报告时间：2026年4月22日（周三）上午 09:30

报告地点：气象楼423会议室

主  持 人： 周洋 副教授

报告题目：Challenges in modelling and predicting interannual and longer-term changes of ocean variations in the Northwest Atlantic, with implication for ecosystem and fishery applications

报告专家：鹿有余 研究员

专家简介:

Dr. Youyu Lu, Research Scientist, Bedford Institute of Oceanography, Fisheries and Oceans Canada, is a physical oceanographer with over 30 years of research experience. In research, he combines statistical analyses of observational and numerical modelling data with dynamic analysis, to quantify and understand the forcing mechanism of the space-time variations of a variety of physical oceanographic processes (ocean circulation, sea levels, hydrography, sea-ice, mixing, air-sea interaction, etc.), and their influences on marine ecosystem and environment. His study areas include the deep oceans of the North Atlantic, Arctic, and the equatorial, northeast and northwest Pacific, shelf and coastal seas, and large lakes. He has published 110 refereed articles, supervised about 20 young scholars and co-supervised 10 graduate students at master and PhD levels, and has been PI or co-PI of about 20 research projects supported by the Canadian government and academic funding agencies. He maintains extensive research collaborations with colleagues across Canada, in China and France. Besides research, he also contributes to scientific management, journal editing and reviewing, and serves as a member for various Canadian and international oceanographic committees. 

报告摘要:

Prediction of ocean condition changes at various space-time scales is of great value to ecosystem and fishery applications. However, the predictability, or accuracy in prediction, varies in different regions due to different complexity of ocean dynamics. Here we identify challenges in modelling and predicting ocean condition changes in the Northwest Atlantic, in particularly on the Scotian Shelf, at interannual and longer time scales. This is elucidated through inter-comparison of sea levels during 1993-2023 from tide gauge and altimeter observations, and four ocean models. These include two models covering Canada’s Three Oceans without data assimilation, and two global ocean models including data assimilation, at ¼ or 1/12-degee horizontal resolutions. For the Northwest Atlantic, only the 1/12-degree global ocean reanalysis with advanced data assimilation (GLORYS12v1) possesses skills for coastal sea level variations at time scales longer than 20 months. The low skills of the other three models are attributed to the influence of strong nonlinear dynamics, in particular the interaction of Labrador Current ad Gulf Stream near the Tail of Grand Banks. Further evaluation reveals that ocean temperatures from GLORYS12v1 are in good agreement with observation, hence can be applied to study ecosystem and fishery variations in the past few decades. However, reliable predictions for future changes in the region may need innovative methods (e.g., ensemble prediction and/or Machine Learning) to overcome the challenges of nonlinear dynamics.

欢迎广大师生踊跃参加！

气候系统预测与变化应对全国重点实验室

气象灾害教育部重点实验室

气象灾害预报预警与评估协同创新中心

大气科学学院

2026.4.17