It is time to stop quivering in our boots in pointless fear of the future and just roll up our sleeves and build it.
- Ray Pierrehumbert

Tuesday, April 5, 2011

Cat out of Bag

It's obvious that I'll have to come back to what temperature really means and doesn't mean in climate change. Right now I have another provocative comment I need to follow up on, since a couple of younger readers seem tempted by the trap. I will explain my impression of why undergraduate programs in meteorology exist, and why for the most part, at least as they are constituted in the United States, they should not.

Once upon a time, Abraham Lincoln, who for historical reasons was actually a Republican (the Republican party began in Wisconsin as a radical anti-slavery movement but I digress) suggested spreading around federal money to create agricultural universities in each of the remote new states. (This is why Lincoln's statue sits atop Bascom Hill.) Of course, weather and climate were interests of these agricultural schools, and a tradition of having a state climatologist, more or less a collector of statistics in big pencil spreadsheets, emerged.

In the twentieth century, under the influence of Bjerknes and Rossby, the Norwegians and the University of Chicago, meteorology emerged from folk wisdom and began to be a quantitative science. However, though the equations for the system were known at that time, hand calculation of their evolution was impractical, and machine calculation would not emerge as a realistic possibility until about 1970. On the other hand, by mid-century a number of elegant abstractions had emerged from people looking for ways to find shortcuts with the established physical equations. This was the heyday of mathematical meteorological dynamics.

During this period, the shortcuts were merged with meteorological lore, and skills in weather forecasting measurably improved. Also, aviation emerged, a field which in its early days was enormously concerned with where storms were coming from and where they were going. An old retired fellow who still hung around the Meteorology department in Madison (when I was pursuing my doctorate there) had been a meteorologist for Pan Am, the first airline (? or at least among the first?) to offer commercial trans-Pacific flights. It was not a desk job. When that flight set out for Honolulu and then Tokyo or Sydney, he'd be on board, advising the pilot to duck this way and that.

Now consider a few facts.

First (this was before climate disruption a.k.a. "global warming" was a consideration for more than a handful of very perceptive scientists) the atmosphere is really beautiful and interesting and a wonderful pursuit, mostly classical physics so no need for all that Heisenberg mumbo jumbo, and offers many opportunities for travel and international cooperation. The pay, not fabulous, is decent, In short, mid-century, a meteorology professorship is a very sweet gig for various sorts of nerd.

Second, "meteorologist" was a meaningful career path. Meteorologists, who become specialists in the weather of a particular place, are in demand not just in the universities, but also in the news business, where different outlets really could compete on skill, in aviation, and in the military. So there was demand for training.

Third, there was already the stamp-collecting sort of climatologist at the land grant colleges, and there is a natural connection to agriculture. This leads to a proliferation of meteorology departments at the land grant colleges. Aside from the fading program at Chicago and the emerging one at MIT, the main centers are Madison, Ann Arbor, Oklahoma State, Penn State, Florida State, Texas A&M, Purdue... The private schools just aren't involved for the most part. No Northwestern, no Rice, no Yale, no Vanderbilt, no Cornell.

Now we come to the paradox of how state universities are funded. As I said in the article where I blurted out reference to this problem, the state universities are forever trying to justify their existence to yahoos in state legislatures who have no compunction about free riding on research expenditures elsewhere. There will always be elements in the state assemblies that will not fund anything that doesn't contribute directly to their constituents pocketbooks. Civilization and intellectual bragging rights don't wash for some of those people. You have to convince them that while, yes, you get some money on the side from research, research is a cost of doing business for a great university, sort of like a big-league football team. Your main purpose is to educate young minds, to bring up the future leaders of the great state of W. Or X, or Y or what have you.

Now the university, having succeeded in defending its funding, parcels it out to the various departments. And notice what it does. It parcels out the funds in proportion to how many student-hours the department teaches. This is why "rocks for jocks" courses exist in geology departments and why every department has a course or two in a major lecture hall that is basically light entertainment. But the massive lecture hall freshman class isn't enough. Individual professors prosper by their research, but the department has to keep as many seats full as possible.

So, it emerges that now that operational meteorology is more of a computer app than a skill, academic meteorology is in a very awkward position. Between 1970 and 1990 the skill set they have been teaching becomes largely obsolete, as computers show up that are powerful enough to actually grind through the math in detail. Yes, a few super-meteorologists are needed to program those machines, but the numbers of trainees needed is going to plummet.

The clearest illustration of this is the grotesque "broadcast meteorology track", a poor man's radio/TV/film major where physics and chemistry and calculus are required for no discernible reason. I call this the David Letterman track, where real success is determined by how far and how fast you can hustle your butt away from meteorology.

Yes, Tom Skilling is great. How many markets can support how many Tom Skillings?

The point is this. If you are a meteorology undergraduate at a state funded land grant college, your purpose is basically to inflate the enrollment so as to keep the number of professorships in the department high. You will learn no marketable skill.

Look at yourself very carefully and decide why this shabby career path interests you. Are you obsessed with meteorology or climatology in some way? If not, forget it altogether. This is still a field in decline for people without postgraduate degrees.

If you are obsessed, decide whether you really are scientist material. Do you love mathematics? Do you love solving problems? Are you happiest when your brain is at its smartest? AND you want to know how the atmosphere works. GREAT! Get a degree in a general science where scale is big enough that a strong teaching tradition exists. Physics. Chemistry. Engineering. Electrical engineering. (Yes, electrical engineering is really good for the brain. You'll probably want some fluid dynamics on the side with that one. Focus on systems theory, communication and control, stability theory.) Take a "clouds for jocks" course on the side from your meteorology department for grins. And apply to a top notch atmospheric science program for grad school.

Filling up the undergrad roster is a problem for the meteorology faculty, but it is not a problem for you. Yet. If you make it to faculty level it may be your problem some day. Try to avoid the land grant colleges that have undergraduate programs in meteorology and this mess will never bite you in the butt.

However, note that this situation is to your advantage early in grad school, as getting a teaching assistantship before you have a funded research program lined up is cake easy at those places.

There was a cat I was not supposed to let out of the bag.

Does anybody think I'm wrong, though?

9 comments:

Deech56 said...

Michael - point of information: Cornell was established as a land-grant university. They've got a state Ag school.

elvin said...

"let the cat out of the bag" resulted from unscrupulous persons selling a "pig in a poke" that was actually a cat. i have learned that if you let the cat out of the bag, it doesn't mean the bag is empty.

Ric said...

Yes please get back to temps some time soon. Roger's line of reasoning could easily spill over into claiming that temps are not rising, or won't continue to rise. Or that predictions of temps don't matter, or that we have no idea whose predictions over the last several decades have been nearest the mark.

I'm not trying to put words in Roger's mouth, but we all know those words are in many mouths. So some followup is appropriate.

skanky said...

I always thought that the UKMO only wanted physics and/or maths graduates for its forecasting path. A quick look on their website though shows that they *also* accept meteorology degrees.

I think it does (sort of) back up your point though.

I'm not sure if that was always the case, or if it reflects more on the UK met degrees, these days.

IIRC Reading & Edinburgh are the two big centres for those.

David B. Benson said...

High demand for electrical engineers to work in the power industry. High pay for interesting work.

And any other form of engineering.

Fawgedabout meteors and their 'oligists.

Rob Carver said...

My academic path is a near copy of mt's suggestion (BS physics, MS meteo from OU, PhD Meteo from Penn State), and he's got it pretty much right. I would add that meteo grad students with meteo undergrad degrees didn't have the best grasp on problem solving. That said, the best grad I knew had a PSU meteo undergrad degree.

As someone who's sat on both sides of the job interview table, I would also say that if a meteo undergrad degree excites you, try to get a minor that makes you more than a just another forecaster (CS works best).

Kooiti Masuda said...

First, a little caricaturized story of the sitation in Japan: If you enter a physics course, you will be motivated to solid state physics (semiconductors, superconductors, magnetics, etc.) or elementary particle physics, nothing else.

(Pure) fluid dynamicists have difficulty in survival in physics departments, though they may be welcome in some engineering departments. So, if you enter a small physics course, you have little chance to experience classical fluid dynamics.

In such a situation, I do not recommend "first finish B.Sc in physics" to those who really want to study meteorology.

It would be much better to have a course of physical science, which consists of a core physics curriculum plus various "minors" including meteorology. I do not think such a course currently exists anywhere in Japan.

Such a course existed in 1950s. In the University of Tokyo, the geophysics and astronomy courses were sub-courses of the physics course. In 1970s when I was a student, the courses had been formally separated, but the geophysics curriculum was still a variant of physics curriculum. I did not like it, since I was more interested in the phenomena of climate rather than equations of physics. Also I was not satisfied to see the geophysics professors were indifferent about what is actually taught in the physics courses they claim mandatory. But, in retrospect, perhaps it was (relatively) a good curriculum.

Since then, the earth and planetary science department was established by merging geophysics, geology and geography, and it became distant from the physics department. The reorganization is surely a progress in terms of professional science. But it might have a negative effect to undergraduate education. I cannot comment on the current situation there, since I have not gathered enough information.

EliRabett said...

If you want to learn fluid dynamics, major in aeronautical engineering:) that or mechanical

Corie said...
This comment has been removed by a blog administrator.