Welcome to the first of my Published Research Synponses! As I mentioned in the previous entry, this will become somewhat of a regular feature on Anthonares.net. As this is a first attempt, I will be refining these as I go along, but I want to point out the general structure of the synopsis: 1) the citation lists the authors, the title of the article, the journal in which it appeared, the volume and number of the journal, the pages, and finally the year in which it was published and the authors’ affiliations or institutions, 2) the synopsis describes what the publication is about, what its findings are, and what data it used, 3) the context will provide a bit of background on the research or the field if I feel it will help further understanding, and 4) the general explanations will be where I attempt to demystify some of the gobbledygook phrases such as “computer models.”
I picked this week’s paper because it is a prime example of a really successful application of a computer model. Also, I think that Jupiter is just an amazing planet, and to think that, thanks to this article, we now understand it a little better is just wonderful! Now, if only we could get a better handle on that Great Red Spot. The publication appeared in the journal Nature in the November 10th edition. The journal itself is widely available in community libraries, but to access it online you’ll need to go through an institution or individual who has a subscription (thought Nature.com does have free online content).
Citation: (view online at CiteULike.org)
M. Heimpel1, J. Aurnou2, and J. Wicht3. Simulation of equatorial and high-latitude jets on jupiter in a deep convection model. Nature, 438(7065):193–196, 2005.
1Univeristy of Alberta, Edmonton, Alberta Canada.2 UCLA, Los Angelas, California USA.3Max Planck Institute for Solar System Research, Katienburg-Lindau, Germany.
Synopsis:
The authors describe the results of a computer model that attempts to reproduce the general structure of the cloud bands seen on Jupiter. Their simulations are able to recreate the broad bands seen near Jupiter’s equator as well as the much narrower bands nearer to the poles. Their model also accurately reproduces the wind speeds and directions within these bands which we’ve measured using the Voyager, Galileo, and Cassini space probes. They also describe how their model, if applied to Saturn, Neptune, and Uranus can accurately reproduce most of the features seen there as well. They note, in particular, that the winds on Neptune and Uranus, which move in the opposite direction to those planets’ rotation, are explained well by their model.
They explain that the fundamental physical cause of Jupiter’s cloud bands is a phenomenon whereby random turbulence in deeper layers can, on such a very quickly rotating planet, lead to the formation of very large, stable structures at the cloud tops. The mathematical equations that they use in their work are all those of standard fluid mechanics, and would be very familiar to physical oceanographers working on problems here on Earth. The model is a “three-dimensional” one, which means that they have represented the planet Jupiter as a 3-d array of approximately 10 million cells. Within each of those cells, the computer solves mathematical equations for fluid flow. The combined result is the very cool graphic shown at the right, note the similarities to the picture above.
Context:
Earlier computer modeling of Jupiter’s winds have suffered from two main problems, 1) the models did not accurately reproduce what we see with our space probes and telescopes, or 2) in order to do so, the models used improper assumptions about the composition of Jupiter’s atmosphere. This effort, which seems to make reasonable assumptions about Jupiter’s structure, and reproduces its cloud bands quite well, is still not perfect, however. There are very fine bands of clouds between wider bands that the model does not predict. The authors mention several simplifications that may have led to the slightly overly smooth banding seen in their model.
General Explanations:
Computer Models: This extremely general term refers to the use of computers to solve mathematical equations that would be impossible to perform by hand. Computer models sometimes use Newton’s Laws (as is the case for my computer modeling work with groundwater), while others attempt to mimic the function of the human brain, while still others solve thousands of simple algebraic equations whose answer depends on the answer of all of the others. The most crucial thing to know about computer models is that they are prime examples of the “Garbage In - Garbage Out” statement that is true of most computing. A model is not correct simply because it is done on a computer.
Neither is it true that computer models are never accurate. Often, in my field, I encounter the question “But can you believe your model results?” (I remember once getting into this kind of conversation with a guest at my brother’s wedding after having indulged at the open bar. I’m not sure I put up the best defense of computer modeling that evening). To which the standard answer is: “I believe the model results are perfectly accurate, as long as reality matches perfectly the gross simplifications and rectangular representations I’ve been forced to use.” If that is not the case, which it never is, then I trust my computer models inasmuch as I can know how well I’ve described reality. Even then, I still need to make sure that my computer models produce results that correspond reasonably well to real-world data. Luckily, since computer models are capable of matching real-world data pretty well in lots of situations, we’ve not abandoned them altogether.
