Statement of Research Interests

My research interests are directed toward understanding the dynamics of atmospheric, oceanic, and sea-ice components of the climate system, as well as the ways these components are coupled together. My approach is to use numerical models of intermediate complexity and combine them with state-of-the-art statistical/dynamical data analysis methods to diagnose the modes of climate variability and estimate their predictability. I have also been interested in employing such intermediate models to develop parameterizations of subgrid-scale processes for the use in coarseresolution GCMs, and to guide data analysis.

The climate system varies on a wide range of spatial and temporal scales. Understanding and quantifying natural climate variability is important, among other things, for properly assessing the magnitude of anthropogenic climate change. My research has primarily been concentrated on the large-scale, low-frequency variability (LFV), with time scales longer than those of the atmospheric synoptic eddies. In particular, I have been interested in intrinsic atmospheric intraseasonal variability, with an emphasis on both hemispheres’ annular modes and planetary flow regimes associated with them. For a longer time scale climate behavior, oceanic processes are bound to play an important role due, in part, to ocean’s large inertia and enormous capacity to store heat and carbon dioxide. I have been working on interannual variability associated with El Niño/Southern Oscillation (ENSO), decadal oceanic variability due to instabilities of the wind-driven circulation (WDC), and interdecadal oscillations of the thermohaline circulation (THC).

Intrinsic atmospheric oscillations and ocean’s WDC variability, while occurring within two different fluids and on different time scales, are likely to involve similar dynamics, in which interaction of the large-scale flow with faster mesoscale processes is essential. We have taken initial steps in representing the latter processes as stochastic forcing that affects large-scale circulation; possibility and theoretical justification of such a parameterization are of great interest to me, as it will allow one to substantially reduce computational burden of eddy-resolving simulations. Intrinsic THC variability is known to be sensitive to polar processes, which include, among other things, sea ice. Sea-ice internal time scales are decadal, but it may also serve as an important element in glacial-to-interglacial transitions of the global climate. These involve major changes in the land icesheets, which occur on millenial time scale and are thought to be triggered by the external, to climate system, variations in solar forcing.

In addition to intrinsic and forced climate modes, there is a possibility of inherently coupled atmosphere–ocean–sea-ice behavior. One of the most studied examples of the coupled behavior is ENSO. Interesting potential novel mechanisms on decadal-to-interdecadal scales that I am now studying include interaction of oceanic eddy field with mid-latitude atmospheric LFV, and a nonlocal feedback loop involving joint effect of THC, sea ice, and atmospheric teleconnections from tropics to middle and high latitudes. Both mechanisms might play a role in the Pacific Decadal Oscillation (PDO).

If certain climate modes are dominated by a low number of degrees of freedom, they may be partially predictable. We have developed a methodology of constructing reduced models based on observed or model generated data; these models represent optimally the data’s statistical properties and provide an estimate of predictability associated with a given phenomenon. Our new models describe nonlinear interactions both within large-scale components of the flow and with eddies parameterized stochastically, as well as seasonal cycle. They have shown success in predicting ENSO and mid-latitude LFV, and are now being used in numerous problems related to data analysis.