From the CCSM3 simulation of the last deglaciation, I find that basin-scale OCAPE starts to appear in the North Atlantic and is accumulated over decades at the end of Heinrich Stadial 1. Using a high-resolution eddy-resolving model, I find that this accumulation of OCAPE ultimately overshoots the intrinsic threshold of fluid instability and is released abruptly into kinetic energy of convection. This causes a ~2 °C sea surface warming for the whole Atlantic basin (~700 km) within a month, which can significantly contribute to the abrupt Bølling-Allerød warming. The upper movie shows the abrupt convection occurring in a ~44km wide subdomain, while the lower movie shows the convection in the whole domain (~700km). Details can be found in my paper here.

I wrote a two-dimensional eddy-resolving non-hydrostatic model to study the ocean deep convection. Weakly stratified ocean columns are widely observed in wintertime polar oceans such as in the Weddell Sea and Greenland Sea, having cold fresh water (blue in the movie) overlying warm salty water (red). I find that thermobaricity builds up Ocean Convective available Potential Energy (OCAPE) in these ocean columns. OCAPE can be released into kinetic energy and hence powers the abrupt deep convection, with further acceleration from Cabbeling instability. Convection occurs very locally, at about 0.1-1km horizontal scales. Within only a few day, the ocean column is totally mixed due to the convection. Similar convection can occur in real ocean such as the wintertime Weddell sea, as found in my simulation. Details can be found in my papers here and here.

This movie shows the surface temperature of the Weddell Gyre (run by NASA Ames; animation from Andrew Stewart). Due to the smallness of Rossby radius of deformation and the blowing of surface wind, the Weddell gyre is highly turbulent especially when close to the Antarctic continental shelf. I aim to understand the heat/salt exchange between the gyre interior and the shelves driven by the interaction between eddies, wind-driven mean flow, and sea-ice, while also evaluating the associated impacts on the production/export of the densest water mass of the world ocean. My long-term goal is to understand the role of local ocean processes around the continental shelf in driving global ocean circulation.

This cool movie shows the sea surface velocity of the global ocean, from ECCO2 products. It is clear that the ocean is full of mesoscale eddies, especially at the Southern Ocean and the western boundary region of subtropical gyres such as the Gulf Stream and the Kuroshio Current. These mesoscale eddies account for the majority of oceanic kinetic energy. An important energy source for these mesoscale eddies comes from the release of Available Potential Energy (APE) through Baroclinic instability. I aim to better characterize the general relations between APE and Eddy Kinetic Energy (EKE). Details can be found in my paper here.


This is the satellite observation (year 2002-2007) for the Antarctic sea-ice trend. In contrast to the dramatic loss of sea-ice in Arctic, Antarctica actually achieves a larger sea-ice extent in recent decades. The trend of sea-ice is related to the wind-driven sea-ice drift, and the local thermodynamic production/melting of sea-ice. I am interested in the dynamic interaction between the Southern Ocean and the sea-ice, and the resulting impacts on the trend of Antarctic sea-ice extent. See my paper here.

Research Overview


My research aims to apply mathematical and modeling techniques to analyze large datasets and develop successful quantitative models or statistical algorithms to study stochastic systems such as ocean turbulence, climate, and other stochastic process like financial time series.


I have worked on modeling and developing analytical steps to examining how small-scale ocean eddies (<50km), which are ubiquitous by observations but not resolved in climate models, may strongly affect the global climate including the global heat budget and large-scale ocean circulation. See my paper here.


I have been excited in analyzing big datasets and applying statistical tools to study how atmosphere-ocean-ice interaction plays a key role in the global sea ice budget and its decadal evolution, using satellite and modeling tools. See my paper here.


I have developed simulations and algorithms on studying ocean deep convection and analyzing how it affects the sea-ice budget and the global Meridional Overturning Circulation. These convective events produce key ocean deep water, feed large-scale ocean circulation, and efficiently transport heat, carbon and other tracers. See my papers here and here,


I also implemented quantitative modeling and statistical analysis on mesoscale dynamics such as the turbulent Weddell Gyre and its interaction with Antarctic Circumpolar Current. See my paper here. I also developed statistical algorithms/methodologies to examine the global pattern of mesoscale eddies and its close relationship with Available Potential Energy. See my paper here.


I used modeling and analytical methods to study the mechanism of abrupt climate-change events related to paleoclimate. See my paper here.


I have improved the statistical model and strategies for retrieving greenhouse gases, using remote sensing and radiative transfer theory. See my papers here and here.



Animations of my research

This is my idealized simulation of the Weddell Gyre using a two-layer shallow water model. The surface wind stress curl drives the Ekman pumping/suction, which increases the slope of isopycnals. This generates the Available Potential Energy (APE) locally, which is released into kinetic energy through Baroclinic instability. This process flattens isopycnals and powers Mesoscale eddies of the gyre, as shown in the movie. I find that topography strongly modulates the potential-vorticity budget of the gyre, which hence breaks the Sverdrup balance at interior and the prediction from the classic Stommel–Munk gyre model. see my paper here.


This movie shows the surface relative vorticity around the Gulf Stream (using a 1/48º model run by NASA/AMES. See my paper here). There are energetic eddies at scales smaller than 50km over the global ocean, which are not resolved in climate models. I am currently studying the impacts of these small eddies on the global climate, e.g., by the vertical transports of physical and biogeochemical tracers, and powering the large-scale ocean circulation by upscale energy cascade.


See below for another movie for the Southern Ocean case (animations by Zhan Su).