My research focuses on the mathematical and physical aspects of the climate system. Of primary interest to me are the physical mechanisms that are responsible for the climate of the Earth, in the present, in the past (paleoclimate), and in the future (climate change). This means addressing questions such as: How does the interaction between the atmosphere and the ocean affect climate? What is the global impact of El Niño? I am also interested in the statistical issues that arise in trying to quantify climate change. Can we simply blame the next big hurricane, or the next major drought on human activities? Or, should we carefully consider the statistical and physical aspects of natural climate variability before drawing such conclusions?
One of important goals of my research is to help improve predictions of climate. Surface boundary conditions play an important role in determining climate predictability. Foreknowledge of the sea surface temperature or the sea ice distribution can help predict the evolution of atmospheric flow on time scales of months to years. Currently my predictability research is focused on the tropical Atlantic sea surface temperatures and on long-term rainfall trends in the Sahel region of Africa. Listed below are some highlights of my recent research in this area.
- Was the prolonged drought in Sahel due to deforestation? We used a numerical model of the atmosphere and land system (created by NASA) to address this question. Observed sea surface temperatures during the 20th century were used as the boundary condition to carry out numerical integrations using the model. No changes were made to the land conditions. We found that the model could simulate the Sahel drought even without the deforestation effect, implying that the drought was most likely caused by changes in the oceanic conditions (Giannini, Saravanan, and Chang; Science, 2003; see figure below).
Figure: Indices of Sahel rainfall variability. Observations used the average of stations between 10°N and 20°N, 20°W and 40°E. Model numbers were based on the ensemble-mean average of gridboxes between 10°N and 20°N, 20°W and 35°E. The correlation between observed and modeled indices of ( JAS) rainfall over 1930–2000 is 0.60. (Time series are standardized to allow for an immediate comparison, because variability in the ensemble mean is muted in comparison to the single observed realization. The ratio of observed to ensemble- mean standard deviations in the Sahel is 4.)
- What is the role of air-sea interaction in tropical Atlantic predictability? Although not as well-known as El Niño, the tropical Atlantic region is perhaps one of the regions of the world which has a rather predictable climate, at least on seasonal timescales. Some of this predictive skill arises simply from the influence of the adjoining Pacific El Niño phenomenon. The remainder of the predictive skill is believed to arise from air-sea interaction that is local to the tropical Atlantic. We have carried out several studies using a hierarchy of coupled and uncoupled numerical models which show that thermodynamic air-sea interaction may play an important role in persisting sea surface temperature anomalies and make them more predictable. (Figure)