Groups lost at ptp_flags: f220 41.0% (expected 10.9%, typical 0.6%); f280 27.6% (expected 6.1%, typical 0.6%)
Groups lost at other steps: f220 7.3% (expected 2.0%, typical 0.7%)
Observations with every group lost: f220 77.6% (expected 30.2%, typical 6.6%); f280 77.6% (expected 30.2%, typical 6.6%)
… of which N_dets ≤ 10: f220 79.0% (expected 44.5%, typical 8.7%); f280 79.0% (expected 41.2%, typical 7.5%)
Wafers: ws0 in 43% of CMB observations (usually 100%). ws5 in 84% of CMB observations (usually 100%). ws6 in 94% of CMB observations (usually 100%). (+10 more below)
61 normal · 8 worse but expected for this PWV · 2 better · 5 on too few groups to judge · 25 worse
DetailsMore worse than usual:
Groups lost at noise: f280 11.6% (expected 5.3%, typical 1.9%)
Weekly interval 2026-10-11 00:00 to 2026-10-18 00:00 UTC. "Usual" = every earlier 7-day window with a normal amount of data, stepped by a week, since the baseline start (satp2 from 2026-04-04). Time plots show the interval and the 8 weeks before it.
Range bar: box: typical (middle half, 25–75%) of earlier windows of the same length with a normal amount of data; whiskers: 10–90%; tick: median; ◇: expected at this interval's PWV; ●: this interval. Status: worse at least 3 robust standard deviations (IQR / 1.35) on the bad side of the median and not explained by PWV; weather as far out but as expected for this PWV; better as far out on the good side; normal otherwise; watch a departure on fewer than 300 groups.
1.1 Dataset coverage
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
CMB observations / week
69
123 (105–137)
normal
69
123 (105–137)
normal
Data acquired (h / week)
51
95 (79–108)
normal
51
95 (79–108)
normal
Duty cycle
30.6%
56.8% (46.9%–64.0%)
normal
30.6%
56.8% (46.9%–64.0%)
normal
1.2 Top-level statistics
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
Detector yield
42.4%
55.5% (48.4%–62.0%)
38.2%
normal
32.1%
52.4% (43.8%–57.9%)
32.4%
normal
Mapping efficiency
2.6%
26.6% (19.0%–34.2%)
5.7%
normal
2.0%
25.9% (17.7%–31.9%)
3.4%
normal
Per-detector NET [µK√s]
1126.7
869.2 (753.2–944.4)
1292.3
normal
3281.6
2352.1 (2036.2–2482.3)
3549.2
normal
Array NET [µK√s]
16.75
12.96 (11.58–15.69)
23.99
normal
60.11
35.43 (32.15–44.55)
69.92
normal
Map depth [µK·arcmin]
553.5
136.5 (106.7–168.3)
273.8
worse
1787.1
372.2 (301.4–491.2)
801.5
worse
Samples flagged
17.2%
16.6% (15.4%–18.8%)
21.5%
normal
16.1%
15.8% (13.8%–17.5%)
20.5%
normal
Gold: this interval, shown with the 8 weeks before it. Small points: each observation; thin line: daily; squares: weekly. Shaded band and dashed line: the typical week (25–75% and median of weekly values since the baseline start, low-data weeks left out). Dotted / dash-dot: baseline / goal where the wiki page has them. Choose the quantity in the menu.
2.1 Attrition
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
Reached init archive
19.1%
88.0% (80.7%–92.6%)
49.5%
worse
20.5%
89.1% (80.5%–92.4%)
51.3%
worse
Reached proc archive
19.1%
88.0% (80.5%–92.6%)
49.5%
worse
20.3%
89.1% (80.0%–92.4%)
51.3%
worse
Lost before init
80.9%
12.0% (7.4%–19.3%)
50.5%
worse
79.5%
10.9% (7.6%–19.5%)
48.7%
worse
Lost init → proc
0.0%
0.0% (0.0%–0.1%)
0.01%
normal
0.2%
0.0% (0.0%–0.1%)
0.02%
normal
Observations with every group lost
77.6%
6.6% (1.1%–11.0%)
30.2%
worse
77.6%
6.6% (1.1%–11.0%)
30.2%
worse
Gold: this interval, shown with the 8 weeks before it. Small points: each observation; thin line: daily; squares: weekly. Shaded band and dashed line: the typical week (25–75% and median of weekly values since the baseline start, low-data weeks left out). Dotted / dash-dot: baseline / goal where the wiki page has them. Choose the quantity in the menu.
Groups per day in this interval, by outcome.
2.1a Why groups never reached the archive (share of expected groups)
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
noise
7.8%
1.9% (0.6%–3.0%)
5.6%
weather
11.6%
1.9% (0.7%–2.7%)
5.3%
worse
noisy_subscan_flags
22.2%
7.7% (3.7%–11.9%)
28.9%
normal
34.7%
7.4% (4.0%–15.8%)
33.2%
weather
cut_bad_dist
2.6%
0.3% (0.1%–0.6%)
3.1%
weather
3.5%
0.4% (0.2%–0.6%)
3.0%
weather
ptp_flags
41.0%
0.6% (0.2%–2.0%)
10.9%
worse
27.6%
0.6% (0.0%–1.9%)
6.1%
worse
other steps
7.3%
0.7% (0.1%–1.3%)
2.0%
worse
2.1%
0.4% (0.1%–0.9%)
1.0%
normal
LoaderError: no metadata
0.0%
0.0% (0.0%–0.0%)
normal
0.0%
0.0% (0.0%–0.0%)
normal
crash
0.0%
0.0% (0.0%–0.1%)
normal
0.2%
0.10% (0.0%–0.2%)
normal
not attempted / not in logs
0.0%
0.0% (0.0%–0.0%)
normal
0.0%
0.0% (0.0%–0.0%)
normal
2.1b The PWV and detector-count filter
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
Excluded by PWV < 2 mm and N_dets > 10
100.0%
27.5% (9.7%–37.4%)
93.9%
weather
100.0%
27.2% (8.9%–37.4%)
93.9%
worse
… of which PWV ≥ 2 mm
91.5%
23.8% (6.2%–37.4%)
91.5%
23.8% (6.2%–37.4%)
… of which N_dets ≤ 10
79.0%
8.7% (3.0%–14.9%)
44.5%
worse
79.0%
7.5% (2.3%–13.8%)
41.2%
worse
2.2 Detector-level cuts (fraction of 860 per wafer)
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
Starting
83.0%
84.0% (83.9%–84.1%)
watch
75.2%
76.2% (76.1%–76.3%)
watch
Total cut
31.6%
27.8% (21.2%–31.5%)
33.3%
normal
39.0%
23.7% (15.9%–27.9%)
34.1%
normal
Final yield (per wafer)
51.3%
56.3% (52.6%–62.8%)
50.5%
normal
36.2%
52.4% (47.8%–60.4%)
42.0%
normal
Gold: this interval, shown with the 8 weeks before it. Small points: each observation; thin line: daily; squares: weekly. Shaded band and dashed line: the typical week (25–75% and median of weekly values since the baseline start, low-data weeks left out). Dotted / dash-dot: baseline / goal where the wiki page has them. Choose the quantity in the menu.
2.2a Cut steps
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
detcal_nan_cuts
7.6%
1.8% (1.7%–4.9%)
4.3%
normal
14.1%
3.5% (2.7%–7.1%)
7.0%
watch
noisy_subscan
9.8%
10.1% (7.0%–12.4%)
14.6%
normal
7.6%
5.5% (3.2%–6.4%)
8.5%
normal
det_bias_flags
5.9%
1.4% (0.7%–3.0%)
4.2%
watch
8.5%
1.9% (0.9%–3.7%)
5.9%
watch
noise
2.2%
1.6% (1.1%–2.6%)
3.6%
normal
5.3%
2.6% (1.5%–4.8%)
8.5%
normal
dark_dets
3.8%
3.8% (3.8%–3.8%)
3.8%
normal
0.0%
0.0% (0.0%–0.0%)
0.0%
normal
ptp_flags
0.7%
1.8% (1.3%–2.2%)
1.1%
normal
1.7%
3.2% (2.6%–3.6%)
2.4%
normal
trends
0.7%
0.4% (0.2%–0.8%)
0.7%
normal
0.9%
0.4% (0.3%–0.9%)
0.8%
normal
cut_bad_dist
0.3%
1.0% (0.8%–1.1%)
0.3%
normal
0.03%
0.5% (0.2%–0.5%)
0.01%
normal
fp_flags
0.0%
0.0% (0.0%–0.0%)
0.0%
normal
0.0%
0.0% (0.0%–0.0%)
0.0%
normal
2.3 Sample flags
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
noisy_subscan
13.3%
11.6% (10.2%–14.2%)
16.7%
normal
12.1%
10.6% (8.5%–12.3%)
15.6%
normal
turnaround
3.9%
4.9% (4.8%–5.1%)
better
4.0%
5.0% (4.8%–5.2%)
better
smurfgaps
0.0%
0.0% (0.0%–0.0%)
normal
0.0%
0.0% (0.0%–0.0%)
normal
glitches
0.00%
0.01% (0.01%–0.01%)
0.00%
normal
0.00%
0.01% (0.00%–0.01%)
0.00%
normal
Gold: this interval, shown with the 8 weeks before it. Small points: each observation; thin line: daily; squares: weekly. Shaded band and dashed line: the typical week (25–75% and median of weekly values since the baseline start, low-data weeks left out). Dotted / dash-dot: baseline / goal where the wiki page has them. Choose the quantity in the menu.
2.4 Efficiency and retention
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
Observing efficiency: all CMB obs
6.7%
49.1% (40.0%–56.0%)
worse
7.1%
49.1% (41.2%–56.2%)
worse
… + PWV < 2, N_dets > 10
6.7%
49.1% (38.7%–55.8%)
worse
6.7%
49.1% (40.3%–55.8%)
worse
… + excluding flagged samples
6.4%
47.6% (37.4%–54.1%)
worse
6.5%
47.7% (38.9%–54.1%)
worse
Preprocessing efficiency (all groups)
11.8%
58.0% (50.1%–67.5%)
23.2%
weather
10.0%
59.5% (51.9%–71.0%)
20.8%
weather
… archived groups only
57.9%
65.4% (60.3%–73.1%)
57.4%
normal
45.6%
67.1% (61.1%–76.9%)
51.8%
normal
Loading × observing
22.4%
47.6% (38.9%–50.9%)
normal
20.1%
43.2% (35.3%–46.2%)
normal
Data retention (det-hours retained / acquired)
57.9%
65.4% (60.3%–73.1%)
57.4%
normal
45.6%
67.1% (61.1%–76.9%)
51.8%
normal
Detector-hours acquired / week
46k
391k (333k–459k)
worse
45k
354k (300k–415k)
worse
Detector-hours retained / week
27k
280k (197k–337k)
normal
20k
260k (194k–316k)
normal
Gold: this interval, shown with the 8 weeks before it. Small points: each observation; thin line: daily; squares: weekly. Shaded band and dashed line: the typical week (25–75% and median of weekly values since the baseline start, low-data weeks left out). Dotted / dash-dot: baseline / goal where the wiki page has them. Choose the quantity in the menu.
Monthly detector-hours acquired and retained, and retention (line), as on the wiki page.
3 Noise
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
White noise Q (stats DB, ×10⁻⁶)
158.3
121.8 (109.3–136.5)
165.7
normal
190.9
142.4 (127.0–159.1)
195.3
normal
White noise T (stats DB, ×10⁻⁶)
110.4
81.3 (75.9–94.6)
112.5
normal
127.4
93.6 (87.3–108.6)
128.3
normal
f_knee Q [mHz]
46
18 (12–26)
49
normal
42
13 (10–20)
39
weather
NET per obs [µK√s] against APEX PWV. Contours: every observation in the baseline; points: this interval. A point above the contours is worse than the weather explains; ▲/▼: outside the plotted range.
yield per obs [%] against APEX PWV. Contours: every observation in the baseline; points: this interval. A point above the contours is worse than the weather explains; ▲/▼: outside the plotted range.
Distribution of per-observation NET: history (filled) and this interval (outline).
5 Weather
Quantity
f220
f280
This interval
Usual
At this PWV
This interval
Usual
At this PWV
APEX PWV during CMB obs [mm]
4.27
1.25 (0.91–1.77)
4.27
1.25 (0.91–1.77)
APEX PWV through the whole interval, including when not observing.
Share of each day's CMB observations that include the wafer (box: this interval; two weeks before it for context).
Files
data.json: every number on this page. Generated 2026-11-15 23:47 UTC by sat_quality_reports.py from the ISO-SAT v4 260702 archives,
stats DBs and preprocessing logs, the loaded-detector sweeps, the site obsdb and the APEX PWV archive.