学术报告通知：
报告题目：Uncertainty Analysis for Subsurface Environmental Modeling
报告人：叶明 副教授
报告人单位：美国佛罗里达州立大学科学计算系  
语言：中文
时间：2014年9月9日（星期二）15:00
地点：国家重点实验室学术报告厅（农水楼一楼）

报告简介：
   Uncertainty analysis, in a broader sense, includes sensitivity analysis, parameter estimation, and uncertainty quantification. These has been used widely in subsurface environmental modeling. The sensitivity analysis is used to determine the processes and/or parameters that are important to the physical, chemical, and biological phenomena to be simulated. The parameter estimation is to obtain the optimum parameters that yield the best fit to observations of system properties and states. The uncertainty quantification is to evaluate propagation of uncertainty in model parameters, structures, and driving forces to the model outputs. The results of uncertainty analysis can be used to inform science-based decision-making related to environmental protection and regulation. In comparison to decision-making based on a deterministic modeling, decision-making based on comprehensive decision-making can help reduce risk of making wrong or overly conserve decisions. This presentation will give an overview of popular methods of uncertainty analysis. The sensitivity analysis will include local methods that are derivative-based, global methods that are based on screening methods, and global methods that are based on variance decomposition. The parameter estimation methods will cover regression and Bayesian approaches. The uncertainty quantification will be focused on the Markov chain Monte Carlo methods combined with using of sparse-grid surrogate. In addition to the theoretical introduction, applications of the methods will presented, including their uses at the Yucca Mountain and Nevada National Security Site of the U.S. Department of Energy for unsaturated and saturated flow modeling, at the Naturita Site for uranium reactive transport modeling, and Florida for nitrogen transport modeling. The presentation will be closed with a perspective of using these methods in future subsurface environmental modeling.

叶明副教授简介:
   叶明，美国佛罗里达州立大学科学计算系并地球物理流体力学研究所副教授。其研究领域主要是地下水水流及反应溶质运移的数值模拟以及不确定性分析，滨海水动力数值模拟，和基于地理信息系统的地下水软件开发。叶明现为美国地球物理学会《Water Resources Research》杂志副主编、《Journal of Hydrology》杂志副主编。其科研多次获得美国国家自然科学基金，能源部，国防部，原子能管理委员，以及佛罗里达州环保厅的资助。2012 获得美国能源部 Early Career Award ，并选为美国地质学会理事 (Fellow of the American Geological Society)。


欢迎相关专业教师和研究生的光临！