Drought Tolerance and Implications for Vegetation-Climate Interactions in the Amazon Forest
December 19, 2011
Anna Harper
Committee: Scott Denning (advisor), Dave Randall, Chris Kummerow, Francesca Cotrufo (Soil and Crop Sciences)
Abstract
On seasonal and annual timescales, the Amazon forest is resistant to drought, but more severe droughts can have profound effects on ecosystem productivity and tree mortality. The majority of climate models predict decreased rainfall in tropical South America over this century. Until recently, land surface models have not included mechanisms of forest resistance to seasonal drought. In some coupled climate models, the inability of tropical forest to withstand warming and drying leads to replacement of forest by savanna by 2050. The main questions of this research are: What factors affect forest drought tolerance, and what are the implications of drought tolerance mechanisms for climate?We examine forest response to drought in an ecosystem model (SiB3 - the Simple Biosphere model) compared to two rainfall exclusion experiments in the Amazon. SiB3 best reproduces the observed drought response using realistic soil parameters and annual LAI, and by adjusting soil depth. SiB3’s optimal soil depth at each site serves as a proxy for forest drought resistance. Based on the results at the exclusion sites, we form the hypothesis that forests with periodic dry conditions are more adapted to drought.
We parameterize stress resistance as a function of precipitation climatology, soil texture, and percent forest cover. The parameterization impacts carbon and moisture fluxes during extreme drought events. The loss of productivity is of similar magnitude as plot-based measurements of biomass loss during the 2005 drought.
Changing stress resistance in SiB3 also affects surface evapotranspiration during dry periods, which has the potential to affect climate through changing sensible and latent heat fluxes. We examine the effects of forest stress resistance on climate through coupled experiments of SiB3 in a GCM. In a single column model, we find evidence for a more active hydrologic cycle due to increased stress resistance. The boundary layer responds through changes in its depth, relative humidity, and turbulent kinetic energy, and the changes feed back to influence wet season onset and intensity. In a full global GCM, increased stress resistance often decreases drought intensity through enhanced ET and changes to circulation. The circulation responds to changes in atmospheric latent heating and can affect precipitation in the South Atlantic Convergence Zone.