The Canada-France-Hawaii Telescope on Mauna Kea, Hawaii. CFHTLenS uses data imaged by the CFHT Legacy Survey which completed observations in early 2009. Image credit J.-C. Cuillandre.
Cosmological data products
CFHTLenS revisited: Re-analysis of tomographic cosmic shear data from Joudaki et al. 2016
We present updated CFHTLenS cosmic shear tomography measurements extended to degree scales using a covariance calibrated by a new suite of N-body simulations. We analyze these measurements with a new model fitting pipeline in CosmoMC, accounting for key systematic uncertainties arising from intrinsic galaxy alignments, photometric redshift uncertainties, and baryonic effects in the nonlinear matter power spectrum (using HMCODE). The new measurements, cosmology fitting pipeline, and central MCMC chains are available on github.
CFHTLenS 2D and tomographic cosmic shear data pre 2014
Shear simulations: Access the CFHTLenS Clone mock galaxy catalogues and the weak lensing simulation maps here. These N-body simulation products were central to many of the CFHTLenS cosmology papers and are now publicly available for download.
• Shear two-point functions in user-friendly and 'xipm' format.
The 2PCF components xi+ and xi- as a function of angular scales in arc minutes. The user-friendly format contains the total variance for each component. The 'xipm' version
is a single data vector of (xi+, xi-), compatible with cosmo_pmc.
• Covariance matrix.
The total covariance, which is the sum of shot noise D + mixed term M + non-Gaussian cosmic variance term V.
The individual terms: D (diagonal matrix), M and V. These are useful if the cosmology-dependence of M and V is to be employed.
All matrices are in cosmo_pmc-compatible block format.
• The redshift distribution histogram
The normalized number of galaxies in redshift bins, from the sum of all pdfs. The redshift values are the left bin corners. The format is 'hist', compatible with nicaea and cosmo_pmc.
Tarball Download: http://www.roe.ac.uk/~heymans/CFHTLenS/cfhtlens_6bin_tomography.tar
This tarball contains four directories:
For all cosmological applications use either;
i) full_sample - tomographic correlation functions measured from the full galaxy sample which was used in the cosmological analysis of Heymans et al 2013
ii) blu_sample - tomographic correlation functions measured from the blue galaxy sample which was shown in Heymans et al 2013 to have an intrinsic alignment signal that was consistent with zero.
If you wish to use this tomographic data to constrain cosmological parameters without including intrinsic alignment nuisance parameters, then we recommend using the blu_sample data.
For intrinsic alignment modelling you might also find useful the data from our;
iii) red_sample - tomographic correlation functions measured from the red galaxy sample which was shown in Heymans et al 2013 to have a significantly non-zero intrinsic alignment signal.
iv) red_foreground_blu_background - tomographic correlation functions measured using the red galaxy sample as the forground, and the blue galaxy sample as the background in order to maximise the signal-to-noise of the measured red-galaxy intrinsic alignment GI signal.
Each directory contains the following data files:
This is the 6bin tomographic correlation function. The format is PMC format:
The first i=1,5 lines are the xi_plus data at 5 different theta values in this format:
theta(i) xip(i,1,1) xip(i,1,2) xip(i,1,3) xip(i,1,4) xip(i,1,5) xip(i,1,6) xip(i,2,2) xip(i,2,3) xip(i,2,4)......xip(i,5,5) xip(i,5,6) xip(i,6,6)
The following i=1,5 lines are the xi_minus data
theta(i) xim(i,1,1) xim(i,1,2) xim(i,1,3) xim(i,1,4) xim(i,1,5) xim(i,1,6) xim(i,2,2) xim(i,2,3) xim(i,2,4)......xim(i,5,5) xim(i,5,6) xim(i,6,6)
where (i,j) corresponds to the correlation between tomographic bin i and j. With 6 bins there are 21 combinations.
This is the histogram n(z) for each tomographic redshift bin i and has the format: z n(z)
z is taken to be the left bin corner.
This is the covariance matrix: 210x210
The format is PMC again - but to be clear here is the relevant fortran used to create the file.
integer :: nzc, nbins ! nzc = number of tomographic bins, nbins= number of theta bins
parameter(nzc = 6, nbins=5)
integer :: nc
integer :: ntot
do izl = 1,nzc
do izh = izl,nzc
iz = iz + 1 ! this counts the bin combinations iz=1 =>(1,1), iz=1 =>(1,2) etc
do i = 1,nbins
j = (iz-1)*2*nbins
xi(j+i) = xip(i, izl, izh)
xi(nbins + j+i) = xim(i,izl,izh)
! Then we have an ntot=210 data vector of xi p and xi m and the corresponding covariance matrix is written out as follows
do i = 1,ntot
write(94,'(210e20.10)') (Cov(i,j), j=1,ntot)
When inverting the covariance matrix, remember to apply the Anderson/Hartlap correction in section 3.3.1 of Heymans et al 2013. The supplied covariance matrix has n_mu = 1656 and p = 210.
PMC Chains for this data are available on request to Catherine (heymans[at]roe.ac.uk]).
README Download: http://www.roe.ac.uk/~heymans/CFHTLenS/cfhtlens_2bin_tomography.txt
• The covariance matrix in block format.
• Redshift distribution histogram.
All files are compatible with cosmo_pmc, which can be used to sample the cosmological parameter space using this data.
Acknowledgments in publications
If you use our cosmological data products in a publication, please cite the corresponding CFHTLenS paper(s), see the above references. Also note our general rule for acknowledging CFHTLenS data products.