Bijux Pollenomics

bijux-pollenomics connects curated evidence to public maps and reports about pollen, palaeoenvironmental context, archaeology, hydrography, fieldwork, and ancient DNA. Its database preserves captured source identity, family-specific preparation, scientific decisions, publication membership, and the gaps that prevent stronger claims.

Start Here

Open the product guide Open the data guide Learn the domain language Inspect the evidence database Open the report portal How to read the report tree Open the atlas guide Inspect fieldwork evidence

Question	Governing route	Identity to retain
What does the product implement?	product guide	runtime version and product-scope contract
What does a database term mean?	domain language	object type, stable ID, and claim dimension
How was evidence acquired and prepared?	data system	source family, source member, stage, and data revision
Why was a record admitted, qualified, or refused?	curation	governed object, decision, proposed use, and reason
What belongs to a public product?	report portal	manifest, member ID, geography, role, and caveat
What does an atlas marker establish?	Nordic atlas	feature ID, point class, coordinate posture, and time posture
What supports a lake ranking?	lake priorities	lake ID, model, scenario, candidate population, and readiness state

Start From The Object You Actually Have

The shortest trustworthy route depends on what is already in hand. A map marker, a source record, and a scientific statement begin at different authorities even when they eventually resolve to the same evidence graph.

Starting object	First check	Then establish
map marker or table row	product manifest and member identity	admission reason, governing claim, and source locator
source accession, DOI, or registry ID	source-family capture and native member	normalized identity, review posture, and every consuming product
sample or site label	governed identity and aliases	project or source lineage, place and time claims, and ambiguity state
number in a report	observation unit, denominator, scope, and revision	exclusions, unresolved members, and aggregation rule
proposed cross-domain statement	role and observation unit for every family	spatial bridge, temporal bridge, independence, and refusal conditions

A visible feature is proof of product membership, not by itself proof of every scientific statement shown beside it. Conversely, a captured source record can remain valuable evidence without qualifying for any current public product.

Current Product Contract

The implemented runtime is an atlas builder and evidence-publication system. The broader research direction must not be read as a claim that the repository is already the full cross-evidence pollenomics engine.

Available now	Outside the current runtime claim
named source collection and source-preserving preparation	general cross-domain harmonization
governed objects, relations, conflicts, qualifications, and refusals	automatic reconciliation of unlike observation units
declared ranking models and sensitivity outputs	general causal or scientific inference
manifested world, regional, country, atlas, and fieldwork products	workflow-wide semantic replay and interpretation

product-scope and surface-map expose this boundary in machine-readable form. Planned behavior becomes product behavior only after it has an owned interface, state transition, governed output, and evidence-fitness contract.

From Source To Public Claim

flowchart LR
    Source["source release, paper, archive, registry, or API"] --> Capture["captured identity and material"]
    Capture --> Evidence["normalized objects and evidence relations"]
    Evidence --> Decision["claim-specific review and admission"]
    Decision --> Manifest["product manifest and member"]
    Manifest --> View["map, table, report, or field record"]
    View -. "trace backward" .-> Evidence

The chain is reversible. A reader can move backward from a visible member to its decision, evidence, capture, and upstream identity. A source correction moves forward through affected objects, decisions, manifests, and views.

The wheel supplies producer behavior; data/ supplies governed evidence state; report/ supplies checked-in publication state. A package version, data revision, and product manifest answer different reproducibility questions.

How A Claim Earns Trust

Resolve five questions for a consequential claim:

1. Identity: which source, evidence object, decision, and product member is being discussed?
2. Meaning: what observation unit and evidence role does the object carry?
3. Space: what geometry, basis, method, and precision are supported?
4. Time: what source expression, evidence class, interval, and comparability posture are supported?
5. Membership: why did the named product admit, qualify, exclude, or refuse the object?

No single map popup answers all five questions. Use the popup for orientation, then follow its stable identity through the manifest and governing evidence.

Claim	Minimum supporting packet	Insufficient substitute
a source object was captured	source identity, version, retrieval context, member locator, and digest	citation or filename alone
two records identify the same object	stable identities, typed relation, evidence locator, and resolution decision	matching labels or nearby coordinates
a location is exact	locality evidence, coordinate provenance, method, precision, and conflict outcome	plotted point or decimal count
a time comparison is numeric	evidence class, common basis, normalized interval, precision, and comparability decision	contextual period label
a record belongs in a product	governed evidence, admission decision, scope, manifest membership, and caveat	presence in a normalized file

Evidence Surfaces

Surface	Answers	Does not answer
source families	what entered, under which identity, role, and access conditions	record-level publication fitness
database	objects, relations, fact ownership, revisions, and coherent state	whether every object belongs in a product
evidence	identity, locality, coordinate, chronology, taxonomy, and join support	universal comparability across domains
curation	claim-specific conflicts, decisions, recovery, admission, and refusal	new source-native facts
publications	versioned scope, members, non-members, caveats, and renderings	stronger evidence than the database contains
atlas	role-aware spatial comparison and traceability	association, contemporaneity, or causation from proximity
fieldwork	dated visits, locations, media, and bounded observations	lake-wide conditions or sampling readiness

Read Counts As Typed Claims

Counts are meaningful only with their observation unit, population, scope, and revision. Captured rows, normalized objects, reviewed claims, eligible members, published members, map features, and display aggregates are not interchangeable denominators.

An example from animal evidence illustrates the difference: a project recovery count, species-owned sample-foundation count, and admitted atlas-point count describe different governed populations. Their disagreement is expected until an identity-level reconciliation proves otherwise.

Before reusing a number, retain:

• observation unit and stable identity namespace;
• source or database revision;
• geographic, temporal, taxonomic, and product scope;
• eligibility, exclusion, unresolved, and missingness rules; and
• the manifest or review surface that owns the denominator.

Current Integrity Boundaries

The public products remain deliberately smaller than the collected evidence:

• animal sample, locality, chronology, coordinate, and source-recovery gaps remain qualified, excluded, or release-blocking;
• SEAD currently supports inventory and spatial context, not numeric temporal comparison;
• RAÄ is Sweden-specific and supplies no equivalent Nordic registry coverage;
• modern boundaries frame publication scope without adding scientific weight;
• lake rankings are evidence-richness decision support, not field readiness or coring-site selection; and
• field visits document bounded observations without validating nearby layers.

These are product facts, not footnotes. Start with the release refusal and relevant review surfaces before strengthening public language.

Reproduce Or Challenge A Result

1. Name the product manifest and stable member.
2. Recover the admission decision and governing evidence IDs.
3. Inspect source identity, locality, coordinate, chronology, role, and caveat.
4. Confirm the database and runtime revisions used by the product.
5. Recompute only through the owner of the disputed transition.
6. Compare identities, semantics, decisions, populations, and manifested descendants—not only files or rendered appearance.

For installed behavior, continue to operations. For tracked evidence, continue to Pollenomics Data. For checked-in products, continue to the report portal.