Scientific sources and scope

The four-country author-taxonomy study withholds one target country from its training set, assigns native pollen counts to 31 author targets and preserves all excluded or unmapped count mass. Its input manifest, coverage and score table identify the exact support.

Target	Sites	Samples	Assigned/native pollen NISP	Scored/attempted bins
Denmark	1	91	122,998/129,768	24/24
Finland	28	1,797	1,229,019/1,260,281	435/435
Norway	52	3,544	2,018,295/2,286,020	1,084/1,178
Sweden	3	228	221,800/240,966	51/54

The 94 Norwegian and three Swedish unsupported bins remain explicit refusals. Denmark's one site cannot establish between-site Danish skill. Country outcomes had already been inspected in earlier development, so these are exploratory scores. The response is a closed 31-target pollen subcomposition, not full pollen, modeled vegetation or measured vegetation. Its MAE should not be compared directly with the earlier 16-label response.

The Swedish metadata-only cohort screen found that all 99 site IDs in a current Neotoma pollen query are already in the retained snapshot. Three further PANGAEA profile handles are already retained under UNASSIGNED country decisions. Three named handles have no exact match among all retained cores; their physical identity, selected pollen-sample ages and permitted use remain unresolved. We did not open their target pollen counts or claim independent confirmation.

Before a core can enter LRA

The physical core, selected age model, age-specific basin geometry, count denominator, mapped taxon concepts, PPE/fall speed use and rights must refer to the same analytical window. The Abborrtjärnen source review matches substantial depth and count evidence but cannot yet identify the LandClim-selected chronology or historical basin series. The Solsø source-lineage review retains a different sample-selection question: LandClim II names 79 pollen samples, while the Neotoma site has 92 source samples and its published age span alone retains 90. The Nordic observed-pollen scores do not bypass those gates.

Our Dalmutladdo Norwegian core review (reports/science/dalmutladdo-lra-input/README.md) matches 72 published, LandClim II and Neotoma samples and all 11 published radiocarbon source controls to one physical core. Its source records report 7.2 ha, 8.00* ha and a radius value of 160 with unstated unit; two printed calibrated age intervals exclude their own centres. Historical basin geometry and the exact count/taxon/PPE use remain unqualified, so this lead also has zero admitted LRA samples.

The existing 31-target source ledger assigns 46,477 of 47,085 native Dalmutladdo pollen counts; 600 are excluded and eight remain unmapped. That closed pollen accounting does not establish an LRA denominator.

For Finnish Laihalampi, the LAIH core and 150 Neotoma source depths match the LandClim II sample count. LandClim I reports 25.50 ha, close to the conditional 25.52 ha area of a 285 m circular radius. The original lake study describes a five-date fitted chronology. The exact selected ages and historical basin series are still unqualified; our source account is at reports/science/laihalampi-lra-input/README.md. No Finnish LRA sample was admitted from this correspondence.

The same source ledger assigns 84,078 of 85,303 native Laihalampi pollen counts; 916 are excluded and 309 remain unmapped. This pollen-only mapping does not qualify the core for LRA.

The HYDE source result checks modeled cropland quantity accounting at selected labels. It is not an observed vegetation series. The vegetation reference review currently admits zero matched measurement pairs for accuracy scoring.

A named lake and nearby archaeology

We resolved the proposed Bjäresjösjön name to the modern official Bjäresjö lake registry record. The source screen finds 371 SEAD display intervals from 13 sites within 20 km and reproduces all 25 existing modeled-window context counts. The nearby Neotoma pollen site and LandClim BJARE sequence remain different, unconfirmed core identities. SEAD chronology remains display context without qualified event propagation. We have therefore withheld a lake-specific ecological join, archaeology weighting and LRA basin transfer.