CAMB: Lensing output

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Jun-Qing Xia
Posts: 22
Joined: January 02 2005
Affiliation: SISSA, Italy

CAMB: Lensing output

Post by Jun-Qing Xia » September 03 2006

Dear all,

I consider the lensing effect on the CMB power spectra.

When I set do_lensing=F in params.ini, the output scalCls.dat is unlensed spectra of TT, EE, TE.
If I set do_lensing=T in params.ini, there are two output files scalCls.dat and lensedCls.dat.
I know the scalCls.dat is spectra of TT, EE, TE, pp, pT and the lensedCls.dat is spectra of TT, EE, BB, TE.
What I confused is the exact means of TT, EE, TE in scalCls.dat and lensedCls.dat respectively.

To my knowledge, when I set do_lensing=T the spectra of TT, EE, TE in scalCls.dat are still unlensed.
And the spectra of TT, EE, BB, TE in lensedCls.dat are lensed power spectra.

If my understanding is correct, the spectra of TT, EE, TE in scalCls.dat should have nothing to do with the "do_lensing" in params.ini.
However I test it and find some difference on small l (0.4% relatively). Is it due to the numerical error or something else?

Thanks!

Jun-Qing Xia

Antony Lewis
Posts: 1941
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: CAMB: Lensing output

Post by Antony Lewis » September 03 2006

When lensing is on the equations are evolved for different ranges of times than without lensing. Since the target default (un-boosted) accuracy is 0.3% there should only be differences of this order. At small \ell 0.4% is of course well below cosmic variance.

Post Reply