STSpencer
commented
3 years ago Alright, sounds good. For the first part, we have nsb skymap runs corresponding to all of those observing conditions on disk, it's just a question of what analysis to perform on them. As for the questions:
Q1: We've got the first two already (I'll collate the results), the NSB levels over the Camera per pixel and mean per TM. I'm not quite sure what you mean by NSB level over pixel, if you mean spatially then subdividing pixels into further sub-elements would probably be computationally prohibitive. If you mean temporally then that's Q2.
Q2: Maximum change of NSB across Camera,adjacent TMs/superpixels, pixels. For pixels over time these are represented by the timescale plots from #7, we've only got the full moon scenario for that currently but running the other nights would be trivial. Running whole cameras/TMs worth of skymap plots over a night would be very slow, even at hour intervals. If you mean spatially the maximum changes between TMs/superpixels then I don't currently have code for that but could write it based on the results from Q1;@watsonjj do you have anything like this anywhere?
Q3: I could do this semi-analytically using the values from the timescale plots * 2048 or 64. Or alternatively I could run NSB at a half hour after (once) each of the four scenarios above to get a better (and more reliable, less sensitive to normalisation) reading, then run these through the same analysis as Q1 for comparison/ subtract one set of values from the other for each observing scenario.
For the second part, we don't have this currently but I think this should be relatively straightforward to figure out, nsb was built with these sorts of calculations in mind. If we assume our 'dark field' is nominal, then we can do a rough nLb to Hz conversion and then look at the observability plots for a given bright source we choose (lets say the Vela Pulsar) with different set observing thresholds in nLb (then compare the live times with the differing thresholds).
I've got the corresponding results for all the four eta carinae runs uploaded to their respective results folders on this repo (etacardarkgauss3,etacarbrightgauss3,etacarmoonpoint53gauss3,etacarnightmarev2, in order of increasing brightness).
Let's attempt to summarise the results in a few basic numbers.
Some possibilities (to be iterated). For each of the following conditions:
Questions:
... so I guess the above can basically be a table. We need to define half and full moon (in a consistent way to CTA/ASWG). I would take a pragmatic approach to the above, extracting numbers from existing plots (e.g. can read off the rate of change of NSB for the camera) and removing parts that require much more work.
As a second part, it would be good to assess the number of "extra" nights we can expect vs maximum (mean) camera level. For example: "An additional 20 hours observation can be obtained if the camera can operate up to a mean level of 20x nominal"
Other ideas?