|
V3.5-pops Low-Frequency Responsivity Maps | |
o The purpose of the Low-Frequency (LF) responsivity maps is to quantify
relative spatial throughput variations over the focal plane from the
entire optical train. This includes surface brightness variations
induced by distortion, e.g., from the non-uniform convergence or
divergence of light rays, optical aberrations, and in general,
residuals from improperly characterized PSFs over the focal plane.
o These products complement the High-Frequency (HF) pixel-to-pixel
responsivity maps, which mostly emphasize detector pixel-to-pixel
gain variations. Some LF component may also be present. The LF maps
emphasize power at spatial frequencies characteristic of the WISE
beam convolved with a grid whose size is determined by the number
of available photometric measurements. The LF maps are derived
from profile-fit photometry off level-1b frames which have been
corrected for detector gain variations using the HF maps. Therefore,
the LF maps may also capture residual detector-gain variations not
removed by the HF maps.
o If the LF spatial variations are significant, e.g., they significantly
exceed temporal throughput variations and impact the requirements, the
LF maps can be applied to level-1b products in the pipeline so that
photometric variations are minimized.
o Here's a brief outline of the method:
1. query profile-fit photometry from L1b frameset mdex tables using a number
of constraints (see summaries below for band-specific constraints);
2. perform source matching across all scans and in-scan frame overlaps.
All multiple apparitions of the same sources are grouped;
3. for each group of multiple apparitions, compute the average flux
of the group and normalize each apparition flux by this average.
This ratio therefore becomes a "residual" relative to the average
of all apparitions. Only groups where the number of multiple apparitions
exceeds some minimum are considered. See below for these thresholds.
4. bin the residuals over an N x N grid on the array and increment
ancillary arrays to record their number per bin, accumulated sum,
and RMS's for final average and uncertainty estimation.
5. repeat steps 3-4 on all groups from the master match list in 2.
6. compute the responsivity and metric from the final accumulator arrays.
The relative responsivity per bin is the average of all residuals
therein.
o Overall, the LF responsivity variation for W1,W2 is <~ 2%. The relative
responsivities for these bands appear to be determined to an accuracy of
<~0.78% (S/N >~ 130). For W3,W4 the bulk of the variation is <~ 6% and
the accuracy is <~2% (S/N >~ 50). Five bins (out of 49) in W4 have
residuals between 7-9%. Nonetheless, these bins also have high
uncertainties and can be attributed to low source statistics.
o The LF residual maps are in general qualitative agreement with the
bright-star PSF variance maps on the Survey PSF Page, most notably at
the corners.
o The results below used 15 scans from the v3.5 testbed, i.e., on one side
of each orbit from the 30 available scans. These are very preliminary.
We plan to regenerate the LF maps using many more scans following
deployment of v3.5. The accuracy of the LF map crucially depends on the
number source matches across all scans and their effective depth-of-coverage.
The depth determines the degree to which each array can be randomly sampled
to get a robust estimate of the array-average flux (with which we use to
normalize all apparitions). The goal would be to use >~100 overlapping
scans (e.g., near a pole).
See below for JPGs.
Number of unique input sources read = 169921 Number of bins: 9 x 9. Minimum number of same-source apparitions across frame overlaps: 11 S/N >= 5 Chi-square <= 2.0 WPRO mag. range = 13.5 - 13.65 Active deblend flag = 0 Number of components in fit = 1 Match radius = 2.0 arcsec Relative response per bin: 0.982870 0.993002 0.986097 0.989560 0.991286 0.990784 0.995966 0.988877 0.982894 0.989595 0.996229 0.997795 0.995517 1.003152 0.996424 1.005672 0.998073 0.989028 0.995472 0.998095 1.001640 1.003451 0.999102 1.004365 1.008908 0.995101 0.992598 0.999974 1.008278 1.009011 1.003958 0.998199 1.000077 0.998164 1.000077 0.986678 1.002779 1.007037 1.013828 1.014426 1.005301 1.005510 1.002176 1.002189 0.990220 0.990326 1.008320 1.016318 1.010264 1.009939 1.006971 1.006726 1.002457 0.999673 0.998428 0.997377 1.008346 1.011548 1.006777 1.008649 1.009152 1.001609 0.988090 1.002179 1.003181 1.001717 0.994727 0.997280 0.997482 1.001620 0.995767 0.992225 0.981768 0.993462 0.990879 0.989531 0.991494 0.999737 1.001832 0.996025 0.994581 RMS in relative response per bin: 0.027590 0.029580 0.030585 0.036000 0.030555 0.028691 0.028596 0.028526 0.023561 0.034050 0.030619 0.026868 0.034916 0.029174 0.027176 0.033323 0.029248 0.033684 0.035822 0.032089 0.028591 0.028283 0.030235 0.033986 0.027228 0.026832 0.025237 0.035363 0.032047 0.030215 0.027520 0.030139 0.035550 0.033097 0.030429 0.026747 0.031992 0.031322 0.031306 0.028643 0.033084 0.030868 0.033976 0.031887 0.026905 0.037310 0.034067 0.032401 0.028143 0.029286 0.028661 0.029745 0.030135 0.029935 0.032647 0.025425 0.033372 0.027696 0.031053 0.028291 0.029235 0.035405 0.033607 0.028560 0.027819 0.023650 0.033083 0.031948 0.031804 0.029362 0.028545 0.034853 0.024775 0.025306 0.029141 0.031420 0.033056 0.030852 0.032761 0.029699 0.024958 Number of relative response measurements per bin: 66 112 185 181 179 132 134 86 37 36 65 115 94 125 149 97 83 60 73 131 154 160 134 111 90 52 21 69 120 139 163 234 205 141 89 78 29 71 88 115 119 99 77 59 24 81 116 138 223 201 147 106 88 41 45 56 127 201 140 130 134 104 59 43 96 128 147 129 113 76 47 26 74 105 160 191 184 124 160 97 40 1-sigma uncertainty [~RMS/sqrt(N)] in relative response per bin: 0.003396 0.002795 0.002249 0.002676 0.002284 0.002497 0.002470 0.003076 0.003873 0.005675 0.003798 0.002505 0.003601 0.002609 0.002226 0.003383 0.003210 0.004349 0.004193 0.002804 0.002304 0.002236 0.002612 0.003226 0.002870 0.003721 0.005507 0.004257 0.002925 0.002563 0.002156 0.001970 0.002483 0.002787 0.003225 0.003028 0.005941 0.003717 0.003337 0.002671 0.003033 0.003102 0.003872 0.004151 0.005492 0.004146 0.003163 0.002758 0.001885 0.002066 0.002364 0.002889 0.003212 0.004675 0.004867 0.003398 0.002961 0.001954 0.002624 0.002481 0.002526 0.003472 0.004375 0.004355 0.002839 0.002090 0.002729 0.002813 0.002992 0.003368 0.004164 0.006835 0.002880 0.002470 0.002304 0.002273 0.002437 0.002771 0.002590 0.003015 0.003946
See below for JPGs.
Number of unique input sources read = 135583
Number of bins: 9 x 9.
Minimum number of same-source apparitions across frame overlaps: 10
S/N >= 5
Chi-square <= 2.0
WPRO mag. range = 13.0 - 13.15
Active deblend flag = 0
Number of components in fit = 1
Match radius = 2.0 arcsec
Relative response per bin:
1.000316 1.002795 0.999790 1.001630 1.001471 0.992627 0.996322 0.985233 1.001823
0.994805 1.007156 1.006907 1.002199 1.006585 1.000061 1.001986 0.998268 1.008757
1.002960 0.994001 1.003412 0.995759 1.002647 0.997318 1.002833 0.999081 0.998746
1.012528 1.005262 0.993681 1.007693 0.997830 0.998697 1.001820 1.001706 1.012140
1.012997 0.988181 1.006340 0.996088 0.997343 0.998989 1.003155 0.987815 0.997380
0.992336 1.009648 0.993218 0.997091 0.998438 1.003296 0.989567 0.994960 1.003207
1.001579 0.994168 1.001799 1.001273 0.997104 0.998148 1.000115 0.987968 1.002468
1.001040 1.008508 0.999274 1.004988 1.010114 1.007058 0.995469 1.013461 1.003427
0.982162 1.002794 0.996200 1.001022 1.000605 1.001315 0.995471 1.001282 0.997435
RMS in relative response per bin:
0.029595 0.039077 0.037222 0.036326 0.031753 0.035779 0.045452 0.020394 0.029829
0.025985 0.039284 0.032724 0.035878 0.036648 0.032631 0.027586 0.033583 0.036122
0.034311 0.034743 0.032915 0.030732 0.033625 0.031421 0.035579 0.033875 0.032380
0.037153 0.036674 0.033326 0.035358 0.038967 0.030531 0.036984 0.033895 0.032095
0.035843 0.034096 0.037166 0.036225 0.034946 0.037438 0.038645 0.037014 0.040294
0.041818 0.035503 0.032435 0.031618 0.030550 0.033758 0.031029 0.033472 0.035532
0.045291 0.031928 0.031183 0.035775 0.039135 0.035308 0.040937 0.031347 0.034018
0.026738 0.039435 0.040029 0.038224 0.033185 0.033635 0.034372 0.045638 0.034868
0.036818 0.030602 0.034357 0.034720 0.036413 0.035126 0.030441 0.035831 0.040428
Number of relative response measurements per bin:
67 83 104 109 97 79 64 39 44
32 32 42 68 63 72 45 39 29
53 72 113 88 61 85 32 59 24
39 55 80 106 120 95 75 62 46
20 40 62 47 61 47 44 33 17
37 72 83 109 134 81 60 60 23
6 54 68 61 76 75 70 53 38
32 57 74 79 100 66 51 13 26
57 66 84 112 95 82 64 53 41
1-sigma uncertainty [~RMS/sqrt(N)] in relative response per bin:
0.003616 0.004289 0.003650 0.003479 0.003224 0.004025 0.005681 0.003266 0.004497
0.004593 0.006944 0.005049 0.004351 0.004617 0.003846 0.004112 0.005378 0.006708
0.004713 0.004095 0.003096 0.003276 0.004305 0.003408 0.006289 0.004410 0.006609
0.005949 0.004945 0.003726 0.003434 0.003557 0.003132 0.004271 0.004305 0.004732
0.008015 0.005391 0.004720 0.005284 0.004474 0.005461 0.005826 0.006443 0.009773
0.006875 0.004184 0.003560 0.003028 0.002639 0.003751 0.004006 0.004321 0.007409
0.018490 0.004345 0.003781 0.004580 0.004489 0.004077 0.004893 0.004306 0.005518
0.004727 0.005223 0.004653 0.004301 0.003319 0.004140 0.004813 0.012658 0.006838
0.004877 0.003767 0.003749 0.003281 0.003736 0.003879 0.003805 0.004922 0.006314
See below for JPGs.
Number of unique input sources read = 142193 Number of bins: 9 x 9. Minimum number of same-source apparitions across frame overlaps: 11 S/N >= 5 Chi-square <= 2.0 WPRO mag. range = 10.1 - 12.2 Active deblend flag = 0 Number of components in fit = 1 Match radius = 2.0 arcsec Relative response per bin: 0.977958 0.964801 0.963639 0.980862 0.974556 1.013548 0.994642 0.982932 0.938179 1.015514 0.991253 0.987647 1.019227 1.003014 0.997045 1.009229 0.961307 0.977089 0.931742 1.006804 0.992065 0.994190 1.011616 1.029603 1.027410 1.030598 0.977791 0.991338 1.027469 1.011519 1.008330 1.002509 1.030899 1.019055 0.976556 0.969782 1.025081 1.022607 1.026095 1.026545 0.982601 0.999129 0.976331 0.977900 1.007953 0.979270 1.004327 0.986122 1.027005 0.993856 0.987256 1.029190 1.006877 0.971386 0.955846 0.962692 1.024646 0.986020 1.039157 1.004525 0.961898 0.959825 1.069332 0.986202 1.017373 1.005943 1.025336 1.004628 1.021605 0.991354 1.018200 1.033490 1.010395 0.989985 0.995201 0.973727 1.006513 1.035413 0.981661 0.985138 0.945004 RMS in relative response per bin: 0.132712 0.147257 0.123700 0.152936 0.120173 0.121429 0.133201 0.145957 0.138930 0.120513 0.134634 0.136113 0.147055 0.122759 0.121044 0.112193 0.117604 0.153034 0.115034 0.152796 0.127451 0.134787 0.130039 0.121444 0.128364 0.128402 0.136916 0.162901 0.141435 0.138354 0.121817 0.142279 0.120662 0.155142 0.136399 0.100729 0.175705 0.103478 0.141917 0.143535 0.142738 0.121935 0.114517 0.129838 0.175648 0.159914 0.133294 0.133014 0.115434 0.127312 0.129093 0.091622 0.169810 0.114824 0.149667 0.126948 0.138927 0.136509 0.114333 0.118182 0.117996 0.144294 0.144754 0.145520 0.105048 0.147202 0.133136 0.123439 0.108826 0.128181 0.127038 0.122884 0.127013 0.131322 0.107674 0.125932 0.123603 0.144302 0.135681 0.106427 0.143233 Number of relative response measurements per bin: 55 117 112 119 176 133 111 75 26 26 67 102 135 142 104 69 41 41 40 73 104 150 165 95 119 59 27 43 91 111 191 123 101 93 88 48 23 78 81 113 101 126 88 61 14 55 112 155 160 143 139 126 77 50 48 77 106 106 101 192 88 59 25 50 63 101 117 137 120 72 57 29 52 101 78 135 178 136 116 70 25 1-sigma uncertainty [~RMS/sqrt(N)] in relative response per bin: 0.017895 0.013614 0.011689 0.014020 0.009058 0.010529 0.012643 0.016854 0.027247 0.023634 0.016448 0.013477 0.012657 0.010302 0.011869 0.013506 0.018367 0.023900 0.018189 0.017883 0.012498 0.011005 0.010124 0.012460 0.011767 0.016716 0.026350 0.024842 0.014826 0.013132 0.008814 0.012829 0.012006 0.016088 0.014540 0.014539 0.036637 0.011717 0.015769 0.013503 0.014203 0.010863 0.012208 0.016624 0.046944 0.021563 0.012595 0.010684 0.009126 0.010646 0.010950 0.008162 0.019352 0.016239 0.021603 0.014467 0.013494 0.013259 0.011377 0.008529 0.012578 0.018785 0.028951 0.020580 0.013235 0.014647 0.012308 0.010546 0.009934 0.015106 0.016827 0.022819 0.017613 0.013067 0.012192 0.010839 0.009264 0.012374 0.012598 0.012720 0.028647
See below for JPGs.
Number of unique input sources read = 34952 Number of bins: 7 x 7. Minimum number of same-source apparitions across frame overlaps: 10 S/N >= 5 Chi-square <= 2.0 WPRO mag. range = 6.5 - 10 Active deblend flag = 0 Number of components in fit = 1 Match radius = 2.0 arcsec Relative response per bin: 0.979094 1.078270 0.977708 1.046372 1.057893 0.978492 1.018834 1.103107 0.976496 0.981455 0.998970 1.022223 0.998226 0.999296 0.994185 1.042635 0.961825 1.076097 1.007534 1.063388 0.984408 1.012210 0.989702 0.999155 1.014389 1.046743 0.932342 0.911926 1.017011 0.944194 1.004247 1.036322 1.039978 1.089607 0.929988 0.937583 0.957196 1.005775 0.957997 0.960850 0.974238 1.010957 1.019822 0.961497 0.968366 0.992167 0.954545 0.991657 0.952810 RMS in relative response per bin: 0.153404 0.147310 0.143148 0.158240 0.143192 0.127173 0.115803 0.165973 0.142741 0.151959 0.169201 0.110466 0.119044 0.141887 0.139428 0.173352 0.106597 0.149387 0.135025 0.151243 0.128388 0.126175 0.131869 0.121535 0.133839 0.130769 0.144919 0.103397 0.163693 0.114906 0.148573 0.144224 0.151564 0.186833 0.113204 0.132658 0.124882 0.135607 0.128141 0.124035 0.154048 0.126894 0.137392 0.134787 0.139202 0.141021 0.113253 0.128821 0.171241 Number of relative response measurements per bin: 23 46 67 66 51 28 33 33 68 77 35 59 60 37 46 34 55 36 89 59 18 32 54 41 62 64 39 18 46 44 44 82 59 35 26 26 53 54 95 47 47 29 28 57 85 73 47 52 42 1-sigma uncertainty [~RMS/sqrt(N)] in relative response per bin: 0.031987 0.021720 0.017488 0.019478 0.020051 0.024033 0.020159 0.028892 0.017310 0.017317 0.028600 0.014381 0.015369 0.023326 0.020558 0.029730 0.014374 0.024898 0.014313 0.019690 0.030261 0.022305 0.017945 0.018981 0.016998 0.016346 0.023206 0.024371 0.024135 0.017323 0.022398 0.015927 0.019732 0.031581 0.022201 0.026016 0.017154 0.018454 0.013147 0.018092 0.022470 0.023564 0.025965 0.017853 0.015099 0.016505 0.016520 0.017864 0.026423
W1 = top left; W2 = top right; W3 = bottom left; W4 = bottom right.
See color bar for scale. All are at same stretch.
 |
Gaussian-smoothed versions of above:
 |
W1 = top left; W2 = top right; W3 = bottom left; W4 = bottom right.
See color bar for scale. All are at same stretch.
 |
W1 = top left; W2 = top right; W3 = bottom left; W4 = bottom right.
See color bar for scale. All are at same stretch.
 |
W1 = top left; W2 = top right; W3 = bottom left; W4 = bottom right.
See color bar for scale. All are at same stretch.
 |