Is there a reasonable rule-of-thumb number of exposures per band over the ten-year survey in the WFD? I see ~800 or 825 cited as the total number of exposures, and I have a sense that there will be more (g, r, i) than (u, y) exposures, but after a bit of searching around I haven’t found a reference for the approximate number of each.
The overall distribution of bands in baseline 3.4 is ugrizy = (25.3, 40.4, 100.0 , 102.1, 90.1, 79.9) normalized to r=100, or normalizing to sum=825, ugrizy= (47.7, 76.1, 188.4, 192.4, 169.8, 150.6). I imagine that, since these averages include the NES and DDF, they’re a bit skewed, but probably not too far off?
1 Like
Further sleuthing revealed https://pstn-051.lsst.io/ with:
Name u g r i z y
Baseline 0.31 0.44 1 1.00 0.90 0.90
Checking the latest bleeding-edge simulation, I’m getting median number of visits in the WFD area of 54, 66, 173, 174, 156, 151 in ugrizy respectively.
In general, there is a reference distribution described in the SRD, which has been further modified by various SCOC recommendations (see https://pstn-053.lsst.io/, https://pstn-055.lsst.io/ and https://pstn-056.lsst.io/), but the goal for the filter balance in the WFD has actually remained rather stable over many years.
Peter’s numbers above should be fairly close to the ratios you have in your message, and other simulations (such as you can find in the metric results under the “General Info Metrics” category on the MAF pages at
https://usdf-maf.slac.stanford.edu) should for the most part show similar ratios as well.
The overall number of visits in WFD will vary a bit - commissioning will tell us whether we can go to a single snap instead of 2x15s exposure per visit, which would boost the overall survey efficiency. The amount of downtime we attempt to include will also modify the total visits a bit. So overall numbers from 800 to 825 or slightly beyond these ranges are all reasonable.