| | 1 | == 2014-09-17 == |
| | 2 | |
| | 3 | Peter Draper has reported issues with the stack and warp backgrounds in PV3 reprocessed data. The relevant links are [http://ps1sc.ifa.hawaii.edu/PS1wiki/index.php/DRAVG_telecon#Items_for_September_2nd_meeting The DRAVG page] and [http://astro.dur.ac.uk/~pdraper/panstarrs/mdbackgrounds/ Peter's analysis of a stack]. |
| | 4 | |
| | 5 | I reproduced Peter's analysis for that particular stack, and obtained largely identical results and numbers as he presented. Histograms of the object SKY values have a positive bias in the warps, whereas histograms of the SKY values from the stack do not have the same bias. I believe this is the result of two effects. |
| | 6 | |
| | 7 | 1. The warp bias is caused by the intrinsic undersubtraction of the background measurement process. Due to the fact that we know a given image is more likely to have positive stars in it than negative features, we measure the background level by finding the best fit to the full histogram of pixels as well as for the lower half. When these disagree, we assume that the bright end is contaminated with stars, and use the lower half estimate instead. This results in the background values tending to under subtract the sky by a very small amount. The image is then quantized, which doesn't help. Fix that bit. |
| | 8 | |
| | 9 | When the background is re-measured by doing photometry on the warp (or by measuring the sky level during the stacking process), the quantization of the pixel values reduces the sensitivity of the fitting process. This produces a somewhat random sigma-dependent background level that is stored in the warp background model. |
| | 10 | |
| | 11 | 2. The random seed value of the stack level photometry can change the mean stack SKY levels by about a count in the tests I ran. Fixing the seed and switching between psphot and psphotStack produced identical SKY levels, confirming that the background estimate in the two programs is running identically. To check the accuracy of the background value for the stack itself, I |
| | 12 | |
| | 13 | |
