Vakili and I spent some time in the morning discussing the next steps in his work on PSF interpolation. He can show that his method is better than any polynomial interpolation, and it also provides a probabilistic PSF—that is, it returns a probability distribution function over PSFs. We debated whether to write a paper using this on LSST simulation data, which is fake data (bad) but where we know the truth so we can assess accuracy (good), or else using this on SDSS data, which is real data (good) but where we have no extrinsic handle on truth (bad). His assignment is to figure out just how mature LSST simulation outputs are in the relevant regards. We also discussed applying for the NSF WPS call (PDF).
"We debated whether to write a paper using this on LSST simulation data, which is fake data (bad) but where we know the truth so we can assess accuracy (good), or else using this on SDSS data, which is real data (good) but where we have no extrinsic handle on truth (bad)"
ReplyDeleteWhy not both? Also, have you considered using DECam data (stare vs TDI mode?).
PS: Looking forward to the paper!
PhoSim is "mature" in the sense that it has been run a lot. (In the bizarre world of supercomputing freakonomics, having a code that takes a lot of CPU time means you get even more CPU time!) I've not used it myself, but if you want to fire photons through a stochastic atmosphere, it's the tool for the job. If you want "truth", be prepared to fire a *LOT* of photons. (As an aside, Aaron Meisner's models of the WISE PSF span 8 orders of magnitude.) The way you specify the properties of the PSF is often way far removed from the PSF you get, so you may know the "truth" in the sense that you know the random number seed that went into a large stochastic generative model.
ReplyDeleteGalSim may provide more of the truthy truth you desire, and realistic enough PSFs to be interesting. Also, it doesn't require a supercomputer.