Numerical experiments of tropical deforestation with general circulation models (GCMs) and Regional Climate Models (RCMs) show the importance of land-cover teleconnections in the climate system but ignore multiscale feedbacks of regional-global and Eco-Hydro-Climatological (EHC) processes. In parallel to these scientific investigations, the evaluation and improvement of the Ocean-Land-Atmosphere Model (OLAM) has continued. Unlike typical GCMs, OLAM is uniquely designed to simulate multiscale interactions with its unstructured grid that can change resolution in any part of the global domain, as well as its fully interactive vegetation dynamics parameterization. Both features are essential for understanding the intricate processes that govern land-atmosphere interactions (including vegetation dynamics, clouds, precipitation and radiative processes) in a changing climate.
This project will understand and quantify the global EHC changes under a changing climate driven by increasing atmospheric carbon dioxide concentration and deforestation of tropical regions. Specifically, the investigators use OLAM to address two main issues: (i) What are the impacts of evolving deforestation in the Amazon basin on the multiscale EHC processes and feedbacks (including fires and other ecosystem disturbances), can we explain them and quantify them? (ii) Are these impacts affected by climate change (triggered by increasing atmospheric concentrations of carbon dioxide)? In the numerical experiments to be conducted for this investigation, deforestation patterns will be derived from socio-economical scenarios of the Amazon basin development, and the carbon dioxide scenarios adopted by the Intergovernmental Panel on Climate Change (IPCC)will be used.
The work will contribute to the evaluation and improvement of OLAM, which will help advance future climate research and operational weather prediction. It will be disseminated though seminars, lectures, teaching, and standard forums of publication. OLAM fulfills the criteria of a "community" model as it is freely available and supported. A post-doc will be trained and supported.
