Workshop: summary
Day 2: Curriculum Development
Webcast from MIT by Glenn Flierl was partially 'translated' by the colleague John Marshall talking to him on the cell phone due to audio difficulties. Glenn addressed how to integrate experiments, data, and models at a graduate level. In his class, students begin to form concepts through exploring dynamical processes using atmospheric and oceanic data. This is accompanied by projects with theory and experiments. 
The topics include 2D PV inversion, geostrophic adjustment, vortex interactions, and Rossby waves. As an example, to learn about fronts, an idealized front between two PV contours is inroduced. Analogies from data include locations of Ertel PV front on isentropic surfaces (atmosphere) and Gulf Stream (ocean). Then in the tank experiment, a central vortex is produced using a rotating disk, leading to the waves and instability of shear and formation of Stewartson layers. Later vortices begin to merge. Finally, numerical models (including simple ones based on Matlab) are used to explore theory further. Using numerical experiments in conjunction with data and laboratory experiments is effective in integrating theory learned in class with observations. The integrative approach is received well by students, and serves as an effective learning tool.
Jianhua Lu talked about idealized climate models as tools for the understanding of some basic concepts of climate dynamics. An idealized climate model with a slab ocean (5m depth), fixed relative humidity, 1x or 2x CO2 values, and critical lapse rate for convection decreasing from 6.5-9.8K/Km. The model is too simple for prediction of climate change, but provides basic physics and dynamics (forces and feedbacks) of climate system. Questions for students may include: Why is temperature chosen as the output? Why is water vapor, lapse rate, and clouds chosen to be dependent upon temperature? Why does radiative forcing have to be at TOA? How to diagnose dynamical feedback? Then he went on to explain that in TOA-based climate feedback analysis all local dynamic feedbacks and part of non-local dynamic feedbacks are lumped into a single feedback agent: the lapse rate. The same model may be used to explain vertical and horizontal structure of climate change. The subsequent discussion on the role of GCMs in education was lively but inconclusive, indicatng that the community is still grappling with the issue.
The second day sessions were followed by a group photograph and dinner.