Power Spectra with CAMB

Use of Healpix, camb, CLASS, cosmomc, compilers, etc.
Post Reply
Joseph Wraga
Posts: 6
Joined: February 22 2018
Affiliation: Drexel University

Power Spectra with CAMB

Post by Joseph Wraga » March 08 2018

I was going through the tutorial for CAMB (http://camb.readthedocs.io/en/latest/CAMBdemo.html ) and I notice the TT spectrum is shifted downward compared with the Planck 2015 results( https://arxiv.org/pdf/1502.02114.pdf ) . For example, the first peak in the demo is around 5300 [tex] \mu K^2[/tex], while for Planck the peak is around 5700 [tex] \mu K^2[/tex]. I notice the same thing when trying to compare the BB spectrum with the best fit values (third entry in the data set column in http://irsa.ipac.caltech.edu/data/Planc ... lary-data/ ). In both cases it seems multiplying the CAMB results by a constant factor would make them match the Planck observation, but I do not see why that is the case.

Antony Lewis
Posts: 1386
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: Power Spectra with CAMB

Post by Antony Lewis » March 09 2018

The demo is just using some random parameters. Specifically it probably doesn't set a realistic value of A_s. Always check with print(pars) that everything is correct after setting values.

Joseph Wraga
Posts: 6
Joined: February 22 2018
Affiliation: Drexel University

Power Spectra with CAMB

Post by Joseph Wraga » March 10 2018

Yes, that was the problem. Adjusting [tex] A_s[/tex] according to Planck (default value was slightly too small) makes the CAMB result match the best-fit model.
Thank you!

Post Reply