1-read thoroughly at line-level for flaws, gaps and internal...

Prompt

1-read thoroughly at line-level for flaws, gaps and internal inconsistencies. 2- how can the work be improved or meaningfully generalized? propose useful upgrades: Introduction The viability and obstruction machinery of the calculus — survivable sets, robust multipliers, breach-time certifi- cates, and probe design — has so far been developed and validated on deliberately small exact instances. This pa- per applies it to the Northern cod stock (NAFO divisions 2J3KL) and certifies the obstruction structure of its col- lapse on the public assessment record. The forecast layer The certified windows Proposition 1 (the two windows are sub-unitary at worst case). From the Table A2 series: the 1983–1989 win- dow’s six annual net multipliers have minimum ρg = 41800 836.00/902.94 = 45147 ≈ 0.9259 (1986), and the 1991– 1993 window’s two annual net multipliers are 38195 73451 ≈ 2021 0.520 (1992) and 7639 ≈ 0.265 (1993). Both windows’ worst-case multipliers are strictly below unity. 1Proof. Each multiplier is the exact ratio of consecutive Ta- ble A2 SSB entries in kt over 100, computed in the script as a rational; the minimum and the two bad-window val- ues are the displayed fractions by direct comparison of the six and two ratios respectively. Each displayed fraction is below 1 because its numerator is smaller than its denom- inator (41800 < 45147, 2021 < 7639). an output of the assessment), recorded at archival time. Proposition 5 (the reconstruction certifies a near-closed fishery). The catch reconstruction gives removals of 1,314 t (1994) and 413 t (1995) — four and five orders of mag- nitude below the fishery’s operating scale in the preceding Proposition 2 (the moratorium obstruction). Across decade — and its 2015 row (4,436 t) agrees with the official 1992–1993 the realized removals fell from 40,956 t to removals table and with the assessment report, a cross-file 11,392 t (official removals table), approaching the closed consistency check. The Proposition 2 witness therefore fishery; the realized net multiplier of the same pair of years stands on removals effectively at zero. is 2021 7639 ≈ 0.265. At the realized rates, removing the fishery did not arrest the decline: the obstruction is witnessed in Proof. The three rows are read from the locked recon- struction file; the agreement at 2015 is the exact equality the dynamics, not in the catch level. 4436 = 4436 across the two files (and matches the as- Proof. The removals are the certified table rows for 1992 sessment report’s removals figure, as recorded at archival and 1993; the multiplier is Proposition 1’s 1993 ratio, time). The comparison with the preceding decade’s scale whose numerator and denominator are the realized 1993 is the ratio against the 1983–1989 removals rows, all above and 1992 SSB values. A multiplier of 2021 7639 with removals 130,000 t in the same file. at 11,392 t against a biomass of tens of thousands of kilo- tonnes certifies that the decline is dominated by the non- fishing component of the net rate; the statement is a wit- 0.6 The reference decade was already ness about the realized window, not a counterfactual claim marginal about other catch paths. Proposition 6 (the reference point is the decade’s own mean, and the decade was not robustly viable). (i) The 0.4 Breach-time certificates limit reference point 884.6 kt equals the 1983–1989 mean Proposition 3 (breach bounds and the realized timeline). of the Table A2 SSB series (884.58, rounded), and the With Flim = 276 kt: (i) from the end-1991 reading 734.51 decade’s readings stay within ±6.4% of it: the ratios run 4180 ≈ 0.945 (1986) to 94075 kt, one step at the window’s worst multiplier already cer- from 4423 88460 ≈ 1.064 (1987). (ii) 27600 2021 tifies breach, since 7639 ≤ 73451 ; (ii) under the realized Under the window’s worst-case multiplier ρg = 41800 45147 sus- two-step sequence the stock is below Flim in exactly two tained from the 1989 reading 921.66 kt, the stock breaches 27600 15 years — 101.05 < 276 in 1993, the xteNCAM file’s SSB- Flim = 276 kt within sixteen years: ρ16 g ≤ 92166 < ρg . No over-Blim ratio 0.29 at 1993 confirming the same fact in removals policy certifies indefinite above-floor viability at the independent assessment; (iii) under the limit reference worst-case realized multipliers, since even zero removals point = 884.6 kt the 1991 reading 734.51 is already below inherits the net multiplier’s strict decline. the reference point. The certificates match the realized timeline exactly. Proof. (i) The mean is the exact rational sum of the seven readings over seven, equal to 88458 100 up to the certified 2021 2021·73451 Proof. (i) is the exact inequality 7639 = 7639·73451 against rounding of the record; the displayed ratio bounds are the 27600 73451 , checked by cross-multiplication of integers in the exact minimum and maximum of the seven ratios. (ii) The script (2021 · 73451 < 7639 · 27600). (ii) is the direct com- sixteenth and fifteenth integer powers of 41800 are com- 45147 parison 101.05 < 276 on the Table A2 row for 1993, with pared against 27600 by exact cross-multiplication in the 92166 the xteNCAM ratio column read as the locked decimal script. The final sentence: any policy with non-negative 0.29 at 1993. (iii) is 734.51 < 884.6. No step of the proof removals has a net multiplier at most the zero-removals uses floating point. multiplier, which at worst case is ρg < 1, and a strict con- traction crosses any fixed positive floor in finite time (at 0.5 The survey instrument and the catch most ⌈log(Flim /z0 )/ log ρg ⌉ steps, the integer certified by the exact power comparison). record Proposition 4 (the model-independent survey confirms the collapse). The fall research-vessel survey index falls from 2,127,417 (1989) to 1,117,670 (1991), 239,740 (1992), 90,709 (1993), and 21,797 (1994): a certified fac- 21797 239740 tor of 2127417 ≈ 0.0102 over 1989–1994, and 1117670 ≈ 0.214 across 1992 alone. The instrument is independent of the assessment models that produced the biomass series, so the collapse is witnessed twice, by two instruments.Corollary 7 (what the reference point encodes). The ref- erence point’s construction — the 1980s mean — is pre- cisely the certification that the healthy decade sat at its own failure boundary: the stock was marginal at Blim -scale via- bility before the regime shift, which is the viability reading of the same fact the assessment literature records as the reference point’s provenance. Proof. All four values are rows of the locked survey file; the factors are their exact ratios. Independence of the assessment model is a provenance property of the file (the survey index is an input collected by research vessels, notProof. Immediate from Proposition 6(i)–(ii): the mean- rule fixes the boundary inside the decade’s realized range, and the worst-case contraction certifies non-viability of the sustained worst case within sixteen years. 20.8 Table 1: The certified windows and instruments (all quanti- ties exact; SSB in kt, removals in t, survey index in model- independent units). Delimitations Every claim is a certificate about the realized windows of the certified series — not a forecast, not a stochas- Quantity Value Source file tic model, and not a counterfactual catch analysis: the net multipliers conflate the mortality regime with all re- Decade worst mult. (1986) 41800/45147 Table A2 1992, 1993 multipliers 0.520, 0.265 Table A2 movals (the archival diagnostics rule), so no statement Removals 1992, 1993 40,956, 11,392 Table 1 (2025) here identifies what a different catch path would have Removals 1994, 1995 1,314, 413 Reconstruction Survey index 1989, 1994 2,127,417, 21,797 Survey T3 done. The biomass series are assessment outputs (model- Covariate 1990, 1991 5783, 138 Capelin file based), treated as the data of record with their published Outcome years 2016–2021 340–440 kt DB extract Breach of 276 kt by 1993 (realized) A2 + xteNCAM uncertainty untouched; the survey index and the covariate are independent instruments and carry their own sampling error, untouched here. The covariate series is observed- 0.7 The covariate alarm and the outcome years only; the one-year lead is a statement about the years certified rows, robust to no claim about unobserved years. Single stock; two windows; no averaging beyond the refer- Proposition 8 (the covariate alarm leads the mortality ence point’s own construction. crossing). In the observed capelin series, the acoustic es- timate falls from 5783 kt (1990) to 138 kt (1991) — a 138 ≈ 0.024 — one step before the Ta- 0.9 Verification methods certified factor of 5783 ble A2 mean natural mortality diagnostic crosses M = 1 (0.403 in 1990 to 1.002 in 1991, diagnostic column, not a Every number, fraction, inequality, power, ratio, and tem- model input). The one-year lead makes the covariate the poral claim in this paper is regenerated by the accom- certified probe datum of the regime shift: the monitoring panying script in exact integer and rational arithmetic, standard library only, directly from the locked files: the instrument moved before the mortality regime did. six good-window and two bad-window multipliers; the re- Proof. Both readings are observed rows of the locked movals rows of both removals files and their 2015 cross-file capelin file (no interpolation, per the archival discipline); equality; the survey and covariate factors; the breach com- the factor is their exact ratio. The mortality crossing is parisons by integer cross-multiplication; the sixteenth and the diagnostic column’s consecutive values straddling 1. fifteenth powers of the decade’s worst multiplier; the seven The proposition asserts only the temporal order of the reference-point ratios with their exact minimum and max- imum and the mean rule; the outcome-year comparisons; two certified events, which the rows display directly. and every headline number of this text asserted present Proposition 9 (the outcome years: partial recovery, no verbatim. return to the reference scale). The per-year database ex- tension reads the stock at 340, 433, 394, 419, 440, 411 kt for 2016–2021 (same assessment family, 2021 vintage): the stock re-crosses the 276 kt auxiliary threshold by 2016 and stays above it through 2021, while its maximum (440 kt, 44000 ≈ 49.7% of the reference point. 2020) remains at 88460 The recovery proceeded under a modest fishery (4,436 t in 2015), completing the bidirectional reading: the dynamics, not the catch level, set the trajectory on the way up as on the way down. 0.10 Conclusion Applied to the public record of the Northern cod, the calculus certifies what the narrative record suggests and the data now proves exactly: the healthy decade was al- ready marginal at the auxiliary threshold (the reference point is its own mean); the collapse proceeded with the fishery closing (the moratorium obstruction, witnessed at two orders of magnitude of removals); the breach time- line is exactly two years from the last pre-collapse read- ing; two model-independent instruments — the survey and the covariate — moved before or with the regime crossing; and the partial recovery re-crossed the auxiliary threshold without approaching the reference scale. Viability on this stock was never a catch-optimization problem; it was a monitoring problem. That is the applied content of the obstruction calculus, now certified. Proof. The seven values are the locked database rows for 2016–2021 (exact integers in kt); the crossing and maxi- mum comparisons are direct rational comparisons against 276 and 884.6; the ratio 440/884.6 = 44000/88460 is re- duced exactly in the script. The 2015 removals row is the cross-checked file value of Proposition 5. Remark 10 (the monitoring consequence). Read together: catch level is not the certifiable instrument at the real-