Collaborative Research: Southern Ocean Convection in climate models: controls and Impacts

点击翻译按钮获取中文摘要

Despite its critical importance for the global heat and carbon cycles, the Southern Ocean, the vast ocean surrounding the Antarctic continent, is the least understood region of the world ocean, particularly because of sparse observations and our incomplete understanding of high latitude processes such as deep sea convection, the mixing and subsequent homogenization of water column properties over thousands of meters depth. Many climate models simulate strong open sea deep convection within the Weddell and occasionally the Ross Sea, with periods ranging from decades to centuries. This process brings a large amount of heat from the subsurface and releases it to the atmosphere, changing ocean and atmospheric temperatures and winds globally, and producing large holes in sea-ice called polynyas. Interestingly, the largest Weddell Sea polynya observed since the 1970s was seen from satellite pictures this past (2017) October, making an understanding of polynyas ever more urgent. The mechanisms underlying Southern Ocean convective variability, the reasons that it differs widely across climate models, and the impacts of variability on ocean circulation, heat and salt cycles remain unclear. This makes it difficult to determine whether models simulate (or fail to simulate) Southern Ocean convection for the right reasons, whether such events are likely to reoccur and whether past changes in convection are exerting an impact on the course of modern climate. This proposal will directly inform the Intergovernmental Panel on Climate Change (IPCC-AR6) assessment project, focusing on a previously understudied area, expanding our fundamental knowledge of the Southern Ocean and its role in the global climate system, and developing new model diagnostics. By providing insights into the relationship between deep ocean convection and ocean/climate indices on multiple timescales, this proposal will potentially inform design of future observational networks. The proposed work will enhance education and dissemination of research via undergraduate teaching and strong outreach activities both in West Philadelphia, where Dr. Marinov will partner with a public elementary school, and in Baltimore, where Dr. Gnanadesikan will continue his strong involvement in the Science Olympiad.The project is organized around four major scientific questions and underlying hypotheses:1. What produces SO deep convection and determines its frequency and intensity across coupled climate models? 2. What are the global oceanic implications of Southern Ocean convection? 3. How do changes in convection impact the transient climate response and climate sensitivity? 4. How do eddy parameterizations and model construction affect the representations of SO convection and resulting oceanic teleconnections? The team will run a set of new simulations and PIs will analyze (a) a suite of models already run at Johns Hopkins (JHU), (b) a suite of new model runs at the Geophysical Fluid Dynamics Laboratory (GDFL), (c) models submitted to the IPCC-AR5/IPCC-AR6 as part of the Coupled Model Intercomparison Project versions 5 and 6 (CMIP5,6). PIs will examine how different models become "preconditioned" for convection by producing a buildup of heat at depth or salt anomalies at the surface; how convective variability produces changes in key water masses and how heat, salt, mass transport signals evolve over time. They will examine the impact of convective variability within a large ensemble of the same coupled climate model, as well as within versions of the GFDL models which differ in model construction (hybrid vs. level coordinate), resolution (1 vs. ¼ degree ocean), and eddy mixing coefficients. A side benefit of the proposal will be building a set of novel metrics than can be used for routine benchmarking and evaluation of GFDL and CMIP6 models. These metrics will help evaluate strengths/weaknesses of SO simulations, including tracking corrections in long-standing model errors from one model generation to another.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.