The current CSV aggregation uses FMP Area == BSAI, both pollock target categories, the single Scypho jellies group, and the sum of retained and discarded Nontarget Estimate (mt) records. BSAI includes BS and AI. Annual BSAI totals reconcile to BS plus AI and to retained plus discarded totals. The current export contains no exact duplicate rows, and all BSAI jellyfish weight records are populated. All 2019–2022 BSAI jellyfish detail-record counts are one, so weighting by that count produces the same totals.
Source-history finding
The September 10, 2024 assessment source at revision c29d1e0 contains jellyfish totals close to the current CSV totals. The November 5 source at revision d38de13 approximately doubled the 2019–2022 totals. In 2019 the November total is exactly twice the September total at stored precision; other years differ slightly as well as showing the same twofold pattern. The earlier 2003–2018 jellyfish totals stayed unchanged across these revisions. The November summary was carried into the 2024 SAFE table.
The factor-of-two pattern therefore predates the current processing and is already present in the historical summary CSV. The 2024 Quarto table code reads that precomputed CSV and formats it; it supplies no doubling step. The upstream raw export or aggregation operation that introduced the larger historical totals remains unavailable in the inspected record. Duplication in that upstream preparation is a plausible explanation, not an established mechanism. The current table retains the directly summed CSV estimates.
Other nontarget columns, 2019–2022
The November summary approximately doubles each of the six same-named historical columns Jellies, MiscFish, SeaStar, Osmerid, Grenadier, and Eelpouts relative to September. All 24 year-column ratios are within 0.001 of two. These are comparisons of historical summary column labels; they do not establish that every historical label represented the same biological grouping as the current explicit species-name mapping.
The other four columns (Sea pen, Anemone, Snails, and Other) change non-uniformly. Numerical matches across columns indicate changed column alignment or allocation. For example, November 2019 Snails (1.287252353 t) equals twice September Sea pen (0.643626176 t), while November Sea pen (0.651518564 t) equals twice September Anemone (0.325759282 t). The same two cross-column patterns occur in 2020–2022. These are numerical relationships, rather than confirmation of a particular upstream coding error. A blanket division by two would retain ambiguous labels and fail to reconcile all groups.
The current Table 57 is rebuilt directly from named species in the new export: Jellies = Scypho jellies; MiscFish = Misc fish; SeaStar = Sea star; Osmerid combines the explicitly listed osmerid groups; Grenadier combines named grenadier groups; and the remaining columns use their published source-name mapping. BSAI records and both catch dispositions are included. The retained current table values receive no scaling adjustment. Its category mapping and excluded squid/sculpin/seabird scope are separate from the historical column-label comparison.
All 40 year-column comparisons · Current species-to-column map
Attaching package: 'dplyr'
The following objects are masked from 'package:stats':
filter, lag
The following objects are masked from 'package:base':
intersect, setdiff, setequal, union
| Jellies |
2.000 |
2.000 |
2.000 |
2.000 |
| MiscFish |
2.000 |
2.000 |
2.000 |
2.000 |
| SeaStar |
2.000 |
2.000 |
2.000 |
2.000 |
| Osmerid |
2.000 |
2.000 |
2.000 |
2.000 |
| Grenadier |
2.000 |
2.000 |
2.000 |
2.000 |
| Eelpouts |
2.000 |
2.000 |
2.000 |
2.000 |
| Sea pen |
1.012 |
9.859 |
3.115 |
1.895 |
| Anemone |
11.429 |
7.481 |
0.080 |
0.230 |
| Snails |
2.919 |
2.345 |
3.656 |
3.791 |
| Other |
1.506 |
0.502 |
2.583 |
2.118 |
| Ratios compare the same historical header, including columns with apparent alignment changes. |
Competing explanations checked
- Area: AI contributes only 0.122 t in 2020 and 0.00685 t in 2022, with zero in 2019 and 2021. Including all GOA and inside-water records still falls far below the 2024 SAFE totals.
- Species: Jellyfish are selected directly from the single source group
Scypho jellies; the mapping of other nontarget columns has no effect.
- Disposition: Both retained and discarded weights are included once. In 2019 all BSAI jellyfish records are discarded. In 2020–2022 the retained component is included and is much smaller than the discrepancy.
- Duplicates/units: Exact-duplicate rows are absent. The source field is already metric tons, and no record-count expansion or unit conversion is applied.
| 2019 |
3,889.126 |
7,778.251 |
3,889.126 |
3,889.126 |
0.000 |
2.000 |
| 2020 |
3,149.954 |
6,299.888 |
3,150.056 |
2,822.340 |
327.715 |
2.000 |
| 2021 |
7,829.681 |
15,660.867 |
7,831.182 |
6,387.228 |
1,443.954 |
2.000 |
| 2022 |
7,609.703 |
15,218.266 |
7,608.570 |
6,414.432 |
1,194.138 |
2.000 |
| All catch values are unrounded annual sums in metric tons; ratios are dimensionless. |
Annual audit CSV · Disaggregated filter audit
Historical source: noaa-afsc/ebs_pollock_safe2024, doc/data/byc_nontarg.csv, revisions c29d1e0 and d38de13.