Nonzero Clkg from CambSources at very low ell even when b=0

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Alex Krolewski
Posts: 11
Joined: November 07 2019
Affiliation: UC Berkeley

Nonzero Clkg from CambSources at very low ell even when b=0

Post by Alex Krolewski » June 28 2021

Hi,

We have found a nonzero signal for Clkg, up to 10% of the b=1 case at very low l (<20), even when b=0 in Cambsources. An example script is on github here:

https://github.com/akrolewski/cambources_test/blob/main/example_for_cosmo_coffee.py

(the repo includes the text file for the dN/dz as well). This signal appears whether I set nonlinear=True or False, and doesn't appear to be affected by the Limber parameters either. Do you know what might be going on?

Alex Krolewski

Antony Lewis
Posts: 1936
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: Nonzero Clkg from CambSources at very low ell even when b=0

Post by Antony Lewis » June 29 2021

I think this is because the "counts" source term is defined in the Newtonian gauge. Bias affects the synchronous-gauge density, so the Newtonian one is non-zero (if very small) on large scales even with bias=0. See Eq 31 of http://arxiv.org/abs/1105.5292

Alex Krolewski
Posts: 11
Joined: November 07 2019
Affiliation: UC Berkeley

Re: Nonzero Clkg from CambSources at very low ell even when b=0

Post by Alex Krolewski » June 30 2021

Thanks, that clears it up!

Post Reply