I have an ongoing interest in developing and using advanced statistical methods to solve environmental problems. For example Thompson, Dowd, Shen and Greenberg (2002) used the theory of Markov Chains to give a probabilistic characterization of tidal mixing in a tidal embayment. The economy of the representation makes it useful as the physical basis of ecosystem models. The technique is being used to model reef connectivity in the Caribbean by ECONAR, a project funded by NSERC's Collaborative Research Opportunities program. Genton and Thompson (2003) and Thompson and Shen (2004) used a new type of probability distribution, the skew-t, to characterize storm surges and then show how to predict the frequency of extreme floodings under various climate change scenarios. These ideas were used to estimate flooding risk over the next century at Charlottetown, a low lying coastal city of Atlantic Canada (Thompson et al., 2002).

Below are some plots showing the DEM (Digital Elevation Model) of part of Shediac, NB, Canada and the return periods analysis. Click on an image to enlarge it.

DEM of Shediac area, NB, Canada
Winter return periods	Spring return periods
Summer return periods	Fall return periods