My research focuses on ocean-atmosphere dynamics and interactions relevant to the Earth's climate.

One ongoing project focuses on the formation of deep waters. In this process, warm/fresh surface water loses sufficient heat and fresh water to the atmosphere (by cooling and evaporation) to become dense enough to sink to great depths. Since this is a very complicated process, in which oceanic and atmospheric dynamics and thermodynamics interact on small space and time scales, it is very hard to observe. To overcome this major difficulty, my approach has been to employ a `natural lab'. That is, I study the process in the smaller, better observed northern Red Sea, assuming that some of the insights are applicable to larger scales and other sites. In the past, I used a numerical model to study this phenomenon. More recently, we use corals from the formation region to study long-term variability of formation. We have already used relatively modern corals to construct a simple theory for regional air-sea interactions and the evolution of surface density (the key dynamical variable controlling water mass formation). Currently we are working on a 300 year long coral record, which documents the interplay between tropical modes of climate variability (primarily El Nino-Southern Oscillation and Monsoon), and mid-latitude ones (the North Atlantic Oscillation).

Another project involves understanding and forecasting Eastern Mediterranean rainfall variability. Using atmospheric and surface data and statistical techniques, we are now able to forecast Eastern Mediterranean winter rainfall variability robustly and skillfully, with obvious practical applications. More fundamentally, the work demonstrates and explains the influence of planetary waves on downstream predictability. The inherent predictability of the North Atlantic/European sector, which emerges from this work, is something I will pursue in the near future.

My most recent work (circa mid-2001) addresses predictability of geophysical/climatic phenomena in general, and the North Atlantic Oscillation in particular. The work has so far yielded a novel approach to statistical forecasting that is based on drastically reduced-dimension phase-space history of the studied phenomena. The North Atlantic Oscillation makes an excellent test bed in which to evaluate the method's performance. The work addresses the interannual and decadal timescales separately, and, in both, shows extended, substantial predictive skill that has not been reported or suggested previously.

Education:

• Ph. D. (physical oceanography), 1996, LDEO of Columbia University
• M.Phil. (physical oceanography), 1994, LDEO of Columbia University
• M.A. (physical oceanography), 1993, LDEO of Columbia University
• B.A. 1988, Haifa University/Technion, Israel