Understanding tropical vegetation-climate interactions in the Amazon basin

Our current Amazon work — supported by NSF Partnerships for International Research and Education (Amazon-PIRE) and building on the strong legacy of support from NASA — focuses on integrating remote sensing techniques and ground based measurements to extrapolate our understanding of local controls on carbon cycling in old-growth Amazon forest to derive landscape and regional scale carbon balance. The goal of this work is to build upon our ongoing investigations of how forest demography and disturbance dynamics control carbon cycling in old-growth Amazon forest, which uses long-term eddy covariance observations of net ecosystem exchange of CO2, integrated with classical methods of forest ecology.

Our ground-based site is the Tapajós National Forest near Santarém (see figure, above right, of long-term eddy flux tower for measuring net ecosystem exchange of CO2 in this central eastern Amazon forest of Brazil). The net CO2 flux from this tower shows the forest losing carbon, an observation not previously seen in the Amazon that will help reconcile an Amazon carbon-budget problem (See Saleska et al., 2003). The carbon budget problem arose from previous studies showing carbon uptake of 1-3 Pg C yr-1 in Amazônia alone, comparable to the whole global terrestrial carbon sink.

We collaborated with Chris Martens at UNC to test of the accuracy of our eddy flux measurements using radon gas as a transport tracer (Martens et al., 2004). This work independently confirms our approach to estimating carbon balance using eddy flux measurements.