Given our scientific interest centered on the BAO signal with LSST, and on the corresponding related activities outlined in the preceding section, the LUPM group has defined a work plan in close collaboration with LSST colleagues, with the short-term goal to develop local knowledge of the science, the data, and the instrument, and to quickly increase our collaboration with French colleagues.
Taking part to the French effort around the photometric redshifts will materialize along two different lines :
We are going to finalize our use of the local proprietary galaxy spectra obtained with the FORS2 instrument to critically reassess the use of interpolations based on 6 historical spectra, as was done in \cite{Gorecki_2014}. The key aspect of this effort, beyond the obvious pedagogical opportunity for some of us, is to investigate the possibility to separate better galaxy types and use this information prior to performing a template fitting. Beside the FORS2 spectra, we are also studying the SED atlas recently published \cite{Brown_2014}. This work is lead by E. Giraud, and will proceed in close partnership with the LPSC and LAL LSST team. It is worth noting that we are also in contact with photo-z experts within the OCEVU labex perimeter, especially at IRAP. Finally, a note is already being drafted, that will eventually summarize our study.
We are taking part to the benchmarking of the homegrown LSST-France photo-z code, which is a hybrid template fitting algorithm with a final trained classification step to remove outliers. This code, developed at LAL and LPSC, needs to be compared to other mature software on identical real and simulated catalogs. The LAL team has recently obtained a matched spectro-photo catalog of more than 100000 galaxies that will serve as input. This activity will be collective in nature, for instance with Johann Cohen-Tanugi in charge of shortly deploying and investigating at the CC the SkyNet toolSuite \cite{Graff_2014} in order to compare to recent photo-z benchmarks \cite{2013arXiv1312.1287B}; He is already in contact with Christopher Bonnett.
Beyond the photo-z crucial effort to make BAO measurements competitive with LSST, members of the team are starting to work toward the following goals :
Denis Puy will take part to the theoretical working groups within DESC, especially on combining probes and large scale structure observables and theoretical constraints.
Eric Nuss and Johann Cohen-Tanugi are teaming up with the LAL team to improve and benchmark the homegrown BAO code DirectSim \cite{2010arXiv1009.4769A,2011AAS...21725219A}.
An interesting line of development is lead by Dominique Boutigny (CC-IN2P3, currently at SLAC), who studies GPU versions of a clustering tool usable for BAO.
Last but not least, the team intends to quickly come up to speed on the current status of photometric calibration with LSST. We will benefit from expertise within LSST-France (see for instance author list of \cite{2010AAS...21540119B} and \cite{2010SPIE.7737E..1FJ}), and will bring something specific along the three following lines :
The local expertise on Gaia will serve as a starting point to study in depth the potential usage of the Gaia catalogue to improve upon the baseline calibration model that LSST has adopted \cite{2010SPIE.7737E..1DB}. This work, expected to span several years, will involve most members of the LUPM team. Fabrice Feinstein will coordinate this effort in the lab.
Study of the current performance of the LSST stack software, and improvement thereof, needs large-scale realistic datasets. The CFHTLS provides such an entry point, and LSST-France, helped by the key software developpers of the stack in the U.S.A., is leading the planned “Data Challenge" to reprocess the CFHTLS. In our goal to focus on the photometry measurement, this effort represents an excellent opportunity to
familiarize ourselves with the stack software,
study real images of quality close to what LSST needs to achieve, in order to develop knowledge and intuition about such kind of observations,
take part to the LSST-France effort toward a full-scale computing framework at the CC for LSST processing (a critical aspect of the French contribution to the project, as mentioned above).
Johann Cohen-Tanugi is starting to help LPNHE team with software support and deployment of the electro-optical tests that will be performed there, and will act as a contact person for the SLAC team in charge of the sensor testing software framework. This represents an opportunity to get acquainted with the CCD typical responses and calibrations, a basic element to the photometry calibration, and an important ‘instrumental’ experience for the LUPM team, as it will not have hardware contribution by itself.
We conclude with the mention that this Work Plan has been discussed with and validated by several colleagues, notably Pierre Antilogus, Pierre Astier, Richard Dubois.