If you have any questions, please contact Jack Sayers, jack@caltech.edu.

As a general rule, the Bolocam data should be used as follows:

1) If you would like to fit a parametric model to the Bolocam data, then you should convolve this model with the transfer function of the filtering applied by the Bolocam data processing and compare the result to the filtered image.

2) If you would like to compute aperture photometry from the Bolocam image, or if you would like to study the SZ morphology in a non-parametric way, then you should use the unfiltered image.

Within each tar file, you will find the following fits files.

filtered_image.fits: This is the SZ image produced by Bolocam data reduction pipeline. This processing pipeline significantly filters the SZ signal from the cluster, and produces a highly distorted image. As a result, this image does not represent the true SZ signal from the cluster! The image is 14'*14' with 20" pixels that are in units of uK_CMB. The fits header provides a range of information, including the cluster redshift used for the Bolocam analysis, the properties of the Bolocam PSF (which is approximately Gaussian), and the number of pixels required to equal to solid angle of the PSF. In addition, there is a conversion factor to go from uK_CMB to units of SZ "y". Please note that this conversion factor includes relativistic corrections to the SZ signal based on the X-ray derived electron temperature for the cluster under the assumption that it is isothermal.

filtered_image_noise_realizations.fits: This is a set of 1000 independent noise realizations of the SZ image. These noise realizations include instrumental, atmospheric, and astronomical sources of noise, and fully include all pixel-pixel correlations due to these noise sources. For precise error estimates, a bootstrap monte-carlo using these noise realizations should be performed.

filtered_image_rms.fits: This is the noise RMS in each pixel of the filtered image. The noise covariance matrix of the filtered image is approximately diagonal, and therefore the RMS/pixel is a good, but not perfect, representation of the noise. Therefore, these RMS values can be used to quickly make error estimates. However, error estimates derived from these pixel RMS values should be treated with caution, and the noise realizations described above should be used if precise error estimates are required.

filtered_image_signal_transfer_function.fits: This file contains two extensions representing the real and imaginary parts of the signal transfer function due to the Bolocam data processing. They should be used to filter a candidate model of the SZ signal in an identical way to how the Bolocam SZ image was filtered. Specifically, they should be used as follows. 1) Compute the 2D Fourier transform of the candidate model, using the convention of [0,+dk,+2dk,..,-2dk,-dk] for the angular frequency of the pixels in the Fourier transform. 2) Multiply the Fourier transform of the candidate model with the complex-valued signal transfer function. 3) Fourier transform the result back to image space. This will produce an image of the candidate model that has been filtered identically to the Bolocam SZ image.

unfiltered_image.fits: This file contains the Bolocam SZ image after deconvolving the effects of the filtering due to the Bolocam data processing. As a result, this image provides a true representation of the SZ signal from the cluster. However, the deconvolution results in significant noise on large angular scales within this image. As a result, this image is truncated to a size of 10'*10'. The information within the header of this file is identical to what is included in the filtered_image.fits file.

unfiltered_image_noise_realizations.fits: This is a set of 1000 independent noise realizations of the SZ image, and is analogous to the set of noise realizations for the filtered image.

gnfw_fit_map.fits: This is an image of the best-fit gNFW model to the Bolocam SZ image. The details of this gNFW fit are given in Czakon et al. (arXiv:1406:2800). This image has 2" pixels to allow for comparisons with high resolution data. However, caution should be used, as the Bolocam PSF has ~1' resolution. The image is given in units of SZ "y".

gnfw_fit_mcmc.fits: This file contains a total of 6 extensions, with each extension corresponding to a different parameter in the gNFW fit to the Bolocam data as described in the header. Each extension contains 1000 values for that parameter, corresponding the results from the bootstrap monte-carlo of the gNFW fit. These values can therefore be used to estimate uncertainties and correlations between the gNFW fit parameters.