Something positive from Hurricane Katrina: a means of data driven meteorology education

In my seven or so years of teaching meteorology, I have found that one of the least engaging, least motivating means of presenting a concept is to use an example from a textbook. These textbook examples are most often black and white, and as such, somewhat lacking in attraction. Additionally, they are always presented from one visual perspective, such as a cross section or a view from above. Another limitation of a textbook example is that there is nothing beyond the printed page. If the textbook example shows heights of two different pressure levels, it says nothing to the student who is interested in temperatures. Or winds. In addition to being lifeless, the textbook example is often presents a case from so far in the past as to be nearly irrelevant to the students of today.

One of the greatest weather phenomena to occur in the recent past is Hurricane Katrina. The significant loss of life and property resulting from this storm, and the resulting coverage by the media, has given this event prominence in the minds of students. Students are interested in Katrina, certainly more-so than upper tropospheric atmospheric dynamics.

Consider the question "What events led to the formation of Katrina?" One answer is that a weak low level circulation moved beneath a region favored for uplift due to the position of a tropical upper tropospheric trough, or TUTT. One might even sketch a simple drawing on the board to illustrate the point. The classic schematic of a TUTT comes from Fitzpatrick et al., (1995):

What if instead of showing this image, or drawing a similar sketch on the board, the students could SEE the upper level trough in the actual data, and then also see the lower level circulation. And then see their motion to the point where the upper level and lower level features work together to lead to the genesis of Katrina?

Unidata has developed an Integrated Data Viewer, IDV, a Java-based software framework for analyzing and visualizing geoscience data. Within this framework, data sets can be used to illustrate concepts. Using IDV, I created a 'bundle'. This bundle consists of pointers to data repositories – in this case the data is from the NCEP/NCAR reanalysis. When run, the bundle loads in these data sets, and plots upper level (250 hPa, approximately 12,400 meters altitude) heights as black contours, the lower level (925 hPa, approximately 800 meters altitude) heights as color filled. I have drawn in the locations of the low level tropical wave with a yellow line, the upper trough with a red line. In addition, I have added in text describing the scene. The image below is what a student would see when they load the bundle:



Viewing these data, as presented, is not the end of the learning experience, it's just the start. This bundle contains 5-days worth of reanalysis data. Students can animate this image, and see how both the upper level and lower level heights evolve over time to see the reason why Katrina formed where it did, not sooner, not later.

In addition to looking at evolution over time, students can also explore additional data from the atmosphere. Initially, the bundle shows the heights of the 925 hPa and 250 hPa surfaces. As the entire reanalysis fields are available, students can add layers of data, perhaps moisture. Indeed, a bundle could be made with any number of data sources: satellite, radar, reanalysis, surface observations, balloon soundings, and profiler data. As a bundle developer, if I find a data set that I consider critical to the lesson but worry that this data set may not be available in the future, I can use the THREDDS data repository to house such data sets. That is one of the great benefits of using IDV – the data sets are stored on a common server, and do not have to be installed on the machines of all the users of the data. The bundles themselves are relatively small, and are easy to share.

Resources

About the Author

Brian Etherton is an assistant professor in the Department of Geography and Earth Sciences at the University of North Carolina Charlotte. Dr. Etherton's research focuses on numerical weather prediction, ensemble forecasting, and data assimilation. He teaches courses in dynamic meteorology and physical meteorology, and is working to make online versions of both courses.