jwst-gc pipeline monitor

Stage ladder

Products per filter

* = the product name carries no _o<obs> token and this field has more than one observation, so the count cannot be attributed to 6778/o001.

Astrometry — m2 checkpoint

“tiles ok” is not a tolerance. measure_offset_grid runs with no max_off_mas, and astrometry_offsets sets off_ok=True whenever that is None — so N/N counts tiles whose offset histogram had a coherent peak, however large the offset. The column that carries the gate is worst tile, against 15 mas.

Provenance

No *.prov.json sidecars.

Queue

Findings

• fail
  F090W: m2 built no visit consensus — 3 visit(s) attempted, 0 exposures ingested

  The checkpoint ran and recorded `passed: false` without measuring a single exposure, so the filter has NO m2 astrometry: no exposure was compared with its visit consensus and no tie to the reference frame was made. Every per-exposure count on this card is 0 because the list is empty, which is a different state from "all exposures agree".

  checkpoint_m2_F090W_latest.json

  what is affected, and why

  The usual cause is that one physical (visit, exposure, module, filter, vgroup) identity resolves to more than one per-frame catalog, and build_visit_consensus refuses rather than picking one — correctly, since picking one would tie the field to whichever it happened to read. That happens where tokened and untokened catalogs sit in one directory, or where grouped-fit (`_group_`) or stale module-level variants sit alongside the per-detector ones. Re-catalog the filter at m2, or retokenise the untokened catalogs, then re-run the checkpoint.

  Diagnostic writeupfull diagnostic writeup (PDF)

  recorded failure
  ngc6334 F090W/nrcb F090W visit 1 [m2]: duplicate exposure identity: visit consensus (ngc6334 F090W/nrcb F090W visit 1 [m2]): 20 exposure identity/ies ingested more than once -- duplicate per-frame catalogs for the same (visit, exposure, module, filter, vgroup): [(('1', 1, 'nrcb1', 'F090W', '02103'), 2), (('1', 1, 'nrcb2', 'F090W', '02103'), 2), (('1', 1, 'nrcb3', 'F090W', '02103'), 2), (('1', 1, 'nrcb4', 'F090W', '02103'), 2), (('1', 2, 'nrcb1', 'F090W', '02103'), 2), (('1', 2, 'nrcb2', 'F090W', '02103'), 2), (('1', 2, 'nrcb3', 'F090W', '02103'), 2), (('1', 2, 'nrcb4', 'F090W', '02103'), 2)]. Check for grouped-fit (`_group_`) or stale module-level variants alongside the per-detector ones.
  ngc6334 F090W/nrcb F090W visit 2 [m2]: duplicate exposure identity: visit consensus (ngc6334 F090W/nrcb F090W visit 2 [m2]): 20 exposure identity/ies ingested more than once -- duplicate per-frame catalogs for the same (visit, exposure, module, filter, vgroup): [(('2', 1, 'nrcb1', 'F090W', '02103'), 2), (('2', 1, 'nrcb2', 'F090W', '02103'), 2), (('2', 1, 'nrcb3', 'F090W', '02103'), 2), (('2', 1, 'nrcb4', 'F090W', '02103'), 2), (('2', 2, 'nrcb1', 'F090W', '02103'), 2), (('2', 2, 'nrcb2', 'F090W', '02103'), 2), (('2', 2, 'nrcb3', 'F090W', '02103'), 2), (('2', 2, 'nrcb4', 'F090W', '02103'), 2)]. Check for grouped-fit (`_group_`) or stale module-level variants alongside the per-detector ones.
  ngc6334 F090W/nrcb F090W visit 3 [m2]: duplicate exposure identity: visit consensus (ngc6334 F090W/nrcb F090W visit 3 [m2]): 20 exposure identity/ies ingested more than once -- duplicate per-frame catalogs for the same (visit, exposure, module, filter, vgroup): [(('3', 1, 'nrcb1', 'F090W', '02103'), 2), (('3', 1, 'nrcb2', 'F090W', '02103'), 2), (('3', 1, 'nrcb3', 'F090W', '02103'), 2), (('3', 1, 'nrcb4', 'F090W', '02103'), 2), (('3', 2, 'nrcb1', 'F090W', '02103'), 2), (('3', 2, 'nrcb2', 'F090W', '02103'), 2), (('3', 2, 'nrcb3', 'F090W', '02103'), 2), (('3', 2, 'nrcb4', 'F090W', '02103'), 2)]. Check for grouped-fit (`_group_`) or stale module-level variants alongside the per-detector ones.

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F090W_j6778 visit 2: consensus not formed

  The exposures of this visit could not be brought into a common frame, so nothing downstream of m2 is trustworthy for it.

  checkpoint_m2_F090W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F090W_j6778 visit 3: consensus not formed

  The exposures of this visit could not be brought into a common frame, so nothing downstream of m2 is trustworthy for it.

  checkpoint_m2_F090W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F090W_j6778: lowest offset-peak contrast 2.00 < 5

  A low peak contrast means the offset histogram found no real tie. It does NOT mean "no offset": a shift larger than the search window has zero true pairs inside it, so a gross misalignment reads as low contrast (the brick-1182 v001 signature).

  checkpoint_m2_F090W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupfull diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F090W_j6778: 30/52 exposures misaligned vs their visit consensus

  An exposure is misaligned when its offset exceeds 2.0 mas AND is significant against the peak error bars. At m2 this CORRECTS the offsets table and stops the run; the crf frames must be regenerated before cataloging continues.

  checkpoint_m2_F090W_j6778_latest.json

  what is affected, and why

  Affected: 7/7 detectors — spread across detectors, so the frame as a whole moved: look at the offsets table and whether these exposures were regenerated from _cal after the table changed. fix_alignment SKIPS a frame that already has a RAOFFSET header, so correcting the table alone leaves stale frames in place; regenerate the working copy from _cal instead of re-applying on top.

  Misaligned by detector

  nrca17/10

  nrca27/10

  nrcb35/9

  nrcb45/9

  nrca33/5

  nrca42/6

  nrcb11/3

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  visit	detector	dRA (mas)	dDec (mas)	off (mas)	contrast	pairs	window ("")
  1	nrca1	-10.7	-1.5	10.8	14.0	115	10.0
  1	nrca1	-1.5	-1.9	2.4	20.0	265	10.0
  1	nrca1	1.7	2.8	3.3	27.0	309	10.0
  1	nrca2	-12.6	0.8	12.7	11.0	106	10.0
  1	nrca2	2.0	1.6	2.5	47.0	774	10.0
  1	nrca3	-1.5	-7.4	7.5	5.0	35	3.0
  1	nrcb4	-7.7	1.2	7.8	29.0	290	10.0
  1	nrcb4	-6.4	1.5	6.6	25.0	238	10.0
  1	nrcb4	5.6	2.0	6.0	41.0	685	10.0
  1	nrcb4	5.9	-0.8	5.9	33.0	590	10.0
  2	nrca2	-8.0	-5.6	9.8	27.0	232	10.0
  2	nrca2	4.0	-1.8	4.4	27.0	305	10.0
  2	nrca2	5.6	-0.3	5.6	21.0	329	10.0
  2	nrca3	-2.9	-3.3	4.4	41.0	586	10.0
  2	nrca3	4.8	4.0	6.2	27.0	396	10.0
  2	nrcb3	-10.6	-6.3	12.3	17.0	214	10.0
  2	nrcb3	4.7	0.4	4.8	20.0	303	10.0
  3	nrca1	8.1	5.1	9.5	16.0	317	10.0
  3	nrca1	-3.5	-0.5	3.6	22.0	365	10.0
  3	nrca1	1.4	3.1	3.4	26.0	327	10.0
  3	nrca1	2.2	2.5	3.3	27.0	370	10.0
  3	nrca2	-6.5	-0.1	6.5	21.0	485	10.0
  3	nrca2	11.0	5.7	12.4	33.0	488	10.0
  3	nrca4	-10.9	1.3	11.0	25.0	385	10.0
  3	nrca4	8.8	4.5	9.8	25.0	362	10.0
  3	nrcb1	-7.3	-1.8	7.5	14.0	310	10.0
  3	nrcb3	-9.1	2.1	9.4	39.0	1193	10.0
  3	nrcb3	8.5	5.3	10.1	55.0	1057	10.0
  3	nrcb3	22.1	-13.3	25.8	6.0	255	10.0
  3	nrcb4	5.1	4.6	6.9	5.0	64	10.0

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F187N_j6778: 76/120 exposures misaligned vs their visit consensus

  An exposure is misaligned when its offset exceeds 2.0 mas AND is significant against the peak error bars. At m2 this CORRECTS the offsets table and stops the run; the crf frames must be regenerated before cataloging continues.

  checkpoint_m2_F187N_j6778_latest.json

  what is affected, and why

  Affected: 8/8 detectors — spread across detectors, so the frame as a whole moved: look at the offsets table and whether these exposures were regenerated from _cal after the table changed. fix_alignment SKIPS a frame that already has a RAOFFSET header, so correcting the table alone leaves stale frames in place; regenerate the working copy from _cal instead of re-applying on top.

  Misaligned by detector

  nrca412/15

  nrcb312/15

  nrca110/15

  nrcb110/15

  nrcb210/15

  nrcb49/15

  nrca37/15

  nrca26/15

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  visit	detector	dRA (mas)	dDec (mas)	off (mas)	contrast	pairs	window ("")
  1	nrca1	-8.1	0.2	8.1	241.0	2180	3.0
  1	nrca1	-6.4	0.6	6.5	303.0	18204	10.0
  1	nrca1	-1.8	-1.0	2.0	388.0	23446	10.0
  1	nrca2	-4.9	0.2	4.9	278.0	26961	10.0
  1	nrca2	-3.1	0.4	3.1	328.0	32730	10.0
  1	nrca3	-6.7	-0.7	6.8	350.0	19922	10.0
  1	nrca4	-3.5	0.0	3.5	117.0	6881	10.0
  1	nrca4	2.5	-1.1	2.8	325.0	14024	10.0
  1	nrca4	5.2	0.5	5.2	308.0	17237	10.0
  1	nrca4	5.3	-1.1	5.4	419.0	20451	10.0
  1	nrcb1	3.2	-1.4	3.5	83.0	1017	10.0
  1	nrcb1	3.0	0.4	3.0	72.0	1647	10.0
  1	nrcb1	2.8	-0.8	2.9	78.0	1417	10.0
  1	nrcb2	-4.7	-0.5	4.7	69.0	210	3.0
  1	nrcb2	-3.2	-0.2	3.3	62.0	1498	10.0
  1	nrcb2	3.3	0.8	3.4	162.0	1130	3.0
  1	nrcb2	3.0	-1.2	3.2	167.0	7023	10.0
  1	nrcb3	-7.2	-0.1	7.2	76.0	2743	10.0
  1	nrcb3	-5.5	0.1	5.5	79.0	2448	10.0
  1	nrcb3	2.2	2.4	3.2	214.0	749	3.0
  1	nrcb3	2.1	1.3	2.5	185.0	661	3.0
  1	nrcb4	-6.1	0.7	6.1	311.0	21657	10.0
  1	nrcb4	-4.5	0.5	4.5	307.0	22584	10.0
  1	nrcb4	1.6	1.5	2.1	635.0	53689	10.0
  2	nrca1	-2.7	-1.2	2.9	122.0	505	3.0
  2	nrca1	-6.0	-5.2	8.0	178.0	5403	10.0
  2	nrca1	-1.0	2.1	2.3	201.0	7764	10.0
  2	nrca1	2.7	-1.0	2.9	277.0	13745	10.0
  2	nrca1	4.3	0.8	4.3	317.0	19057	10.0
  2	nrca2	-3.1	-1.1	3.3	428.0	3047	3.0
  2	nrca2	-2.6	2.1	3.3	629.0	45149	10.0
  2	nrca3	-3.3	-4.0	5.2	271.0	7406	10.0
  2	nrca3	0.6	3.5	3.6	130.0	442	3.0
  2	nrca3	2.4	-0.3	2.4	84.0	1971	10.0
  2	nrca3	3.7	0.7	3.8	113.0	2413	10.0
  2	nrca4	-1.2	-1.7	2.1	205.0	676	3.0
  2	nrca4	-4.5	-6.0	7.5	208.0	692	3.0
  2	nrca4	-0.3	2.0	2.0	198.0	8238	10.0
  2	nrca4	1.9	-1.3	2.2	337.0	14791	10.0
  2	nrca4	3.3	0.3	3.4	279.0	16660	10.0

  showing 40 of 76

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F200W_j6778: 73/120 exposures misaligned vs their visit consensus

  An exposure is misaligned when its offset exceeds 2.0 mas AND is significant against the peak error bars. At m2 this CORRECTS the offsets table and stops the run; the crf frames must be regenerated before cataloging continues.

  checkpoint_m2_F200W_j6778_latest.json

  what is affected, and why

  Affected: 8/8 detectors — spread across detectors, so the frame as a whole moved: look at the offsets table and whether these exposures were regenerated from _cal after the table changed. fix_alignment SKIPS a frame that already has a RAOFFSET header, so correcting the table alone leaves stale frames in place; regenerate the working copy from _cal instead of re-applying on top.

  Misaligned by detector

  nrca211/15

  nrca311/15

  nrcb110/15

  nrcb210/15

  nrca49/15

  nrcb39/15

  nrcb47/15

  nrca16/15

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  visit	detector	dRA (mas)	dDec (mas)	off (mas)	contrast	pairs	window ("")
  1	nrca1	-5.9	0.8	5.9	1,240.0	41350	3.0
  1	nrca1	-3.8	0.7	3.9	1,249.0	38958	3.0
  1	nrca2	-7.5	0.9	7.6	2,129.0	67260	3.0
  1	nrca2	-5.7	1.5	5.9	2,698.0	87979	3.0
  1	nrca2	2.1	-0.8	2.3	1,334.0	218897	3.0
  1	nrca3	-2.8	-0.4	2.8	1,585.0	49636	3.0
  1	nrca3	3.5	-2.2	4.1	1,418.0	40142	3.0
  1	nrca3	4.1	0.3	4.1	1,046.0	39381	3.0
  1	nrca3	4.5	-1.1	4.6	1,256.0	36669	3.0
  1	nrca4	-7.5	0.8	7.5	910.0	18161	3.0
  1	nrca4	-1.7	-1.1	2.0	1,513.0	36251	3.0
  1	nrcb1	-6.3	1.7	6.5	192.0	4900	10.0
  1	nrcb1	-4.4	1.2	4.5	195.0	5927	10.0
  1	nrcb1	1.8	-1.0	2.1	230.0	7695	10.0
  1	nrcb1	3.0	2.3	3.8	318.0	15623	10.0
  1	nrcb1	3.0	-0.0	3.0	183.0	16982	10.0
  1	nrcb2	-2.4	-0.7	2.5	122.0	2849	10.0
  1	nrcb2	3.5	-1.8	3.9	202.0	6679	10.0
  1	nrcb2	3.9	2.0	4.4	749.0	15382	3.0
  1	nrcb2	3.9	-0.7	3.9	539.0	12284	3.0
  1	nrcb3	-7.2	-0.4	7.2	330.0	39297	10.0
  1	nrcb3	-5.7	-0.4	5.7	341.0	42046	10.0
  1	nrcb3	1.7	3.9	4.3	994.0	13352	3.0
  1	nrcb3	1.9	1.6	2.5	869.0	11450	3.0
  1	nrcb4	-6.3	-0.1	6.3	872.0	40510	3.0
  1	nrcb4	-4.6	-0.7	4.7	1,171.0	37911	3.0
  1	nrcb4	0.9	1.9	2.1	1,392.0	101429	3.0
  2	nrca1	-4.9	-4.2	6.5	739.0	14154	3.0
  2	nrca1	-0.6	2.2	2.3	636.0	18335	3.0
  2	nrca2	-2.8	-0.5	2.8	1,326.0	53820	3.0
  2	nrca2	-5.9	-3.6	6.9	2,143.0	54591	3.0
  2	nrca2	-2.1	1.6	2.6	1,018.0	93026	3.0
  2	nrca3	-5.9	-3.0	6.6	521.0	30387	10.0
  2	nrca3	0.6	-2.4	2.5	273.0	36110	10.0
  2	nrca3	3.0	-1.6	3.4	395.0	45659	10.0
  2	nrca4	-3.7	-2.9	4.7	1,025.0	12596	3.0
  2	nrca4	0.7	3.1	3.2	821.0	22068	3.0
  2	nrca4	2.4	1.2	2.7	1,294.0	36773	3.0
  2	nrcb1	-4.7	-4.6	6.6	94.0	1333	10.0
  2	nrcb1	-2.2	0.7	2.3	613.0	13600	3.0

  showing 40 of 73

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F200W_j7213: 123/158 exposures misaligned vs their visit consensus

  An exposure is misaligned when its offset exceeds 2.0 mas AND is significant against the peak error bars. At m2 this CORRECTS the offsets table and stops the run; the crf frames must be regenerated before cataloging continues.

  checkpoint_m2_F200W_j7213_latest.json

  what is affected, and why

  Affected: 8/8 detectors — spread across detectors, so the frame as a whole moved: look at the offsets table and whether these exposures were regenerated from _cal after the table changed. fix_alignment SKIPS a frame that already has a RAOFFSET header, so correcting the table alone leaves stale frames in place; regenerate the working copy from _cal instead of re-applying on top.

  Misaligned by detector

  nrcb417/20

  nrca116/18

  nrca316/20

  nrcb216/20

  nrca415/20

  nrcb115/20

  nrca214/20

  nrcb314/20

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  visit	detector	dRA (mas)	dDec (mas)	off (mas)	contrast	pairs	window ("")
  1	nrca1	-4.8	4.4	6.5	52.0	479	10.0
  1	nrca1	-5.8	4.1	7.1	54.0	500	10.0
  1	nrca1	-2.6	-1.5	3.0	40.0	399	10.0
  1	nrca1	-2.2	-1.8	2.8	40.0	399	10.0
  1	nrca1	-5.0	-1.9	5.3	39.0	373	10.0
  1	nrca1	-3.2	-1.4	3.4	39.0	369	10.0
  1	nrca1	6.3	2.5	6.7	59.0	560	10.0
  1	nrca1	4.4	3.0	5.3	58.0	140	3.0
  1	nrca2	-2.5	3.8	4.6	823.0	19802	3.0
  1	nrca2	-3.4	3.9	5.2	828.0	19969	3.0
  1	nrca2	-4.2	-2.8	5.1	1,256.0	32003	3.0
  1	nrca2	-2.7	-2.7	3.8	1,226.0	30693	3.0
  1	nrca2	-1.9	0.7	2.0	1,325.0	34862	3.0
  1	nrca3	-1.4	3.5	3.7	143.0	3215	10.0
  1	nrca3	-3.4	2.7	4.4	141.0	3173	10.0
  1	nrca3	-1.8	-5.2	5.5	60.0	571	10.0
  1	nrca3	0.0	-5.0	5.0	33.0	561	10.0
  1	nrca3	4.0	-0.4	4.0	27.0	401	10.0
  1	nrca3	2.3	0.4	2.4	32.0	429	10.0
  1	nrca3	2.9	-0.9	3.0	30.0	378	10.0
  1	nrca4	-1.7	6.3	6.5	980.0	27031	3.0
  1	nrca4	-3.6	5.8	6.8	985.0	27117	3.0
  1	nrca4	-0.3	-5.5	5.5	712.0	27775	3.0
  1	nrca4	1.5	-5.4	5.6	968.0	27144	3.0
  1	nrca4	5.3	-2.3	5.8	672.0	15333	3.0
  1	nrca4	6.3	-2.3	6.7	1,206.0	34046	3.0
  1	nrca4	3.8	-1.6	4.1	1,160.0	34962	3.0
  1	nrcb1	-1.3	4.7	4.8	614.0	10684	3.0
  1	nrcb1	-3.6	4.2	5.5	615.0	10589	3.0
  1	nrcb1	0.7	-7.0	7.1	325.0	59768	10.0
  1	nrcb1	3.2	-6.8	7.5	410.0	61257	10.0
  1	nrcb1	7.5	-2.3	7.9	123.0	2559	10.0
  1	nrcb1	5.7	-1.6	5.9	130.0	2989	10.0
  1	nrcb1	0.8	3.8	3.9	140.0	5767	10.0
  1	nrcb1	5.7	-0.8	5.7	138.0	5508	10.0
  1	nrcb2	-0.5	2.9	3.0	831.0	21144	3.0
  1	nrcb2	-2.8	2.3	3.6	976.0	21163	3.0
  1	nrcb2	1.0	-8.6	8.7	189.0	15554	10.0
  1	nrcb2	3.4	-8.3	8.9	241.0	14578	10.0
  1	nrcb2	6.1	-2.6	6.7	164.0	4126	10.0

  showing 40 of 123

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F277W_j6778: 17/30 exposures misaligned vs their visit consensus

  An exposure is misaligned when its offset exceeds 2.0 mas AND is significant against the peak error bars. At m2 this CORRECTS the offsets table and stops the run; the crf frames must be regenerated before cataloging continues.

  checkpoint_m2_F277W_j6778_latest.json

  what is affected, and why

  Affected: 2/2 detectors — spread across detectors, so the frame as a whole moved: look at the offsets table and whether these exposures were regenerated from _cal after the table changed. fix_alignment SKIPS a frame that already has a RAOFFSET header, so correcting the table alone leaves stale frames in place; regenerate the working copy from _cal instead of re-applying on top.

  Misaligned by detector

  nrcalong10/15

  nrcblong7/15

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  visit	detector	dRA (mas)	dDec (mas)	off (mas)	contrast	pairs	window ("")
  1	nrcalong	-7.9	0.2	7.9	1,587.0	31510	3.0
  1	nrcalong	-6.5	0.9	6.6	2,324.0	35785	3.0
  1	nrcalong	3.0	1.5	3.3	3,372.0	57198	3.0
  1	nrcalong	3.0	-1.1	3.1	3,380.0	61629	3.0
  1	nrcblong	-7.6	-0.1	7.6	671.0	75010	10.0
  1	nrcblong	-0.8	-2.1	2.3	1,124.0	12655	3.0
  1	nrcblong	1.6	-1.5	2.2	1,538.0	16254	3.0
  2	nrcalong	-2.5	-2.1	3.3	1,635.0	17983	3.0
  2	nrcalong	-5.8	-5.5	8.0	1,866.0	19197	3.0
  2	nrcalong	3.2	-0.5	3.3	1,868.0	33888	3.0
  2	nrcalong	4.4	1.3	4.6	2,394.0	34663	3.0
  2	nrcblong	-2.8	-1.7	3.3	1,207.0	11625	3.0
  2	nrcblong	1.8	1.4	2.3	1,077.0	83772	10.0
  3	nrcalong	6.3	4.0	7.4	2,621.0	32080	3.0
  3	nrcalong	-3.9	-2.9	4.8	2,548.0	32139	3.0
  3	nrcblong	5.3	2.7	5.9	3,050.0	50741	3.0
  3	nrcblong	-5.8	-3.7	6.8	3,907.0	63066	3.0

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F335M_j6778: 16/30 exposures misaligned vs their visit consensus

  An exposure is misaligned when its offset exceeds 2.0 mas AND is significant against the peak error bars. At m2 this CORRECTS the offsets table and stops the run; the crf frames must be regenerated before cataloging continues.

  checkpoint_m2_F335M_j6778_latest.json

  what is affected, and why

  Affected: 2/2 detectors — spread across detectors, so the frame as a whole moved: look at the offsets table and whether these exposures were regenerated from _cal after the table changed. fix_alignment SKIPS a frame that already has a RAOFFSET header, so correcting the table alone leaves stale frames in place; regenerate the working copy from _cal instead of re-applying on top.

  Misaligned by detector

  nrcblong11/15

  nrcalong5/15

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  visit	detector	dRA (mas)	dDec (mas)	off (mas)	contrast	pairs	window ("")
  1	nrcalong	-7.6	0.1	7.6	718.0	61890	10.0
  1	nrcalong	-6.1	-0.1	6.1	789.0	73105	10.0
  1	nrcalong	1.1	1.8	2.1	1,616.0	14205	3.0
  1	nrcblong	-6.9	0.7	7.0	482.0	12854	10.0
  1	nrcblong	-5.0	0.5	5.1	482.0	14639	10.0
  1	nrcblong	2.4	2.4	3.4	710.0	18373	10.0
  1	nrcblong	3.0	-0.2	3.0	460.0	20561	10.0
  2	nrcalong	-2.9	-2.5	3.9	746.0	24078	10.0
  2	nrcblong	-1.6	-1.6	2.2	652.0	21243	10.0
  2	nrcblong	-5.3	-5.6	7.7	790.0	21567	10.0
  2	nrcblong	3.8	0.4	3.8	555.0	27298	10.0
  3	nrcalong	-2.9	-4.8	5.6	1,188.0	53027	10.0
  3	nrcblong	-6.6	1.3	6.7	1,239.0	80240	10.0
  3	nrcblong	-0.8	-3.0	3.1	1,358.0	10958	3.0
  3	nrcblong	2.0	1.4	2.4	1,604.0	11522	3.0
  3	nrcblong	4.1	0.6	4.1	1,218.0	10911	3.0

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F470N_j6778: 20/30 exposures misaligned vs their visit consensus

  An exposure is misaligned when its offset exceeds 2.0 mas AND is significant against the peak error bars. At m2 this CORRECTS the offsets table and stops the run; the crf frames must be regenerated before cataloging continues.

  checkpoint_m2_F470N_j6778_latest.json

  what is affected, and why

  Affected: 2/2 detectors — spread across detectors, so the frame as a whole moved: look at the offsets table and whether these exposures were regenerated from _cal after the table changed. fix_alignment SKIPS a frame that already has a RAOFFSET header, so correcting the table alone leaves stale frames in place; regenerate the working copy from _cal instead of re-applying on top.

  Misaligned by detector

  nrcalong11/15

  nrcblong9/15

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  visit	detector	dRA (mas)	dDec (mas)	off (mas)	contrast	pairs	window ("")
  1	nrcalong	-7.8	-0.8	7.9	1,461.0	89148	10.0
  1	nrcalong	-6.1	-0.0	6.1	983.0	10516	3.0
  1	nrcalong	2.9	-0.7	3.0	1,757.0	14671	3.0
  1	nrcalong	3.3	-2.0	3.9	2,214.0	15619	3.0
  1	nrcblong	-5.8	1.1	5.9	1,065.0	35228	10.0
  1	nrcblong	-4.4	1.2	4.6	1,075.0	37294	10.0
  1	nrcblong	2.5	2.3	3.3	1,295.0	45941	10.0
  1	nrcblong	2.3	0.5	2.4	965.0	48196	10.0
  2	nrcalong	-2.4	-1.7	2.9	1,478.0	67304	10.0
  2	nrcalong	-5.5	-5.4	7.8	1,624.0	67744	10.0
  2	nrcalong	3.6	-2.1	4.1	1,746.0	83680	10.0
  2	nrcalong	5.4	0.1	5.4	954.0	83964	10.0
  2	nrcblong	-3.9	-4.5	6.0	1,481.0	57068	10.0
  2	nrcblong	0.1	3.4	3.4	929.0	77045	10.0
  2	nrcblong	2.2	0.8	2.3	1,519.0	89290	10.0
  2	nrcblong	4.0	2.3	4.6	1,763.0	86200	10.0
  3	nrcalong	6.1	3.0	6.8	1,753.0	84424	10.0
  3	nrcalong	-3.7	-6.3	7.3	1,829.0	84719	10.0
  3	nrcalong	-0.3	-2.2	2.2	999.0	74724	10.0
  3	nrcblong	-2.7	-3.8	4.6	2,031.0	12246	3.0

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F470N_j7213: 31/40 exposures misaligned vs their visit consensus

  An exposure is misaligned when its offset exceeds 2.0 mas AND is significant against the peak error bars. At m2 this CORRECTS the offsets table and stops the run; the crf frames must be regenerated before cataloging continues.

  checkpoint_m2_F470N_j7213_latest.json

  what is affected, and why

  Affected: 2/2 detectors — spread across detectors, so the frame as a whole moved: look at the offsets table and whether these exposures were regenerated from _cal after the table changed. fix_alignment SKIPS a frame that already has a RAOFFSET header, so correcting the table alone leaves stale frames in place; regenerate the working copy from _cal instead of re-applying on top.

  Misaligned by detector

  nrcblong16/20

  nrcalong15/20

  Diagnostic writeupD2 — internal astrometric repeatability · full diagnostic writeup (PDF)

  visit	detector	dRA (mas)	dDec (mas)	off (mas)	contrast	pairs	window ("")
  1	nrcalong	-1.6	1.8	2.4	501.0	1907	3.0
  1	nrcalong	-2.6	5.1	5.7	632.0	2181	3.0
  1	nrcalong	-4.5	4.7	6.5	675.0	16315	10.0
  1	nrcalong	-2.2	-4.1	4.7	536.0	13210	10.0
  1	nrcalong	-0.3	-4.1	4.1	350.0	13615	10.0
  1	nrcalong	4.1	-0.9	4.2	451.0	12854	10.0
  1	nrcalong	2.7	-0.4	2.7	377.0	12894	10.0
  1	nrcalong	3.7	-0.5	3.7	422.0	15110	10.0
  1	nrcblong	-2.6	3.6	4.4	826.0	2580	3.0
  1	nrcblong	-3.6	3.4	5.0	831.0	18081	10.0
  1	nrcblong	-4.1	7.1	8.2	878.0	2769	3.0
  1	nrcblong	-2.7	-3.4	4.4	723.0	2221	3.0
  1	nrcblong	-0.0	-3.2	3.2	391.0	2250	3.0
  1	nrcblong	5.0	-1.3	5.2	704.0	2475	3.0
  1	nrcblong	3.4	-0.9	3.5	639.0	2530	3.0
  1	nrcblong	4.7	-0.1	4.7	423.0	18388	10.0
  1	nrcblong	3.1	-0.4	3.1	543.0	2615	3.0
  2	nrcalong	5.4	-2.2	5.8	750.0	22010	10.0
  2	nrcalong	2.8	-1.6	3.2	704.0	22572	10.0
  2	nrcalong	-3.1	-1.2	3.3	574.0	20146	10.0
  2	nrcalong	1.3	-2.4	2.8	583.0	2517	3.0
  2	nrcalong	-3.2	6.4	7.2	801.0	20880	10.0
  2	nrcalong	-1.9	4.8	5.2	736.0	20492	10.0
  2	nrcalong	-6.5	4.8	8.1	731.0	17155	10.0
  2	nrcblong	4.2	-1.9	4.6	995.0	29273	10.0
  2	nrcblong	1.6	-1.5	2.2	852.0	3706	3.0
  2	nrcblong	-3.0	-1.7	3.5	950.0	29619	10.0
  2	nrcblong	1.7	-2.9	3.4	946.0	29673	10.0
  2	nrcblong	-2.2	5.4	5.8	998.0	28865	10.0
  2	nrcblong	-1.2	3.9	4.1	890.0	29263	10.0
  2	nrcblong	-4.8	4.3	6.4	1,019.0	27128	10.0

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F187N_j6778 visit 1: weakest tile peak contrast 3.00 < 5

  At least one tile has no real tie, so its offset is noise.

  checkpoint_m2_F187N_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F200W_j6778 visit 2: weakest tile peak contrast 4.00 < 5

  At least one tile has no real tie, so its offset is noise.

  checkpoint_m2_F200W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F277W_j6778 visit 2: weakest tile peak contrast 4.00 < 5

  At least one tile has no real tie, so its offset is noise.

  checkpoint_m2_F277W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F335M_j6778 visit 1: weakest tile peak contrast 4.00 < 5

  At least one tile has no real tie, so its offset is noise.

  checkpoint_m2_F335M_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F335M_j6778 visit 2: weakest tile peak contrast 4.00 < 5

  At least one tile has no real tie, so its offset is noise.

  checkpoint_m2_F335M_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F335M_j6778 visit 3: weakest tile peak contrast 4.00 < 5

  At least one tile has no real tie, so its offset is noise.

  checkpoint_m2_F335M_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F470N_j6778 visit 1: weakest tile peak contrast 4.00 < 5

  At least one tile has no real tie, so its offset is noise.

  checkpoint_m2_F470N_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• fail
  F470N_j7213 visit 1: weakest tile peak contrast 3.00 < 5

  At least one tile has no real tie, so its offset is noise.

  checkpoint_m2_F470N_j7213_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  2 filter(s) have catalogs that cannot be attributed to 6778/o001: F200W, F470N

  These filters belong to more than one observation of this field, and the per-filter merged catalogs are written with no _o<obs> token, so the file on disk is whichever observation ran last. Counts for these filters describe the directory, not this observation.

  catalogs/

  what is affected, and why

  Diagnostic writeupD1 — field overview · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F187N_j6778 visit 1: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F187N_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F187N_j6778 visit 2: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F187N_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F187N_j6778 visit 3: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F187N_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F200W_j6778 visit 1: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F200W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F200W_j6778 visit 3: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F200W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F277W_j6778 visit 1: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F277W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F335M_j6778 visit 1: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F335M_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F470N_j6778 visit 1: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F470N_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F470N_j6778 visit 2: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F470N_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F470N_j7213 visit 1: reference tie recorded as could-not-verify (apply_ok=false)

  A correction is applied only when the tie is coherent AND the gross cross-check (100 mas) passes AND the per-tile map is clean. This one was not, so no correction was made — the frame keeps whatever it had.

  checkpoint_m2_F470N_j7213_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F090W_j6778: 3 exposure(s) could not be verified

  Recorded as could-not-verify: the reference tie was not coherent, or the gross cross-check (100 mas) failed, so no correction was applied.

  checkpoint_m2_F090W_j6778_latest.json

  what is affected, and why

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F090W_j6778 visit 1: worst tile 78.4 mas at cell (5,0) (> 15 mas)

  1/1 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F090W_j6778_latest.json

  what is affected, and why

  1/1 cells exceed 15 mas, and every one of them is on the mosaic EDGE — that is usually thin coverage (few exposures, so few pairs per cell) rather than a distortion error. Check the pair counts in the table before treating it as a defect. The bulk tie for this visit is 71.61 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (5,0)	78.4	-75.5	-21.2	19.0	6788

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F090W_j6778 visit 2: worst tile 93.8 mas at cell (1,3) (> 15 mas)

  1/1 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F090W_j6778_latest.json

  what is affected, and why

  1/1 cells exceed 15 mas, and every one of them is on the mosaic EDGE — that is usually thin coverage (few exposures, so few pairs per cell) rather than a distortion error. Check the pair counts in the table before treating it as a defect. The bulk tie for this visit is 70.58 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (1,3)	93.8	-91.1	-22.5	17.0	4632

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F090W_j6778 visit 3: worst tile 83.0 mas at cell (4,2) (> 15 mas)

  3/3 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F090W_j6778_latest.json

  what is affected, and why

  3/3 cells exceed 15 mas, and every one of them is on the mosaic EDGE — that is usually thin coverage (few exposures, so few pairs per cell) rather than a distortion error. Check the pair counts in the table before treating it as a defect. The bulk tie for this visit is 80.54 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (4,2)	83.0	-74.8	-36.1	25.0	16604
  (4,4)	82.4	-80.3	-18.5	23.0	18615
  (4,5)	75.5	-68.8	-31.1	25.0	5172

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F187N_j6778 visit 1: worst tile 22.0 mas at cell (1,5) (> 15 mas)

  31/33 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F187N_j6778_latest.json

  what is affected, and why

  5/33 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 0.43 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (1,5)	22.0	14.3	-16.8	9.0	1466
  (1,2)	18.7	17.6	-6.4	39.0	29302
  (0,0)	18.3	-8.2	16.3	30.0	26745
  (3,5)	17.9	15.1	-9.6	12.0	6506
  (3,3)	16.1	13.9	-8.2	17.0	40463

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F187N_j6778 visit 2: worst tile 20570.0 mas at cell (0,4) (> 15 mas)

  30/30 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F187N_j6778_latest.json

  what is affected, and why

  4/30 cells exceed 15 mas, and every one of them is on the mosaic EDGE — that is usually thin coverage (few exposures, so few pairs per cell) rather than a distortion error. Check the pair counts in the table before treating it as a defect. The bulk tie for this visit is 0.57 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,4)	20,570.0	-19,839.5	5,432.9	5.0	19185
  (2,5)	18.3	14.1	-11.6	23.0	9695
  (3,0)	15.9	-8.5	13.4	33.0	60390
  (1,5)	15.7	13.7	-7.7	11.0	23776

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F187N_j6778 visit 3: worst tile 45049.5 mas at cell (0,3) (> 15 mas)

  32/32 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F187N_j6778_latest.json

  what is affected, and why

  7/32 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 1.36 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,3)	45,049.5	39,789.7	-21,124.3	7.0	39399
  (2,0)	32.8	32.8	1.7	14.0	12704
  (3,1)	20.3	-6.5	-19.3	31.0	26447
  (2,2)	17.2	-15.7	6.8	26.0	22216
  (1,4)	17.1	14.8	8.6	16.0	9161
  (0,2)	15.9	-13.5	8.5	11.0	72192
  (5,4)	15.8	-3.8	-15.3	13.0	72832

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F200W_j6778 visit 1: worst tile 54080.4 mas at cell (5,0) (> 15 mas)

  35/35 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F200W_j6778_latest.json

  what is affected, and why

  24/35 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 0.53 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (5,0)	54,080.4	23,722.1	-48,599.9	7.0	38784
  (0,4)	42,313.6	-27,038.6	-32,547.7	11.0	76588
  (5,2)	27,265.6	6,265.8	-26,535.8	8.0	63562
  (5,3)	8,579.0	-586.7	8,559.0	8.0	57875
  (2,2)	30.5	-15.6	26.2	9.0	9849
  (2,3)	27.6	23.6	-14.3	20.0	28183
  (0,3)	26.3	-17.3	19.8	7.0	34833
  (3,1)	24.8	-23.1	8.9	13.0	12162
  (0,1)	22.8	17.9	14.0	18.0	80151
  (1,4)	22.0	17.6	13.2	10.0	14818
  (0,0)	21.1	-7.6	19.7	12.0	14504
  (1,0)	20.3	7.8	18.8	17.0	20309
  (1,2)	19.8	17.0	-10.1	19.0	20243
  (4,5)	19.8	-10.8	16.6	10.0	7449
  (2,1)	19.7	-11.2	-16.1	15.0	66318
  (1,3)	19.2	-15.8	10.9	21.0	37675
  (2,5)	19.0	12.7	-14.1	14.0	4231
  (2,0)	17.3	-15.1	8.3	15.0	25511
  (3,0)	17.1	-15.3	-7.7	19.0	46303
  (3,5)	16.8	15.5	-6.5	14.0	20160
  (4,1)	16.7	12.9	10.5	16.0	22829
  (1,5)	16.3	11.1	-11.9	6.0	2635
  (4,3)	16.0	9.8	-12.6	8.0	2699
  (3,4)	15.3	0.1	15.3	11.0	60500

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F200W_j6778 visit 2: worst tile 56615.4 mas at cell (3,5) (> 15 mas)

  32/33 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F200W_j6778_latest.json

  what is affected, and why

  18/33 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 0.93 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (3,5)	56,615.4	-51,502.1	23,512.5	6.0	28199
  (0,4)	55,762.1	-38,246.2	-40,578.9	8.0	72674
  (0,0)	54,327.9	-12,168.2	-52,947.7	7.0	150309
  (3,4)	52,773.9	8,190.7	-52,134.5	8.0	29800
  (5,5)	49,960.8	38,200.0	-32,200.0	4.0	6670
  (5,2)	45,847.8	-44,191.2	12,213.2	9.0	78829
  (5,1)	45,091.2	1,498.9	-45,066.3	12.0	74015
  (1,4)	41,491.4	-41,384.9	2,970.6	9.0	82327
  (0,1)	37,912.2	-35,290.7	-13,853.0	9.0	49037
  (3,3)	27.3	27.3	-1.5	22.0	51171
  (1,1)	22.3	4.6	-21.8	12.0	15441
  (3,0)	21.9	-17.6	13.0	20.0	32571
  (4,4)	20.0	16.9	-10.7	23.0	56776
  (0,3)	19.9	13.9	14.3	9.0	80811
  (0,2)	16.2	5.5	15.2	12.0	50096
  (3,1)	15.5	-10.7	-11.3	21.0	13064
  (2,1)	15.5	-14.9	-4.2	20.0	64093
  (4,1)	15.2	9.3	-12.1	16.0	10315

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F200W_j6778 visit 3: worst tile 58489.1 mas at cell (4,0) (> 15 mas)

  32/32 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F200W_j6778_latest.json

  what is affected, and why

  17/32 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 0.48 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (4,0)	58,489.1	-47,110.1	34,664.3	10.0	80737
  (5,5)	50,606.6	-47,767.7	-16,711.6	7.0	48794
  (4,5)	48,322.9	-29,015.1	-38,642.3	10.0	76523
  (0,3)	28,618.2	-20,927.7	19,520.0	6.0	42129
  (3,0)	24,362.2	-17,827.3	16,604.3	11.0	79450
  (4,1)	41.1	38.6	14.0	16.0	24861
  (3,1)	39.6	-13.9	-37.1	17.0	10987
  (2,0)	31.3	29.5	-10.3	13.0	56011
  (2,5)	29.2	29.0	-3.8	14.0	33839
  (1,5)	28.0	28.0	0.8	17.0	82351
  (2,2)	24.5	-18.6	16.0	17.0	13283
  (1,2)	17.8	8.1	15.8	17.0	19282
  (0,2)	17.5	-3.8	17.1	12.0	75698
  (0,4)	17.4	-11.9	-12.7	6.0	23597
  (1,4)	17.1	16.9	2.8	11.0	54915
  (1,3)	16.6	-8.2	14.5	21.0	24293
  (2,1)	15.9	-12.2	-10.2	23.0	29345

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F200W_j7213 visit 1: worst tile 76.5 mas at cell (2,0) (> 15 mas)

  27/27 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F200W_j7213_latest.json

  what is affected, and why

  27/27 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 47.69 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (2,0)	76.5	-60.4	47.0	22.0	116320
  (1,2)	73.8	-51.3	53.1	10.0	6856
  (3,0)	67.5	-60.6	29.7	28.0	27359
  (0,1)	66.9	-60.1	29.4	9.0	4144
  (1,1)	63.4	-47.1	42.5	13.0	3070
  (4,0)	62.5	-53.0	33.1	11.0	27135
  (1,0)	62.2	-51.3	35.3	26.5	86332
  (0,0)	59.5	-48.1	35.0	10.0	6900
  (1,5)	59.4	-26.7	53.0	9.0	47707
  (4,4)	58.9	-45.6	37.3	20.0	3054
  (4,2)	57.4	-51.4	25.6	35.0	38860
  (3,3)	55.2	-49.9	23.8	12.0	3295
  (5,5)	54.3	-18.9	50.9	5.0	698
  (0,2)	53.8	-42.2	33.3	6.0	2100
  (3,1)	53.7	-51.3	15.8	25.0	6544
  (2,2)	52.6	-47.9	21.6	14.0	1602
  (3,2)	50.7	-45.3	22.8	53.0	51624
  (4,3)	49.4	-41.4	27.0	13.0	1681
  (2,3)	49.4	-36.2	33.6	6.0	1455
  (1,3)	49.2	-32.3	37.1	23.0	56591
  (1,4)	48.4	-31.9	36.4	16.0	13602
  (4,1)	47.6	-42.2	22.1	24.0	27448
  (2,4)	47.0	-24.4	40.2	20.0	31720
  (2,1)	45.5	-40.2	21.2	13.0	20872
  (2,5)	41.3	-29.1	29.3	26.0	14823
  (3,5)	39.5	-20.3	33.9	11.0	6081
  (3,4)	30.4	-18.1	24.4	11.0	933

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F200W_j7213 visit 2: worst tile 57355.7 mas at cell (5,5) (> 15 mas)

  28/28 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F200W_j7213_latest.json

  what is affected, and why

  28/28 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 37.81 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (5,5)	57,355.7	51,063.7	-26,118.3	9.0	41514
  (5,3)	54,444.7	31,726.1	44,245.7	9.0	64741
  (4,0)	53,228.9	-25,791.4	46,563.1	7.0	31145
  (2,5)	1,057.2	960.5	441.5	6.0	769
  (0,1)	59.2	-50.5	31.0	10.0	22550
  (1,1)	53.4	-52.6	9.5	17.0	5279
  (4,2)	50.3	-50.2	-2.7	14.0	15489
  (2,2)	49.6	-48.7	9.5	21.0	10741
  (2,0)	49.1	-48.6	6.9	12.0	69392
  (0,0)	48.5	-47.3	10.6	17.0	18994
  (4,1)	47.1	-44.3	-16.1	13.0	79378
  (0,2)	45.5	-45.5	0.8	15.0	20484
  (2,1)	45.4	-43.8	11.9	21.0	13725
  (1,2)	44.9	-43.4	11.5	29.0	40782
  (3,1)	44.8	-43.7	9.6	23.0	36717
  (1,3)	41.9	-39.6	13.6	17.0	3119
  (1,0)	41.0	-38.0	15.5	13.0	4695
  (4,5)	40.3	-39.8	6.3	14.0	9069
  (3,0)	40.1	-40.0	1.3	14.0	21950
  (2,4)	39.0	-29.4	25.5	9.0	1132
  (3,2)	34.7	-34.3	5.5	11.0	7897
  (4,3)	34.0	-30.3	-15.5	17.0	46253
  (4,4)	32.9	-32.6	-4.5	14.0	22325
  (5,4)	32.2	-32.1	2.8	19.0	18294
  (2,3)	31.8	-31.6	3.8	8.0	2530
  (3,3)	31.3	-30.8	-5.6	11.0	2123
  (3,5)	23.8	-23.1	-5.7	8.0	323
  (3,4)	18.7	-16.6	-8.6	7.0	1959

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F277W_j6778 visit 1: worst tile 57935.2 mas at cell (5,4) (> 15 mas)

  33/33 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F277W_j6778_latest.json

  what is affected, and why

  27/33 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 0.77 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (5,4)	57,935.2	-42,804.1	39,042.2	6.0	35026
  (5,5)	56,392.0	-47,293.5	30,714.6	5.0	18130
  (4,3)	54,383.3	-54,189.4	4,588.6	9.0	56656
  (2,4)	52,383.9	-49,411.3	-17,395.4	10.0	58137
  (4,5)	49,871.9	-36,649.2	33,823.7	6.0	18471
  (2,5)	49,148.0	37,521.3	31,744.0	6.0	17777
  (5,2)	42,239.1	-24,965.5	-34,071.6	5.0	21002
  (5,3)	38,607.9	-37,804.9	-7,833.0	7.0	42106
  (5,1)	22,110.3	11,354.1	-18,972.4	7.0	57700
  (3,3)	35.5	24.3	-25.9	12.0	82693
  (4,4)	33.1	25.2	21.5	7.0	3899
  (1,5)	30.6	19.9	-23.2	7.0	309
  (2,1)	29.6	-22.3	-19.5	15.0	30486
  (1,2)	28.9	27.4	-9.1	20.0	45006
  (4,0)	26.6	-23.4	-12.6	14.0	17336
  (2,2)	23.8	-0.1	-23.8	11.0	26968
  (3,4)	23.2	7.4	22.0	15.0	41887
  (3,1)	22.5	-22.5	-0.8	13.0	62570
  (0,0)	20.7	-18.1	10.1	18.0	45516
  (3,0)	19.8	-12.6	-15.3	13.0	15006
  (2,3)	19.5	18.8	5.0	15.0	60269
  (1,0)	19.4	-9.1	17.2	14.0	65335
  (1,1)	18.3	18.2	2.2	16.0	44391
  (0,1)	18.2	17.4	5.3	10.0	33625
  (3,5)	17.0	10.9	-13.0	6.0	18025
  (0,4)	15.6	-3.0	15.3	5.0	5992
  (2,0)	15.5	-12.0	9.8	15.0	11199

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F277W_j6778 visit 2: worst tile 58897.6 mas at cell (0,1) (> 15 mas)

  32/33 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F277W_j6778_latest.json

  what is affected, and why

  33/33 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 69.26 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,1)	58,897.6	-58,897.1	240.6	7.0	43336
  (0,3)	50,548.1	27,824.1	-42,201.0	7.0	44078
  (0,4)	88.7	-79.7	-39.1	6.0	7213
  (3,1)	86.8	-80.4	-32.8	39.0	52684
  (5,4)	86.0	-79.8	-32.2	20.0	19813
  (3,5)	84.9	-80.0	-28.4	9.0	25667
  (3,4)	84.4	-77.7	-33.0	8.0	5976
  (2,1)	82.1	-80.9	-14.2	23.0	31072
  (5,3)	81.5	-77.8	-24.1	26.0	30634
  (3,2)	80.5	-74.9	-29.4	20.0	20118
  (3,0)	80.4	-78.7	-16.5	20.5	96732
  (2,0)	80.3	-77.8	-20.0	32.0	73142
  (5,1)	78.7	-64.0	-45.8	7.0	6359
  (1,0)	76.8	-73.9	-20.7	30.0	40924
  (4,2)	76.5	-70.7	-29.1	23.0	26419
  (4,1)	76.4	-68.2	-34.4	31.0	42744
  (1,2)	76.1	-73.1	-20.9	29.0	25611
  (4,0)	75.4	-71.9	-22.5	17.0	41288
  (0,0)	74.7	-73.4	-13.7	15.0	84320
  (4,5)	74.6	-65.3	-36.1	5.0	735
  (1,1)	73.8	-64.7	-35.5	10.0	64074
  (4,3)	68.5	-62.8	-27.4	29.0	46161
  (4,4)	67.5	-66.3	-12.5	27.0	22738
  (1,4)	65.3	-60.2	-25.4	12.0	71550
  (2,2)	65.1	-60.5	-24.2	24.0	20581
  (2,5)	61.3	-52.7	-31.3	14.0	22460
  (1,3)	59.8	-59.4	-7.2	21.0	41711
  (0,2)	57.9	-57.3	-8.0	15.0	9436
  (2,3)	56.1	-55.2	-10.3	19.0	43012
  (3,3)	52.1	-50.9	-11.4	17.0	86782
  (2,4)	49.8	-49.3	6.6	13.0	61750
  (5,5)	49.4	-49.2	4.4	4.0	1623
  (1,5)	48.6	-47.4	-10.8	16.0	17210

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F277W_j6778 visit 3: worst tile 59359.4 mas at cell (0,3) (> 15 mas)

  32/32 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F277W_j6778_latest.json

  what is affected, and why

  25/32 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 1.84 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,3)	59,359.4	-57,776.9	13,614.7	6.0	42908
  (5,5)	58,996.0	-5,896.0	-58,700.6	7.0	30201
  (0,4)	56,162.5	-55,868.9	-5,735.4	7.0	46205
  (4,0)	53,346.7	-19,424.8	49,684.5	8.0	45066
  (3,5)	52,393.5	-7,359.7	-51,874.0	8.0	45874
  (3,0)	50,700.8	-15,452.0	48,288.7	8.0	46429
  (0,2)	46,338.2	-33,014.0	-32,516.2	5.5	127288
  (4,1)	42,242.5	-36,158.7	-21,839.9	11.0	69491
  (2,5)	33,612.6	-25,315.8	22,111.5	9.0	76328
  (3,1)	38.7	-13.2	-36.4	19.0	44935
  (2,0)	38.4	31.9	-21.4	15.0	65326
  (1,5)	28.4	27.1	-8.4	12.0	37229
  (3,3)	25.3	-17.8	18.0	13.5	105957
  (4,3)	25.3	9.9	-23.2	11.0	13088
  (2,2)	23.3	-20.3	11.5	18.0	49691
  (5,4)	21.6	-18.9	-10.5	11.0	35220
  (4,2)	20.7	-6.1	-19.7	13.0	14514
  (1,3)	20.6	-3.4	20.3	21.0	75303
  (2,1)	20.3	-17.4	-10.6	16.0	10617
  (0,0)	18.7	-17.0	-7.8	11.0	39385
  (3,4)	17.8	11.8	-13.3	16.0	17141
  (4,4)	16.6	-15.1	-6.9	15.0	19142
  (4,5)	16.3	13.9	-8.5	7.0	14055
  (2,3)	16.0	-1.9	15.9	14.0	58145
  (5,3)	16.0	15.7	3.1	11.0	28063

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F335M_j6778 visit 1: worst tile 58271.8 mas at cell (5,2) (> 15 mas)

  31/32 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F335M_j6778_latest.json

  what is affected, and why

  32/32 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 70.63 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (5,2)	58,271.8	-15,400.0	-56,200.0	4.0	9976
  (5,1)	9,690.8	7,414.5	6,239.9	6.0	23932
  (2,1)	86.4	-82.9	-24.2	32.0	15716
  (2,2)	85.9	-80.5	-29.8	20.0	32914
  (0,0)	83.0	-82.7	-6.4	26.0	69457
  (2,0)	82.4	-77.8	-27.1	45.0	41432
  (0,2)	82.3	-78.7	-24.0	22.0	5402
  (3,2)	79.2	-75.8	-22.8	22.0	33458
  (1,3)	77.0	-73.0	-24.3	51.0	28345
  (3,1)	76.5	-74.8	-16.1	24.0	27947
  (3,0)	75.9	-71.1	-26.8	51.0	71737
  (4,0)	73.6	-70.5	-21.1	52.0	86408
  (2,3)	73.5	-67.7	-28.7	38.0	52425
  (1,0)	72.8	-72.0	-10.9	29.0	87534
  (4,5)	72.0	-69.8	-17.7	14.0	10868
  (5,3)	70.9	-68.0	-20.2	13.0	4407
  (4,2)	69.0	-66.8	-17.0	19.0	10956
  (2,5)	68.1	-62.7	-26.7	13.0	2639
  (3,4)	67.5	-66.4	-12.0	14.0	18742
  (3,5)	66.7	-60.0	-29.1	14.0	15850
  (4,3)	64.2	-62.3	-15.8	16.0	5034
  (4,1)	63.4	-61.4	-15.9	40.0	47572
  (3,3)	63.1	-56.7	-27.6	28.0	23658
  (2,4)	63.0	-59.3	-21.1	15.0	4854
  (1,1)	61.4	-58.9	-17.3	32.0	73423
  (1,5)	60.9	-49.1	-36.0	10.0	1730
  (1,4)	60.9	-59.3	-13.9	18.0	11429
  (5,4)	58.7	-55.3	-19.5	6.0	4027
  (4,4)	57.6	-55.4	-15.8	7.0	1744
  (1,2)	57.6	-52.5	-23.7	45.0	55734
  (0,1)	53.6	-50.8	-17.0	25.0	50728
  (5,5)	53.2	-52.7	-7.4	7.0	2683

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F335M_j6778 visit 2: worst tile 41388.6 mas at cell (0,3) (> 15 mas)

  30/31 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F335M_j6778_latest.json

  what is affected, and why

  31/31 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 71.65 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,3)	41,388.6	-28,728.0	29,794.5	5.0	24177
  (3,5)	96.0	-93.1	-23.5	8.0	12854
  (3,4)	94.4	-83.5	-44.1	8.0	17180
  (1,4)	89.2	-88.1	-14.2	13.0	42068
  (3,1)	87.5	-81.2	-32.7	42.0	19929
  (5,4)	84.3	-72.8	-42.6	23.0	8287
  (3,2)	82.0	-77.9	-25.7	21.0	33715
  (2,1)	81.9	-78.4	-23.9	30.0	10386
  (1,1)	81.6	-73.4	-35.6	17.0	25465
  (0,1)	80.2	-78.5	-16.6	11.0	21575
  (3,0)	79.5	-77.8	-16.4	48.0	31106
  (1,2)	79.4	-75.4	-24.8	34.0	46370
  (5,3)	79.0	-73.4	-29.2	36.0	13701
  (1,0)	77.8	-74.3	-22.9	40.0	65785
  (4,1)	76.4	-70.7	-28.9	37.0	71495
  (4,2)	74.6	-69.5	-27.3	25.0	6483
  (2,2)	74.3	-69.5	-26.2	24.0	40345
  (4,3)	74.2	-67.8	-30.2	36.0	58563
  (0,2)	73.7	-72.9	-10.9	18.0	5225
  (0,0)	72.1	-70.7	-14.0	18.0	8866
  (4,0)	71.7	-68.5	-21.1	20.0	56516
  (0,4)	71.2	-63.4	-32.5	7.0	3402
  (2,0)	71.1	-67.8	-21.4	40.0	23882
  (3,3)	70.8	-65.5	-26.8	25.0	51730
  (1,3)	69.5	-65.0	-24.6	38.0	29971
  (2,3)	68.1	-65.3	-19.2	25.0	28293
  (2,4)	68.0	-61.6	-28.9	17.0	39405
  (4,4)	66.4	-58.7	-31.1	32.0	9479
  (1,5)	62.8	-59.9	-18.8	18.0	7252
  (2,5)	60.6	-53.1	-29.2	19.0	8927
  (5,5)	46.6	-44.8	-12.8	4.0	861

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F335M_j6778 visit 3: worst tile 58630.1 mas at cell (0,4) (> 15 mas)

  30/31 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F335M_j6778_latest.json

  what is affected, and why

  10/31 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 1.47 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,4)	58,630.1	-58,257.2	-6,601.6	5.0	16271
  (0,3)	58,310.8	-55,546.6	-17,740.6	4.0	18189
  (2,0)	32.3	31.1	-8.6	10.0	28827
  (3,1)	27.2	-2.5	-27.1	17.0	51566
  (1,5)	27.1	26.9	-2.6	14.0	61449
  (0,2)	23.7	-23.3	4.2	7.0	47994
  (2,2)	21.9	-16.1	14.8	21.0	66802
  (5,4)	18.2	2.0	-18.1	13.0	54959
  (5,5)	16.7	14.0	9.1	6.0	2970
  (4,0)	15.4	-14.7	-4.6	7.0	3372

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F470N_j6778 visit 1: worst tile 48478.4 mas at cell (5,2) (> 15 mas)

  32/33 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F470N_j6778_latest.json

  what is affected, and why

  5/33 cells exceed 15 mas, and every one of them is on the mosaic EDGE — that is usually thin coverage (few exposures, so few pairs per cell) rather than a distortion error. Check the pair counts in the table before treating it as a defect. The bulk tie for this visit is 0.51 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (5,2)	48,478.4	-19,000.0	-44,600.0	4.0	11368
  (0,0)	25.1	-14.6	20.5	35.0	26059
  (2,0)	15.3	-13.9	6.4	42.0	46200
  (1,5)	15.2	9.5	-11.9	9.0	341
  (0,1)	15.2	14.0	6.0	15.0	77926

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F470N_j6778 visit 2: worst tile 58379.2 mas at cell (0,1) (> 15 mas)

  33/33 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F470N_j6778_latest.json

  what is affected, and why

  7/33 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 0.58 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,1)	58,379.2	-57,091.9	-12,192.0	10.0	51840
  (3,4)	57,043.0	-13,477.1	-55,428.1	8.0	46647
  (3,5)	56,550.9	-51,424.4	23,527.2	6.0	21340
  (3,3)	19.6	19.3	3.1	26.0	28563
  (2,5)	18.4	15.3	-10.2	23.0	12951
  (3,0)	16.9	-14.7	8.2	35.0	39194
  (0,0)	16.3	-9.8	13.0	16.0	1065

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F470N_j6778 visit 3: worst tile 30.6 mas at cell (2,0) (> 15 mas)

  31/31 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F470N_j6778_latest.json

  what is affected, and why

  5/31 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 0.40 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (2,0)	30.6	29.4	-8.8	16.0	52247
  (3,1)	24.0	-5.1	-23.5	26.0	29573
  (4,0)	21.9	-21.9	0.2	9.0	7104
  (0,2)	17.5	-7.8	15.7	10.0	59304
  (2,2)	15.6	-13.7	7.5	25.0	23048

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F470N_j7213 visit 1: worst tile 61.2 mas at cell (0,1) (> 15 mas)

  26/28 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F470N_j7213_latest.json

  what is affected, and why

  28/28 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 46.22 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,1)	61.2	-50.8	34.1	7.0	2312
  (2,0)	59.6	-48.5	34.6	43.0	52225
  (3,0)	59.4	-51.7	29.3	32.0	8065
  (1,0)	59.4	-50.2	31.7	57.0	57235
  (0,0)	57.8	-45.1	36.2	13.0	8069
  (4,0)	55.7	-49.3	26.1	14.0	9875
  (1,5)	53.4	-18.4	50.2	10.0	19676
  (0,2)	52.4	-38.4	35.7	8.0	1656
  (4,2)	52.2	-42.7	29.9	40.0	42532
  (1,1)	51.9	-40.6	32.4	13.0	2681
  (1,2)	51.0	-39.5	32.3	9.0	8349
  (2,2)	48.7	-39.8	28.0	13.0	2308
  (1,3)	48.5	-26.8	40.5	27.0	26522
  (4,1)	47.4	-44.2	17.2	30.0	29790
  (2,1)	46.7	-38.0	27.1	13.0	19827
  (3,1)	45.7	-42.1	17.8	26.0	6704
  (3,2)	45.3	-41.0	19.2	51.0	42536
  (2,3)	45.2	-28.6	35.0	4.0	1412
  (3,3)	42.5	-37.2	20.6	19.0	4003
  (1,4)	42.2	-22.9	35.4	29.0	42802
  (2,4)	41.8	-21.2	36.1	20.0	7146
  (4,4)	39.2	-24.5	30.6	16.0	2653
  (4,3)	37.7	-32.3	19.5	12.0	2190
  (5,4)	36.2	-27.1	24.1	12.0	151
  (2,5)	33.6	-12.3	31.3	37.0	60571
  (3,4)	32.9	-21.0	25.3	9.0	1197
  (3,5)	30.5	-13.1	27.5	11.0	3601
  (5,5)	26.0	-17.5	19.2	3.0	65

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• warn
  F470N_j7213 visit 2: worst tile 52.9 mas at cell (0,1) (> 15 mas)

  28/28 tiles are reported "ok", but that counts tiles whose offset histogram had a coherent PEAK — not tiles within tolerance. The m7 cross-band gate is no significant 2" cell above 15 mas, and this cell exceeds it, so the bulk tie being ~0 does not mean the field is flat.

  checkpoint_m2_F470N_j7213_latest.json

  what is affected, and why

  28/28 cells exceed 15 mas. They are not confined to the edge, so this is an interior residual: a distortion or per-detector alignment problem, not a coverage artefact. The bulk tie for this visit is 36.68 mas — a small bulk value cannot cancel a local one.

  Diagnostic writeupD3 — absolute tie to the reference frame · full diagnostic writeup (PDF)

  cell	off (mas)	dRA	dDec	contrast	pairs
  (0,1)	52.9	-46.3	25.6	11.0	22127
  (0,0)	51.3	-48.8	16.1	20.0	4429
  (1,1)	49.9	-48.3	12.7	18.0	6545
  (1,0)	48.5	-46.2	14.6	15.0	3224
  (4,0)	45.2	-45.2	1.0	6.0	2604
  (3,1)	44.5	-43.8	8.2	25.0	10616
  (3,0)	44.5	-44.3	3.7	21.0	27960
  (2,1)	44.1	-42.8	10.3	21.0	30812
  (2,0)	42.9	-42.2	7.7	13.0	4434
  (3,2)	42.8	-41.4	10.8	14.0	4208
  (2,2)	42.7	-41.9	8.3	20.0	6873
  (0,2)	42.2	-41.5	7.4	13.0	13666
  (1,2)	41.8	-40.7	9.5	35.0	35670
  (4,1)	36.4	-35.9	5.9	10.0	718
  (4,2)	36.4	-36.4	1.1	14.0	4728
  (4,3)	36.3	-33.6	-13.8	23.0	33080
  (5,3)	36.0	-36.0	1.3	6.0	3443
  (1,3)	33.3	-32.0	9.2	22.0	16426
  (4,5)	31.9	-31.2	-6.8	10.0	3957
  (4,4)	31.3	-31.2	-2.7	13.0	5784
  (5,4)	31.1	-31.0	-2.0	23.0	13916
  (3,3)	28.6	-28.3	-4.1	9.0	455
  (2,3)	27.7	-27.2	5.3	7.0	204
  (3,4)	26.2	-24.8	-8.1	6.0	1002
  (2,4)	25.7	-23.3	11.0	8.0	115
  (5,5)	21.4	-21.2	-2.9	9.0	4214
  (3,5)	17.6	-17.4	2.8	9.0	271
  (2,5)	17.5	-14.4	9.9	10.0	10519

  Existing figures

  • astrometry_offsets_2026-07-10.png audit_plots
• ok
  m7 cross-band product absent

  catalogs/basic_*_photometry_tables_merged_*.fits
• ok
  m8 cross-band product absent

  catalogs/basic_*_photometry_tables_merged_*.fits
• ok
  7 job(s) in the queue (1 running)

  ngc63346778-o001-m4-finalize, ngc63346778-o001-m5-fanout, ngc63346778-o001-m5-finalize, ngc63346778-o001-m6-fanout, ngc63346778-o001-m6-finalize, ngc63346778-o001-m7-fanout, ngc63346778-o001-m7-finalize

  squeue