Understanding the Water Resources of Amazonia

Michael Coe, Associate Scientist with The Woods Hole Research Center, uses numeric models to study large-scale water resource variability.

At the Woods Hole Research Center, we are developing a unique combination of computer models, fieldwork, and satellite data to better understand the natural and human induced variations in the current and future water resources of the Amazon River basin. The research will provide insight into river functioning and services, as well as provide a tool for policy-makers to make more informed development decisions for the region.

The 6.7 million km² Amazon, one of the most important watersheds on the planet, includes a massive network of rivers, floodplains, streams and wetlands, all playing an important role in modulating the Earth’s hydrologic and biogeochemical cycles. With nearly 20 percent of the Earth’s freshwater discharge, the Amazon carries more water than the nine other largest rivers of the world combined.

Population and development pressures in the last several decades have led to large areas of deforestation in the Amazon, mostly in the eastern and southern portion of the basin. By 2000, about 15 percent of the total land area of the river basin had been deforested (Figure 1a) due to these pressures. Recent rates of deforestation continue to be very high: in 2002 and 2003, the amount of land deforested was the largest ever in a two-year period, at nearly 46,000 km², (according to the PRODES Project of INPE, the Brazilian space agency), an area equal to more than twice the size of Massachusetts. Accelerated world demand for free-range beef and grains in light of mad cow disease and other outbreaks suggest that demand for Brazilian beef will increase, perpetuating high rates of Amazonian deforestation well into the future. Recent work by Woods Hole Research Center affiliated scientist Britaldo Soares, at the Federal University of Minas Gerais, Belo Horizonte, Brazil, suggests that between 30-50 percent of the entire Amazon Basin will be deforested by 2050, depending on what policies are followed [Soares-Filho et al., 2004] (Figure 1b, c). This work provides a graphic and quantitative measure of how government policies may directly affect the forests in the future.

Figure 1. Examples of future deforestation scenarios for Amazon River basin [Soares-Filho et al., 2004]. Forested areas are shown in green, and deforested areas in orange and gray. (a) Actual vegetation in the year 2000 (b) Scenario for the year 2050 under strict governance of deforestation, and (c) Scenario for the year 2050 under "business as usual" deforestation. Refresh page to review animation. Animation: Lefebvre/Ernst, WHRC.

Human land cover and land use changes, including conversion of natural vegetation to pasture and croplands, impact the quantity and quality of surface water resources by: 1) changing how rainfall and sunlight are converted into surface energy, runoff, and river discharge; and 2) changing the properties of the land surface itself and what materials leach from soils. These types of activities have already greatly altered the hydrology and chemistry of many watersheds and coastal marine environments across the globe.

These land cover changes may have a very large impact on the flow of the Amazon River and the goods and services it provides. For example, conversion of vegetation from natural types to crops and pasture causes large changes in the amount and timing of river discharge and flooding. Further, many trees species fruit during the flood season and aquatic fauna have adapted to take advantage of flooded forest resources for food and breeding habitat. These may be highly sensitive to changes in magnitude and timing of discharge and/or in water quality created by planned hydroelectric projects. Therefore, an understanding of how different land use policies, climate, and infrastructure investments impact the River will be critical to planning the sustainable development of water resources while conserving the unique ecology throughout the basin.

One tool we are currently applying to these problems is a comprehensive computer model of the River system. This Terrestrial Hydrology Model with Biogeochemistry (THMB) simulates the flow of water and transport of nutrients from the land surface to the ocean through a complex network of dynamically simulated rivers, floodplains, and lakes. We have already used this model to gain unprecedented insights into how El Niño and longer-term climate variability have affected the River system in the last 70 years.

The next phase of our research is to apply this with other tools to understanding how the River may be affected in the future. We have recently completed preliminary simulations of the impact of potential future deforestation on the discharge and flooded area of the River (Fig. 2a,b).

Figure 2a,b. Preliminary results from a simple version of our model system, which illustrate the potential impact of deforestation in the year 2050 [Soares-Filho et al., 2004] on the THMB simulated mean monthly discharge of a 450,000 km2 portion of the Xingu River (2a) and the total flooded area of the southern tributaries of the Amazon (Purus, Madeira, Tapajós, and Xingú). In 2b, months are shown on the x-axis from September (9) to August (8).

These results indicate that by the year 2050 the discharge of many of the large southern tributaries may be increased by 5-25 percent during the wet season, depending on the land use policies that are put into place. While, the seasonal flooded area may increase by 15-30 percent as a result of deforestation alone. These changes are large enough to have a big impact on river ecology and human structures.

This project will synthesize research results of groups located at the University of Wisconsin, The Woods Hole Research Center, and the Federal University of Viçosa, Brazil in topics that include socioeconomics, land cover/land use change, land surface physics, and hydrology. We will link results from basin-wide models of deforestation dynamics, land surface ecosystems and river hydrology and chemistry with satellite and ground based observations of stream flow, floodplain inundation, and stream chemistry. This combination of models, data, and expertise will enable us to assess how future scenarios of land cover and land use change are likely to influence local hydrology and biogeochemistry of streams and how these local changes will be transported downstream to affect the entire Amazon River system. The value of a model based assessment is that: 1) it can provide a clearer spatial-temporal analysis of potential future changes that are suggested by the observations; how large a change to the stream system could occur and at what proximity to the land cover/use changes, and 2) it can serve as a tool for policy makers; providing both government managers and civil society with a modeling framework for analyzing the costs and benefits of different land-use policy decisions. Land use and water resources policies will be analyzed in light of the assumptions and results of the model to provide recommendations to policy makers and state and federal government resource managers at IBAMA (Instituto Brasileiro de Meio Ambiente e Recursos Renováveis) and ANA (Agência Nacional das Águas) for reconciling land use development with conservation of water resources and aquatic habitat and biodiversity.

©Woods Hole Research Center, 2007