Ice streams are believed to play a major role in determining the
response of ice sheets to climate change and in determining global sea
level. In general the motion of ice streams is a response to gravity,
moderated by drag from the sides and base of the stream. Understanding
the relative importance of these two components is essential for
developing models for future evolution of the polar ice sheets. In
this project* we develop hierarchical Bayesian statistical models which rely
heavily on physical models for ice-stream flow. The intent is that the
output of these models combine physical theory and statistical
information contained in observational data, while managing the
uncertainty in both information sources.
I will review the primary datasets used in our first analysis: these
include data on the surface and basal topography as well as estimated
surface velocities for a portion of the North-East Ice Steam in
Greenland. I will also review the simple physics relations
incorporated into the Bayesian model, and then report on the
preliminary data analyses and our experiences in model development. Of
course, an overview of the full Bayesian model, including our plans
for a Markov Chain Monte Carlo analysis, is presented. A summary of
where we are and what's next concludes the talk.
*This talk is a progress report on an NSF-funded collaborative project
featuring researchers from the Byrd Polar Research Center and the
Department of Statistics at Ohio State.
