Showing posts with label cosmology. Show all posts
Showing posts with label cosmology. Show all posts

Wednesday, January 31, 2007

Life lessons from an experimental cosmologist

As some of you know, I work at NASA. I don't do anything particularly brilliant but just upstairs from me is NASA's Nobel-prize-winning pride and joy, John Mather. I sometimes like to think some of his insight and perseverance will rub off on me as we share a few air molecules via the HVAC. (OK that was pathetic.)

Today I attended a luncheon* featuring Dr. Mather. In fewer than 25 minutes, he breezed through his 20 years with the Cosmic Background Explorer (COBE) (which confirmed that the Big Bang is more than just a clever idea); described the forthcoming James Webb Space Telescope (JWST); and ended with the lessons he's learned from it all, ... which I felt deserved a broader audience. So, all three of you, listen up:

- Aim high: the world will change in the 20-year lifetime of your project.
- Your space-based telescope should only try to do what ground-based telescopes can't do.
- Don't be intimidated by difficulty of your project, if no law of nature actually forbids it.
- It it's not forbidden, it's required (in the case of science and exploration).
- If it's not required, it's forbidden (in the case of project management).
- If you don't test it, it won't work (confidence is no guarantee of success).
- If you do test it, it still won't work BUT you'll have a chance to fix it before it has to work.
- Elementary things do fail (simplicity is no guarantee of success).
- All the hard work is worth it -- there's no substitute for major space missions.

I'm no cosmologist, yet something tells me the good Dr.'s onto some genuine wisdom here.

He gave another interesting gem to -ah- chew on: We atoms make up only about 4% of the universe. The rest is cold dark matter (23%), which no-one knows much about; and dark energy (73%), which is even more mysterious.

(* Lunch or luncheon? After extensive study, I have concluded that the "-eon" suffix is code for "wear a tie". Then again, the required preliminary chit-chat does fit the other definition.)