CAMB: Lensing potential power spectrum

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Sudeep Das
Posts: 2
Joined: December 13 2005
Affiliation: Princeton University
Contact:

CAMB: Lensing potential power spectrum

Post by Sudeep Das » December 12 2006

Does the Lensing potential power spectrum [tex] C_l^{\phi}[/tex] that CAMB churns out include nonlinear corrections from halofit?

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

Re: CAMB: Lensing potential power spectrum

Post by Antony Lewis » December 12 2006

Not by default, but you can set do_nonlinear = 2 in the .ini file to add corrections.

Sudeep Das
Posts: 2
Joined: December 13 2005
Affiliation: Princeton University
Contact:

Re: CAMB: Lensing potential power spectrum

Post by Sudeep Das » March 01 2007

Antony Lewis wrote:Not by default, but you can set do_nonlinear = 2 in the .ini file to add corrections.
Hi Antony,
When I plot [tex]\ell^4 C_\ell^{\phi\phi} [/tex] vs. [tex]\ell[/tex] with do_nonlinear=0 and do_nonlinear=2 I get almost identical plots:

Image


When I do my own Limber approximation based calculation using the power spectrum from CAMB I get the non-linear potential power spectra to be appreciably higher than the linear case for high l's. I think that should be the case. Am I missing something here?

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

Re: CAMB: Lensing potential power spectrum

Post by Antony Lewis » March 01 2007

Probably not using high enough k_eta_max_scalar ? (note CMB lensing for non-BB is insensitive to the high l part of the lensing potential power spectrum; you need k_eta_max_scalar much higher to get the lensing potential accurately than you do to get the lensed C_l accurately)

Gabriela Calistro Rivera
Posts: 1
Joined: August 15 2011
Affiliation: Heidelberg University

CAMB: Lensing potential power spectrum

Post by Gabriela Calistro Rivera » August 15 2011

Hello,

I'm simulating the lensed CMB in a similar way as Lenspix for my Bachelor Thesis.
Using the HEALPix routine synfast I have calculated the gradient of the lensing potential given by CAMB and as next the C_l^dd of the absolute value of the deflection angle. In the ReadMe file of CAMB, they say

C_l^{dd}= [l(l+1)]^2 /(2\pi) Cl^{\Phi\Phi},

where the C_l^{dd} given as output here is equivalent to (l(l+1)) /(2\pi) * C_l(computed)^{dd} , which is my computed deflection angle power spectrum, right?.

But if I plotting this relation, I get a great difference between both! (uploaded figure plot4.eps).

I don't understand the problem, since it seems that my results of the deflection angles are right (going on with the remapping, the produced lensed temperature power spectra fits very well with the theoretical of CAMB).

I wonder if I have a false interpretation of the information given in the ReadMe?

Thanks for the help in anticipation!

Gabriela

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

Re: CAMB: Lensing potential power spectrum

Post by Antony Lewis » August 17 2011

"absolute value of the deflection angle"? You don't want [tex]|\nabla\psi|[/tex], but the power spectrum of the gradient (E)-mode part of spin-1 harmonic transform of the map of [tex]\nabla\psi[/tex] (see e.g. appendix of astro-ph/0502469)

Post Reply