Theme I research includes the development of dynamical models and data assimilation procedures to advance the prediction of climate variability, the development of tools and subroutines to improve models of the global atmosphere and ocean, and the use of these models to enhance and expand the fundamental understanding of the climate system, its variability, and predictability.

LDEO research into climate modeling and forecasting dates back to the mid-1980s when Mark Cane and Steve Zebiak launched the first successful numerical coupled model to forecast El Niño. Much of this early research was funded by NOAA grants and today, LDEO and Columbia University Department of Applied Physics and Applied Mathematics researchers are developing new tools for ENSO prediction. These researchers are advancing methods for numerical data assimilation in an effort to improve the initiation of prediction models and thereby the forecasts themselves. They develop and maintain a homegrown ocean modeling effort (the Lamont Ocean-Atmosphere model or LOAM) and adopt existing global climate models for use in our local computer facilities. These climate models are used to study and understand key processes and phenomena of climate and climate variability, such as the overall role of the oceans in climate, the links between tropics and high latitudes and its implication to decadal variability and paleoclimate phenomena, the role of tropical ocean-atmosphere interaction outside the Pacific in regional and global climate variability, and the phenomenon of abrupt climate response to slow and relatively weak changes in external forcing. Theme I modeling and model development activities are also important components of GFDL research. Accordingly, the present and future modeling studies and model development activities under CICAR enrich the hierarchy of numerical tools and modeling scenarios accessible to GFDL and LDEO scientists, thus advancing climate research at both NOAA and Columbia University.


In the context of CICAR, a special collaborative activity between LDEO and GFDL was recently launched. The project entrains research by graduate students and postdoctoral research scientists and is concerned with the plan to conduct a simulation of climate variability in the last millennium using realistic forcing (solar, volcanoes, and trace gases) derived from best available observations and proxy data. The simulation results will be verified against proxies such as data from corals, tree rings, and mountain glaciers. Project goals for this simulation are to put future climate change in the context of the variability during the Medieval Warm Period and the Little Ice Age and to study the variability of short-term climate fluctuation under slow changes in external forcing.