My research interests include: uncertainty analyses of climate forecasting, terrestrial paleoclimate and paleoelevations, and topics in atmospheric physics (in general).
My current position entails working on the use of the MIT 2D Land-Ocean Climate Model (ref1 and ref2 ) to examine the uncertainty of climate projections and their relation with uncertainty in forcing parameters as well as nonlinear interactions. In particular, we are using methods from climate change detection algorithms to identify regions of our model parameter space that are consistent with changes in radiosonde observations from 1961-1995. By using a model that was developed for addressing uncertainties in cloud feedbacks and deep ocean heat uptake, we have two uncertain parameters in the model which can be adjusted to alter the model's long term behavior in response to climate forcings (e.g. increased concentrations of greenhouse gases or aerosols). By systematically varying these model parameters and therefore the model's climate system properties, we are then able to use the detection diagnostics to provide goodness-of-fit statistics that can then be used to estimate regions of parameter space which are inconsistent with observations. The direct goal of this research is to place constraints on properties of the climate system (climate sensitivity, deep ocean heat uptake, uncertain forcings, ...) that have direct bearing on the prediction of future climate changes. Because we are able to vary these properties in our model, these results have implications on the influence on the climate change detection and attribution problems. The first application of these ideas was presented in Forest et al. (2000) (pdf or html ) in Geophysical Research Letters . (The original Joint Program report is here. ) Two subsequent papers have further elaborated on this idea as well as applying it to multiple climate change diagnostics.
Forest et al. (2001) (in Climate Dynamics) has provided a more complete background for the original GRL results as well as providing a thorough look at the sensitivities of the results to important assumptions in the climate change detection algorithm component of the method.
Forest et al. (2002)
(in Science) has applied this technique to three climate-change diagnostics
derived from upper-air, surface, and deep-ocean temperature records. The
major result is that a joint-probability density function for the climate
sensitivity, the rate of heat uptake by the deep ocean, and the uncertainty
in the strength of the net aerosol forcing has been quantified. From this
joint pdf, estimates for the marginal pdf for each climate system property
has been estimated and so objective estimates of the possible ranges for
each property are presented.
An application of these results is given in
Webster et al. (2003)
where uncertainties in the physical climate system (joint pdfs) are combined
with uncertainties in the future emissions of relevant climatic substances
The combined uncertainty in future emissions coupled with the uncertainty
in response to the resulting forcing then presents a more objective estimate
of the uncertainty in future climate change projections.
I have recently been funded with Prof. Bruno Sanso by the NSF - Collaborations in Mathematics and Geophysics Program , to explore rigorous methods for estimating the probability density functions for climate system properties. (Further discussion)
My resume .
My old site at PAOC (formerly CMPO)