EML Section Guide

A reference for where metadata lives in EML and how the FSDB Data Catalog module uses it. Use this guide to locate sections in EML files (e.g. CF002_24.xml) and to see which fields drive the catalog search and detail pages.

Schema reference: EML 2.2.0 schema · eml.xsd

---

1. EML document structure (major sections)

An EML document has a single root <eml> and one primary resource. For data packages, the structure is:

XPath note: with the EML namespace bound as eml, paths look like /eml:eml/eml:dataset/....

---

2. Dataset structure (major and minor sections)

Under dataset, sections appear in a defined order. Not all packages include every section.

Major sections (direct children of dataset)

Minor sections (nested inside major ones)

Examples the catalog uses:

• Coverage: coverage/geographicCoverage, coverage/temporalCoverage, coverage/taxonomicCoverage, boundingCoordinates, rangeOfDates, calendarDate.
• Methods: methodStep, description, instrumentation, sampling.
• Project: title, personnel, abstract, funding, award, studyAreaDescription.

---

3. Data tables (entities) — dataTable

Each data table is one data entity (e.g. one CSV or similar). In EML these are dataTable elements under dataset. The catalog uses them for search, the detail page “Metadata” section, and download links.

Where data tables live

• Path: dataset/dataTable (multiple; often with id e.g. CF00201, CF00202).
• Example: CF002_24.xml has six dataTable elements (CF00201–CF00206).

Structure of each dataTable

Other entity types (same level as dataTable)

• `otherEntity` — Non-tabular data (e.g. documents, scripts). Same idea: identification, physical, optional attributeList.
• `spatialRaster` — Raster spatial data.
• `spatialVector` — Vector spatial data.

The FSDB catalog detail page and download logic focus on `dataTable` and its physical/distribution/online/url.

---

4. Search index — catalog page (/fsdb-data-catalog)

Search matches the following EML fields (paths relative to dataset unless noted):

DBCODE is also matched, derived from the filename convention DBCODE_VERSION.xml.

---

5. Detail page — dataset record (/fsdb-data-catalog/[DBCODE])

Header (summary row)

The top of the detail page shows a compact summary. People are shown in this order (left to right):

• Creator(s) — From dataset/creator only (no duplicate from associatedParty).
• Principal Investigator — From dataset/associatedParty where role is "Principal Investigator".
• Additional — From dataset/associatedParty where role is empty or unspecified.
• Other Researchers — From dataset/associatedParty where role is "Other Researcher".
• Any other associated-party roles appear after these, in a stable order.

Associated parties with role "Creator" are not shown again in the header (they are covered by Creator(s)).

Other header fields include Organization, dates, version, pub date, EDI Package ID (clickable link to https://portal.edirepository.org/nis/mapbrowse?packageid={packageId}), Source XML, and DOI when present.

Sections and EML sources

---

6. Notes

• Not every EML file has every section; missing sections are simply omitted on the detail page.
• UI labels may differ slightly from EML tag names.
• If expected metadata does not appear, check that it exists in the XML at the paths above. For XPath in code, the EML namespace must be registered (e.g. prefix eml).
• EML 2.1.1 and 2.2.0 use different namespace URIs; the catalog supports both for parsing.