Hello stranger,

I am currently finishing a post-doctoral research appointment at the Georgia Institute of Technology, department of Earth and Atmospheric Sciences, with Dr Kim Cobb. After December 1st, 2008, I will be an assistant professor at the University of Southern California.

Broadly speaking, my research is concerned with the role of the Tropics in long term climate variability. The central actor of this game is of course the El Niño-Southern Oscillation (ENSO) phenomenon, whose behavior I strive to understand on decadal to millennial timescales.

My current research project is "Constraining the tropical Pacific’s role in low-frequency climate change of the last millennium", with Drs Kim Cobb, Michael Mann, Nick Graham, Mike Alexander and Martin Hoerling, funded by NOAA

The first phase of this project entailed reconstructing tropical Pacific sea-surface temperatures (SSTs) back to 1000 A.D, using a network of geological indicators of climate and a sophisticated statistical technique. Here is the general idea: consider the picture painted in the tropical Pacific by the instrumental record of the past 150 years. The temperature field varies as a function of time, latitude and longitude, but we can fold latitude and longitude into a single location index (362 points, each corresponding to a 5 by 5 degree box). The following picture is such a represensation of Kaplan SST over the tropical Pacific, smoothed with a 10-year lowpass filter to highlight slowly evolving phenomena.

As you can see, this looks rather like a tapestry. The question is: how can we weave this climate tapestry back in time ? And how reliable would be such a reconstruction? To answer these questions, I used the coarse threads provided by multiple proxies from around the tropics, and a statistical method called RegEM. The proxies are geological objects (corals, ice cores, stalagmites, sediment cores, tree rings) that sensitively record climate information like temperature and precipitation perturbations, generated by ENSO even in remote regions of the Tropics. By making use of the correlation matrix estimated over the period of overlap between instrumental meansurements and proxy measurements, we can fill in the holes in the tapestry. This does introduce errors and biases, however, so it is just as critical to represent those as fairly as possible.

I am also working on modelling ENSO teleconnections (the interconnectedness of climate regions) during the Medieval Climate Anomaly (around 1000 - 1300 A.D.) . For this I use an atmospheric general circulation model (CAM2) and a mixed-layer ocean model (MLM). More generally, I am interested in modeling tropical climate over the late Holocene to provide a dynamical interpretion for the statistical reconstructions mentioned above.

Student Opportunities : I am looking for original thinkers for a numbers exciting projects at the intersection of paleoclimatology, biogeochemistry, and climate modeling.

Post-doctoral Opportunities : I am looking for a statistician (especially a Bayesian) with interest in geosciences to develop climate reconstruction methods.

Please contact me if interested