Atlas reference
Atlas Handbook
Every layer on this atlas makes a claim, and no claim is self-evident from a coloured line. The Handbook is where each one is stated in plain language: what you are looking at, what it rests on, how much of it was drawn rather than found, and what may be cited.
Nothing here requires knowing how the site is built. The repository stays reachable from every page for anyone reproducing the work, and it is never the explanation.
How to read the Atlas
Three distinctions carry most of the weight. A context layer frames the map and asserts nothing about the past - present-day borders are context, and drawing them over historical material is an anachronism the atlas labels rather than hides. A historical layer is a reconstruction of a past state, and is the only kind of layer that can be cited as a claim about what was there. An evidence layer describes a source: a depicted extent tells you what a map shows, not what existed.
Where a line was drawn by this project rather than taken from a survey, the layer says so and the styling shows it. See the glossary for the vocabulary of confidence, provenance and review.
Sections
-
Layer catalogue
Every documented layer, what it claims and what it rests on.
27 records
-
Methods
How a reconstruction was made, written once and cited by the layers that use it.
1 record
-
Sources and evidence
Ledgers of instruments, witnesses and authorities behind the source-heavy layers.
2 records
-
Glossary
The words this atlas uses about evidence and uncertainty, and what each commits to.
1 record
-
Data fields
Field meanings and controlled vocabularies, for reading the map and reusing the data.
5 records
-
Editorial decisions
Choices that shape what you see, and the reasoning behind them.
1 record
-
Technical and reproducibility
Pipelines, schemas and release manifests. Not needed to read the atlas.
2 records
What is documented so far
27 of 27 published layers have a public record. The remainder are being migrated programme by programme; until then their evidence lives in the repository, which is exactly the situation this Handbook exists to end. The current gap list is published as machine-readable coverage.