Investigate Ecosystem Resiliency and Vulnerability under Extreme Event using Multi-Land Surface Models

Extreme events such as droughts and heat waves have serious and damaging impacts on terrestrial processes. Mostly, these events negatively influence ecosystem services (e.g. food, fiber, and freshwater), human infrastructure, as well as atmospheric chemistry through the release of enormous pulses of carbon dioxide and other greenhouse gases from terrestrial ecosystems. Under climate change, these extreme weather events are likely to shift in both magnitude and frequency at regional and local scales. These potential shifts in extreme events and their consequences, as well as stresses from humans, place an ever-greater demand toward understanding the ecological response of natural as well as managed ecosystem under extreme changes.

This project explores the resiliency and vulnerability of ecosystem to droughts at different time scales and magnitudes using land surface models with diverse sets of bio-geophysical parameters and representations. Four land surface models: two process-based Terrestrial Ecosystem Model (TEM monthly and TEM-Hydro Daily), the multi-layered eddy-covariance based Advanced Canopy-Atmosphere-Soil Algorithm (ACASA), and the Community Land Model (CLM) are used to simulate carbon dioxide and water fluxes between the biosphere and the atmosphere for numerous AmeriFlux stations under normal and extreme drought events. In addition to evaluating model performance with in-situ eddy covariance measurements, this multi-model comparison also examines the strength and weakness of the different model parameterizations and representations. For example, while evapotranspiration simulated by the TEM monthly model shows good agreement with observations over the Duke forest, other land surface models with shorter model timescale perform better in simulating ecological responses during the 2002 summer drought.