FF-Sims

The success of the FFs depends entirely on us (the lensing community) providing magnification maps with robust errors for the larger astronomy community to use and exploit this unique data set. One of the key points that was a persistent source of discussion right from the start in the Science Working Group was the calibration of various modeling approaches. Towards that goal we have decided to provide a set of simulations to calibrate the various modeling techniques. The purpose of such an exercise is to (i) calibrate methods and most importantly the efficacy of error determination (ii) provide feedback in time to the selected preliminary map modelers to ensure that the preliminary maps are well calibrated and (iii) publish the results of this exercise highlighting the strengths and weaknesses of each method so that the community at large can choose which map to use for specific purposes. Participation in this exercise is entirely voluntary however, we would strongly encourage all the selected modelers to do so. We intend to provide you with prompt feedback. The purpose of this exercise is not to set up a competition between modelers but rather provide useful feedback that will hopefully come in handy before the preliminary maps for the FFs need to be submitted to STScI. We would be happy to have modelers submit their reconstructions at any time (obviously well before August 31 would be most useful) on a rolling purpose one cluster at a time. We will promptly send the prognosis back upon submission. In addition to 3 simulated clusters we will also provide a 4th cluster in which the assumption of light tracing mass has been relaxed. Our intention is to have a publication describing this calibration exercise finished by the end of the year. We therefore request participants in this exercise to submit all reconstructions (for 3 or all 4 simulated clusters) by the end of October, 2013.

Input for the reconstructions

Below, we outline our plan of action. The goal of this exercise is to compare reconstruction methods broadly starting with the same inputs.

• The comparison will be requested for 3 simulated clusters (a 4th cluster will be also available for those interested). For making their identification easier, we have chosen names of Greek Gods: Hermes, Hephasestus, Ares (and Athena).
• HST FF-like images are provided for each of the three clusters in FITS format. The images can be accessed through the pages dedicated to clusters. For additional details about the parameters, the PSF etc. please refer to the link "HST images" in the Categories menu on the right side of the page.
• for those of you who want to use additional ground-based simulated data, we also provide a few Subaru images covering a FOV of 30'x30'. Again, the images can be accessed from the pages dedicated to each clusters. For more details, please click on "Subaru images" in the Categories menu.
• Together with the FITS images, we provide catalogs of multiple images. These files are named multimages.cat. To facilitate the visualization of the image sets, we have produced corresponding ds9 .reg files (multimages.reg). For details about the format of these catalogs, see "Multiple images" in the Categories menu. Images have either spectroscopic or photometric redshift.
• There are also catalogs of cluster galaxies available. They are named clgal_cat.txt. Region files are provided too (clgal_cat.reg). See "Cluster galaxies" in the Categories menu for more details.

Output requested from each participating modeling team

Each team should submit the following products for assessment for each of the clusters:

• convergence map for sources at redshift z=10
• shear maps (component 1 and 2) for sources at z=10
• errors on the above maps
• plot of fractional area with magnification > mu as a function of mu and errors - this is to check how integrated quantities compare amongst methods

Submission of the results

All groups will submit their model reconstructions to us. We will do the comparison with the "true set" and return the diagnosis to each modeler that submits the required outputs. After collating the results from all participating groups, we will proceed to work on a paper outlining the various modeling approaches and the results of this calibration exercise. We are open to the idea of what all can go into the paper - only simulation comparisons? or comparison of preliminary maps for the first 2 FF clusters?

Suggested timeline

• Simulated data made available on the website to modelers: JUNE 25
• Modelers requested to send in results for comparison: whenever feasible
• Diagnosis returned to the modelers (results of the comparison): promptly after submission
• Deadline for submission of all reconstructions: 31 OCTOBER
• Planned deadline for paper to be completed and submitted with the results and input from all of you (every participant will appear on the author list) end of the year 2013.