The University of Arizona Brazilian Partners Harvard University The University of Arizona

Research

This Partnership for International Research and Education (PIRE) addresses a fundamental earth system science question that cannot be answered without a strong intellectual collaboration between scientists in the U.S. and in South America: what is the future of Amazon forests under climate change? Models that simulate the coupled interaction of climate with carbon and water cycles, mediated by vegetation, suggest that these forests will collapse due to global warming-induced drying. But other models predict resilience. Current understanding is insufficient to evaluate which models are likely to be correct.

We will decisively address this problem by testing hypotheses about the response of Amazon forests to drought on observable timescales that include climate shifts from the El Nino and Tropical Atlantic cycles. Our integrated research program uses 3 approaches:

  1. Long-term observations of (a) ecosystem fluxes of CO2, H2O and energy on a network of sites, (b) vegetation dynamics and ecophysiology at a core site, and (c) regional to continental scale forest-atmosphere processes, by remote sensing and by aircraft and tower sampling campaigns of atmospheric CO2, trace gases, and biogenic aerosols. Extensive existing and new datasets will allow PIRE students to make observational tests only now possible due to ground-based and satellite infrastructure put in place since the last large ENSO-related drought in 1997/98.

  2. Long-term manipulation of precipitation and CO2 inside a precisely controllable 0.2 ha experimental tropical forest mesocosm, part of the University of Arizona's new program at the Biosphere 2, in order to test mechanisms not readily observable at ambient CO2 levels.

  3. Multiscale model simulations linking carbon and water cycles with evolution of the forest vegetation assemblage, using the Ecosystem Demography (ED) model, general circulation models (GCMs), and high-resolution mesoscale and global models (Brazil-RAMS, OLAM).
Download PDF