Difference between revisions of "Coldspotz June Visit to Pasadena"

From CoolWiki
Jump to navigationJump to search
Line 16: Line 16:
  
 
[[File:8.jpg|600px|Welcome to Planck-land]]
 
[[File:8.jpg|600px|Welcome to Planck-land]]
 +
 +
[[File:9.jpg|200px|Tough work!]]
  
 
'''Concepts Learnt:'''
 
'''Concepts Learnt:'''

Revision as of 18:23, 8 August 2012

Schedule during the Visit: Media:schedule.pdf

SSC/Caltech Rooftop

On the JPL Tour

Chillin' after a hard weeks' work

Chris, da' funny maan

That iphone looks so delicious, I could eat it

Gang@JPL courtesy of Chelen

Presenting results

Welcome to Planck-land

Tough work!

Concepts Learnt:

1. Planck vision is not as sharp as WISE. Need WISE to get better position for source. W4 or W3 is the band of choice since they both sample dust emission. Sometimes multiple W4 or W3 sources might be within the Planck beam. If one of the sources is exceptionally bright, that is most likely our counterpart. Otherwise, there could be multiple sources contributing to the Planck flux density.

2. Fitting a model to data and testing the accuracy of the model using additional data e.g. other Planck bands, IRAS, 2MASS

3. Deriving a new model which fits all the data and using that to conclude nature of the source. Normalization of the model to the measured flux densities is important to get the true best fit.

4. Generating spectral energy distribution of a source (Brightness vs Wavelength). Different classes of objects show different spectral energy distributions. Examples of object classes are asteroids, stars, stars with dust shells/disks, dusty nearby galaxies, planetary nebulae, supermassive black holes, distant star-forming galaxies.

5. Using morphology of sources using SDSS or 2MASS to get independent confirmation of nature of source. Always remember to use the most accurate coordinates for the source - this is NOT the Planck coordinates.

6. Statistics - what is signal to noise, why is the measurement of uncertainty or noise important, how can one estimate whether a fit is good when there are multiple data points with different uncertainties.