CFHT multi-goal wide wide wide survey

[Please circulate and please reply to build an email list--this one is
probably a small subset of those interested.]
At the recent CFHT User's Meeting in Marseille many of you had great
ideas about how a 5-10,000 square degree Megaprime survey could be used
to enable a large range of new science opportunities that span
everything from the solar system to z~7 objects. The basic idea is to go
2-3 mags deeper than the SDSS, placed exactly on top of the SDSS survey
to allow their spectroscopy, photometry and astrometry to be used. An
interesting mid-term idea is to develop a wide field spectrograph to
later acquire spectra.

Aside from the immediate science, this survey could be an important
driver for ALMA, JWST and E-ELT/TMT observations.

Science drivers include:
o large scale structure out to z~1.5 (including BAO measurements, if
competitive at time-of-survey)
o high redshift "drop-outs" (qso's mainly)
o solar system dynamical mapping, including the high latitude population
o galactic structure, with proper motions immediately available at the
SDSS brightness level
o galaxy clusters, for themselves and to build a large sample of strong
arcs which will be important for 8m and ELT observations
o local galaxies in the 10,000 square degree sky.

The strength of this idea is that there is no single science driver, and
many new ones would likely appear as the data were acquired. Ideally the
survey would be managed by a fairly broad committee that would try to
optimize all the science returns with the highest possible observational
efficiency.

It is possible to acquire 10-20 one square degree images per hour. This
means that 10,000 sq degrees cost 500 hours, or about 100 nights for
single filter images. Depending on repeats (and duration of the repeat)
and number of filters this will bring the survey into the 500 night
range of the CFHTLS. There is clearly at least one factor of 2 or so
uncertainty in this estimate, but it is certainly feasible without
taking the whole telescope time and should be of very broad interest and
utility. The data would ideally be available for the supporting
communities and there would be some committment to provide reduced data
products.

Those that are interested should start to develop these ideas to a much
more precise level and be ready for a call for proposals. Ideally we
could all get this underway as soon as possible--a quick start would be
invaluable.