What is the NGMDB and what does it contain?

The National Geologic Map Database is the USGS catalogue of geoscience maps for the United States, mandated by the Geologic Mapping Act. It indexes tens of thousands of maps and reports with metadata, scanned images, and increasingly GeMS-formatted GIS data, and is searchable by location, scale, theme and publisher.

How do I search the NGMDB for geological maps of an area?

Use the NGMDB MapView interface to pan to your area and see footprints of available maps, or the catalogue search to filter by state, scale, theme and author. Each record links to the scanned product and, where available, downloadable GIS data and metadata.

What is the difference between a scanned map and GeMS data?

A scanned map is a raster image of a printed sheet: you can view and georeference it but not query it. GeMS data is a structured GIS database (the USGS Geologic Map Schema) with map units, contacts and faults as features carrying attributes, confidence and source, so it can be analysed directly.

How do I judge whether a discovered geological map is usable?

Check the publication scale against your intended use, the publication date and whether newer mapping supersedes it, the CRS and whether GIS data exists, and the mapping confidence on key boundaries. A 1:250,000 map is reconnaissance, not a site-scale boundary.

NGMDB and Geological Map Discovery

Short answer

The National Geologic Map Database (NGMDB) is the USGS-run catalogue that makes United States geological maps findable, and it is the first place to look before commissioning new mapping or building a GIS over an area. It indexes tens of thousands of maps and reports, lets you discover them spatially through its MapView interface, and increasingly serves structured GIS data in the USGS Geologic Map Schema (GeMS) rather than just scanned sheets. Knowing how to search it — and how to vet what you find by scale, date, and data format — turns "we need a geological base" from a fieldwork project into a discovery task.

What the NGMDB is and why it exists

The NGMDB was mandated by the National Geologic Mapping Act and is maintained by the USGS together with the Association of American State Geologists. Its purpose is durable discoverability: geological maps are expensive to make, they remain scientifically valuable for decades, and without a catalogue they become invisible the moment the original authors move on. The database brings together products from the USGS, state geological surveys, universities, and journals into one searchable index, each with metadata pointing to the actual product.

Two practical entry points matter:

The catalogue search, which filters records by state/area, publication scale, theme (geology, geophysics, hazards, geochemistry), author, and publisher.
MapView, a map interface where you pan to your area of interest and see the footprints of every indexed map covering it, colour-coded and filterable by scale. This is the fastest way to answer "what mapping already exists here, and at what scale?"

A related component, the US Geologic Names Lexicon (GEOLEX), resolves stratigraphic unit names — useful when a discovered map uses a unit name you need to reconcile with neighbouring sheets.

Scanned image vs GeMS data — the distinction that governs reuse

What you can do with a discovered map depends entirely on its format, and the catalogue mixes two very different things.

A scanned product is a raster image (often a GeoPDF or TIFF) of a printed sheet. You can view it, print it, and georeference it into a GIS as a backdrop, but you cannot query a unit, select a fault, or run analysis — it is a picture of data, not data. To use it analytically you would have to digitise it, inheriting all the interpretation that re-drawing implies.

GeMS data is a structured GIS database. GeMS (the Geologic Map Schema) is the USGS standard format for digital publication of geologic maps, and it models a map the way a geologist thinks about it:

MapUnitPolys — the polygon units, each linked to DescriptionOfMapUnits for the authoritative unit text, age, and lithology.
ContactsAndFaults — the line features, carrying type plus ExistenceConfidence, IdentityConfidence, and LocationConfidenceMeters, the numeric horizontal uncertainty of the line.
DataSources — citations so every feature traces back to where it came from.

GeMS ships as a file geodatabase or, increasingly, GeoPackage, so it opens directly in QGIS or ArcGIS. The difference is night and day: with GeMS you can select all Qal polygons, buffer a fault, or join unit descriptions in one step; with a scan you are reading colours off an image.

A discovery and vetting workflow

Suppose you need a geological base for a corridor study in the western US.

Open MapView, pan to the corridor. Note the footprints covering it and their scales. You may find a 1:24,000 quadrangle over part of it and only a 1:250,000 sheet over the rest — a coverage and consistency issue you want to know up front.
Open the most relevant records and check, for each: publication date, scale, publisher, and whether downloadable GIS data (GeMS) exists or only a scan.
Prefer the largest scale and most recent mapping for the parts that matter, but record where you are stitching different scales together — that seam is a real interpretation limit.
Download GeMS data where available. In QGIS, add the GeoPackage/geodatabase, confirm the CRS (ST_SRID-equivalent: check the layer CRS, and reproject with the Reproject Layer algorithm or ogr2ogr -t_srs EPSG:32611 out.gpkg in.gpkg if you need a common UTM).
For scan-only areas, georeference the GeoPDF and treat it explicitly as a backdrop, not as queryable data, until/unless it is digitised.
Reconcile unit names across sheets using GEOLEX where adjacent maps disagree on nomenclature.

Judging usability: scale, date, confidence

Three checks decide whether a discovered map is fit for your purpose.

Scale. Publication scale is the honest statement of detail. A 1:250,000 map is reconnaissance — a contact on it may be hundreds of metres from its true position. A 1:24,000 quadrangle supports much finer work. Never over-zoom a small-scale map and treat its lines as precise; the source scale travels with the data's meaning.

Date and supersession. Older mapping is not automatically wrong, but check whether newer mapping covers the same ground. The NGMDB record dates and footprints make this visible.

Confidence. With GeMS data you get this explicitly — ExistenceConfidence and LocationConfidenceMeters tell you which boundaries are surveyed and which are inferred or concealed. Carry those attributes into your analysis and your cartography so an inferred contact is never styled or trusted like a measured one.

QA and validation

Verify scale matches use. A site decision needs large-scale mapping; do not stretch a regional sheet to do it.
Confirm the CRS of downloaded GIS data before joining or measuring, and reproject explicitly rather than relying on on-the-fly reprojection alone.
Check topology of GeMS data — that contacts close polygons and faults connect — before analysis, especially after reprojection.
Preserve source and confidence attributes through every processing step; losing DataSources or confidence fields severs the map from its provenance.
Document seams where you combine maps of different scale or vintage.

Common pitfalls and why they happen

Treating a scan as data. A raster image cannot be queried; mistaking it for GIS data leads to digitising errors and false precision.
Over-zooming a small-scale map. A 1:250,000 line displayed at street scale implies an accuracy the survey never claimed.
Dropping confidence attributes. GeMS records location uncertainty explicitly; discarding it turns inferred contacts into apparent certainties.
Ignoring supersession. Using an old sheet when newer mapping exists because the catalogue returned it first.
Stitching scales silently. Combining 1:24,000 and 1:250,000 mapping without flagging the seam misrepresents where detail actually exists.

Bathyl perspective

We start geological GIS work by exhausting what already exists, and for US projects the NGMDB is the first stop. When we ingest a discovered map we keep its scale, date, source citation, and per-feature confidence intact, so the base layer remains as scientifically careful as the original survey while becoming something a team can query, overlay, and reason about.