报告题目：The Atlantic/Pacific atmospheric moisture budget asymmetry
报 告 人：David Ferreira副教授
报告摘要：
The contrast between basin-integrated precipitation minus evaporation (P-E) of the Atlantic and Pacific Oceans is about 0.4 Sv (1 Sv = 109 kg s-1). The Atlantic is net evaporative and the Pacific near neutral – a result consistent across datasets. This asymmetry is linked to the higher sea surface salinity in the Atlantic than in the Pacific (at all latitudes) and the absence of deep-water formation in the Pacific. Here, I will discuss the relationship between the inter-basin P-E contrast and the atmospheric moisture transports. First, it will be shown that the P-E asymmetry is primarily a result of greater Pacific precipitation (notably south of 30oN) rather than greater evaporation over the Atlantic. The combined analysis of moisture fluxes across the catchment boundaries of the basins and moisture-tracking Lagrangian technique reveals that the eastward moisture flux across South-East Asia, rather the often-invoked flux across Central America, is the key to the P-E asymmetry. The anomalous flux across South-East Asia (i.e. contrasting with the westward flux found in other basins at the same latitudes) is related to the Somali low-level jet and Asian Summer Monsoon. These circulation patterns, diverting trajectories away from Africa into the westerly flow towards the Pacific, explain the higher efficiency of the moisture import to the Pacific than to the Atlantic/Indian basins.
These results highlight the importance of continental geometry and steady large-scale circulation features in the Atlantic/Pacific P-E asymmetry. Implications for future climate change and paleoclimate applications will be discussed.
报告题目：Time structures of the model error in data assimilation
报 告 人：Javier Amezcua研究员
报告摘要：
In the past, data assimilation has been performed in the context of perfect model, i.e. considering that forecast errors came only from the propagation of the uncertainty in initial conditions.
The role of model error has become more prominent in the last years. This model error comes from many sources including: sub-grid parameterisations, numerical discretisation, lack of understanding of some physical process, and the so-called representation error arising from the mismatch in resolution between observations and models. Since its origin is so varied, model error can come in all types and flavours. We have been exploring the time structure in model error and the consequences of its misrepresentation.
报告时间：2018年11月14日（周三）9:00
报告地点：气象楼423报告厅
主 持 人：刘飞 教授
欢迎广大师生踊跃参加！
专家简介：
David Ferreira, Associate Professor, department of meteorology, University of Reading.
Research Interests: the dynamics of ocean mesoscale eddies and how to represent their effects in climate models.
Javier Amezcua, NCEO Researcher on Advanced Data Assimilation Methods, Data Assimilation Research Centre and Department of Meteorology, University of Reading.
Research Interests: Data assimilation, predictability, meteorology, inverse problems. applied mathematics, numerical modelling, Monte Carlo methods, dynamical systems, fluid dynamics, probability and statistics, stochastic processes, numerical analysis, uncertainity quantificaion, Bayesian problems.
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
2018年11月12日
