for projections of future land cover for the entire basin. Such projections are necessary in order to evaluate the consequences of new policies, or possible changes in current policies. Clearly, the main objective of LBA efforts has been to settle questions about the region’s natural systems, and in particular about interactions between the atmosphere and the vegetative surface. Nevertheless, the ultimate policy ramifications of these LBA findings will be greatly enhanced if they can feed into evaluation of policies that will to a great extent determine the future of that vegetation. Thus, for insight into the environmental impacts of future development in the Amazon region, it is necessary to understand the manner in which both land use and land cover are likely to change at the basin scale.

To date, LBA research has taken two approaches to predicting land cover change, neither adequate to the task of basin projection that we propose to undertake in this work. The first approach, human-behavioral research, has been of high quality but micro scale. It has largely addressed household dynamics, with a special concern for links between demography and land allocation at the level of the farm household. While this work has considerably advanced our understanding of local land-cover dynamics, it has been based on both spatially and temporally restricted samples. One consequence is that the work to date does not easily aggregate to basin-scale, and does not fully account for broad-scale regional impacts on land-cover change, or for those associated with the macro-economy.

The second approach has addressed aggregation by taking the basin in its entirety as the scale. Through the use of simple statistical algorithms, which are implemented with spatial specificity within geographic information systems (GIS), this approach has been able to make regional predictions in a consistent manner across the basin. However, the algorithms implemented are based mainly on assumptions, with little in the way of theoretical foundation or detailed understanding of local or regional land-use dynamics. This undermines their utility in making projections about the impacts of infrastructural changes such as those that might be brought about through large-scale policies along the lines of Avança Brasil.

We propose, in essence, to state a projection model combining elements of these two LBA approaches to land-cover-change prediction, in order to generate basin-wide predictive capability with spatial specificity, and grounded in both economic theory and knowledge of local/regional land-cover dynamics. The theory comes from concepts of the household economy, as well as from models of local and regional interactions of agents. The latter is necessary given the scale at which we plan to state and implement our model. Specifically, we propose to construct a regression model that uses polygon observations for the entire basin, and for multiple years. This model will be capable of generating predictions of expected land-cover changes under any number of possible future scenarios and policies. Further, because its structure will be inferred from actual, observed land-cover changes over both time and space, we will be able to provide an assessment of the errors in model projections.

Consequently, the proposed research addresses the LBA-ECO theme, “land cover and land use change.” By focusing on the construction of an econometric statistical model of deforestation for the entire Amazon basin, it seeks to answer the second-tier NASA research question (F2), “What changes are occurring in the global land cover and land use, and what are their causes?” In answering this question, the research will also provide projections of future land use, with error bounds, thus contributing to our understanding of future land use and vegetative cover scenarios for the Amazon basin. This could prove of value to research addressing second-tier question (P5) regarding the “reliability of models predicting future atmospheric concentrations of carbon dioxide and methane.”