Below is the strawman observing plan that we will discuss in the workshop in October. Changes are still possible! One of the primary goals of the workshop is to solicit input into the observing plan, which will be finalized shortly after the workshop. The available observing time should be regarded as fixed - all other parameters of the survey are up for discussion, but the further away from the strawman outlined below, the more compelling the scientific case for changes will likely need to be.
We expect to devote 150 total nights to the survey over 3 years (50 nights per year). While more time might be available for survey extensions and add-ons, it is unlikely that the total amount of time for the survey proper can be significantly larger than this. For purposes of the workshop we will assume the 150 night total. We will assume that our observations will be spread around moon phase (but some time near full moon will be used for engineering, resulting in a slight decrease in bright time). Given expected weather, we will assume ~1000 hours of observing. We further subdivide this time into 250 hours of "best" image quality, 500 hours of "median" image quality, and 250 hours of "submedian" image quality. We will use only the "best" and "median" conditions for the survey - separate projects will be designed for the "submedian" conditions.
There is an obvious trade-off between number of fields and depth. Our current plan is observe each field for significantly longer than one hour in each filter. This will result in greater depth than the CFHT Legacy "wide" fields, for example. For purposes of this strawman proposal we will assume that will will observe 36 fields (~36 square degrees). We will therefore have ~21 hours of observing time for each field, 14 hours in "median" seeing and 7 in "best" seeing. Obviously observing time is not the same thing as exposure time, but the short ~10s readout time of ODI will keep overheads fairly low.
Our fields will be in three regions of 12 contiguous square degrees each. One field will be at RA~2 (most likely near the celestial equator), one will be at RA~8 (at relatively low galactic latitude, b ~ 20-30) and will will be at RA~14 (declination undetermined). Due to the monsoon seasons, we will not attempt to survey a field near RA~20, although add-on projects in that region would be welcome.
Our "strawman" proposal is to observe in griz, with approximately equal exposure times in each filter. An alternative would be to add u, and reduce the exposure time per filter proportionally. However, we will obtain most of our i-band images in the "best" seeing - this will allow us to build up a deep image in sub-half-arcsecond seeing, thus taking advantage of ODI's excellent image quality. The i-band will also be slightly deeper, to allow i-band observations to be taken in every visit, building up good time-domain coverage. More specifically, in the griz only option we would try to obtain the following distribution of observing times for each field. One can imagine a comparable plan that includes also u.
Each individual exposure would be ~2 minutes of exposure time, which provides a good balance between efficiency and the desire for repeated visits for time domain studies. In any given visit, we would obtain at least 3 dithered exposures in each observed filter. Specifically: