Late Kepler Post

Posted by tigerhawkvok on March 10, 2009 20:16 in kepler , astronomy , news , research , IYA

So, I composed most of a post on the Kepler launch, then, well, forgot to post it. Belatedly, here we are:

So, Kepler. I mentioned it a little while ago, what was I talking about? Well, I was talking about the Kepler mission (kepler.gov) being sent up by NASA, set to launch on 3/6/09 (This is one of those "Whoops, posted late" parts).

Kepler is a sattelite that will observer one portion of the sky for its entire mission, looking for transit events. To understand a transit event, consider a lightbulb. When something passes in front of it, it obscures part of the surface from your view, and thus reduces the amount of light you observe. So, by observing the same starfield for its mission duration, Kepler can detect any minute changes in stellar flux, thus detecting a planet. The amount of flux change then gives you the planetary size (since color/spectral profile -> temperature -> size -> flux ).

However, there is a caveat here. Sunspots can often be regular, long lasting, or otherwise look like planets. So, my own impact on the Kepler project came with my research work with Gibor Basri, in which we wrote routines in IDL to analyze our own star for the influence of magnetograms on stellar luminosity profiles (IE, where are sunspots? How big? etc).

Various identification methods picking out
umbrae and penumbrae

A particularly good example can be seen in the figure to the left, where various versions of the algorithms pick out different features in a sunspot group. It is important to note that the differences are exaggerated -- the sun is very bright, so the "dark" spots are valued about 0.85 on an absolute scale. you can also see on the right an early version of the algorithm picking out by far most of the major solar features, including the harder-to-discern faculae, or unsually bright areas (which are, of course, very hard to observe in photos such as these).

Early feature identification

The detection and automation algorithms we developed were actually fairly robust and accurate, though CPU intensive (running through about 100 photos took about 3-4 hours on my now-deceased laptop Liz), though our funding ran out over here before I could complete an algorithm to reverse-construct a star from its magnetic profile. Thus, by also observing the long-range magnetic profile of the stars (particularly Sun-like stars) we can rule out some categories of false positives. Looking at some more "final" algorithm photos, you can see why this could be relevant:

light contrast

Solar feature identification with
later versions of the algorithm

If a dark spot such as that passed along the star, it is enough of a brightness dip that it could register as a false positive — that is to say, it could look very much like a planet. While prolonged observation is one way to get around this, feature identification is another way. And I had a part in it! Nifty!


Updated: 3/13/09

IYA Entry

Curse of Budget Cuts and the Would-Be Biologist.

Posted by tigerhawkvok on July 02, 2008 16:22 in General , kepler

So, budget cuts have finally caught up with me, and I am now working for Kepler on a solely volunteer basis. In the ensuing job search, I turned to Dr. Padian who hooked me up with one poetential website job, and in adddition hooked me up directly with a job which involves creant a k-12 teacher / publisher site for demonstrations and examples of how to teah "macroevolution" to (particularly) high-school students. This promises to be quite intersting -- and I encourage my readers to keep periodic tabs on the beta site. I will update as the site progresses and I get more information on the objectives. Now, more prep-work on LHS stuff and website work. Of course, after watching Heisenberg eat.

Hawai'i, Move, and Internet

Posted by tigerhawkvok on June 08, 2008 14:10 in General , kepler , computers , lhs

Having just returned from a trip to Hawai'i, I would very strongly recommend it to anyone who is thinking about going -- O'ahu in particular is a beautiful island, and the ocean is beautiful. I've uploaded many photos to my Flickr page, should anyone be curious.

Tomorrow I return to Berkeley via Amtrak. Thankfully, there are power cords -- I fully intend on working more on sauropods and doing some course outlining for the LHS Adapt or Die camp. Later this month, I will talk with Jim, Kevin, and Gibor about recommendations; and, around August, if things go well, I will talk to Erica about recommendations as well.

What will be irritating about the next few days is lack of internet -- its not getting activated until the 12th, which, needless to say, is painfully long from now. I suspect much of my days will be spent setting up the new place, but if I have down time, I will see about working on the stellar models from the UGAstro room in Campbell. I have a few colortable ideas, and in addition I will probably rewrite some of the interface when I get it done in my head. No use having so many programs to call and save files when you can do it better internal to one program or by passage of variables. Relying on saving of structures almost obviates their use.

It really is strking how internet dependant everything is nowadays -- my work with Kepler is strictly network based, and I rely on the internet to keep copies and synchronized versions of my biomechanics work across multiple computers. No internet will be hard, indeed.