What to Expect
Non-stationary climate dynamics pose challenges to engineers and hydrologists in the design of infrastructure, the development of economic tools to hedge against changing environmental risks, and restoration of ecosystems. The forecasts necessary to address future water resources problems require an understanding of thermodynamic intensification of precipitation, structural atmospheric changes, shifts in plant compositions and phenologies, and knowing whether our current hydrologic models are flexible enough to capture these processes. In this talk, Dr. Knighton will first discuss a framework for quantifying the uncertainty in General Circulation Model derived extreme precipitation from the perspective of causal atmospheric mechanisms, and then introduce the possibility of utilizing emergent convolutional neural networks to circumvent these challenges. Next, he will discuss the importance of representing functional plant traits in distributed models with implications for flooding increases
following forest cover change. Hydraulic regulation of transpiration by plant root water uptake (RWU) will be examined with stable water isotopic measurements collected throughout a forested catchment in the Northeast US. He will share preliminary insights into how forest rooting structures can be derived from widely available discharge data with inverse ecohydrological modeling techniques.
James Knighton is a hydrologist and registered Professional Engineer (P.E.) with a Ph.D. in Environmental Engineering from Cornell University and an M.A. in Environmental Studies from the University of Pennsylvania. James’ research centers on applying data analytic techniques in the context of mechanistic models to study relevant ecohydrologic problems. His current research focus includes: 1) utilizing stochastic approaches to develop hydrologic forecasts of atmospheric and discharge extremes change constrained by physically-based models, 2) catchment modeling informed by long-term ecohydrologic datasets to make inferences on rooting zone processes, and 3) evaluating the critical issue of model-climate transferability for physical and biological processes as well as statistical models used to describe uncertainty. Prior to completing his Ph.D., James worked as a P.E. for eight years in the nuclear industry and government performing risk analysis of power generation facilities and urban environments.