Read the scientific accounts offline¶
Download a paper account or an audience briefing, then open it in a PDF reader. Each file includes its explanation, evidence references and limitations. The website build serves the same reviewed bytes as the repository's retained exports. It does not require access to private source data to open these files.
The accounts distinguish published findings, our numerical reproduction, development pollen predictions and independent vegetation accuracy. Current Nordic predictions have no admitted matched vegetation reference; their pollen scores do not measure vegetation accuracy. Read papers and evidence roles before interpreting a reconstruction or choosing a method.
Main-paper accounts¶
| Question and evidence boundary | |
|---|---|
| LandClim II | Regional reconstruction, a worked count-to-result example and the unresolved historical archive correspondence. |
| LOVE and LOVEoptim | Local reconstruction and the difference between the published simulation and later numerical fixtures. |
| REVEALS theory | Regional cover assumptions; original simulation inputs and full primary-text review remain open. |
| LOVE theory | Local vegetation and regional background; original scenario reproduction remains unavailable. |
| Hellman regional validation | The original ten-lake validation and the source fields needed for a new measured-cover comparison. |
| Hellman sampling design | Sampling and parameter sensitivity; original scenario membership and numeric outputs remain unavailable. |
| Southern Sweden validation synthesis | Nine rounded figure labels from two regional compositions, with explicit limits on their interpretation. |
| Småland historical maps | Map coverage, reused pollen anchors and missing numerical vegetation fractions. |
| Namibia modern reference | Ten soil-pollen sites, repeated vegetation radii and a separate exposed development comparison. |
Partial original article access is identified in the corresponding account. The PDFs reproduce our canonical accounts; they are not copies of the original articles or a grant of rights to their source data.
Audience briefings¶
| Intended use | |
|---|---|
| Project introduction | The scientific problem, source roles, one worked finding and how a collaborator can help. |
| Scientific evidence | Concepts, assumptions, published versus reproduced versus evaluated outcomes, refusals and exact source/review needs. |
| Contributor workflow | Installation, data/result ownership, a bounded runnable export example, checks and troubleshooting. |
Editable sources are LaTeX. PowerPoint files have not been delivered; the intended editable collaboration format remains a decision for the readers. All PDF pages were inspected by a model for layout and font embedding. Actual colleague reading, qualified scientific review and accessible tagged-PDF reading remain open.
Use the downloaded files¶
Save a file before disconnecting; its main text, tables and diagrams remain readable offline. Citation and repository evidence links require a connection. The paper PDFs use commit-pinned evidence links whose public availability has not been verified for this local candidate. Follow results and model use for current scope, or the repository's source requests for the exact missing data and reviewer decisions. A missing measurement is not a zero accuracy score.