finding-gis-data

Finding GIS Data

Some general advice for searching the Internet for GIS data, and some recommendations of top sites for specific types of data.
Updated 2025-04-07.

Keith Jenkins kgj2@cornell.edu (GIS Librarian, Mann Library)
Jo Klein jo.klein@cornell.edu (Social Science Research and Spatial Data Librarian, Fine Arts Library)

https://cornell-gis.github.io/finding-gis-data/

Geolode

• Start here: https://geolode.org/
• Search/filter by place or broad topic
• Global data is tagged as “Earth”

Finding GIS Data via Google, etc

• Try the obvious… search for something like: GIS data {placename}
• Some places (like New York state) may have both a data website AND a geodata website
• Data is created and managed at different administrative levels (country, state, county, city, etc.)
• Try both the city AND the county (example, https://geolode.org/?q=angeles )

Parcels

• Often maintained at the county level, possibly state level.
• There are many online map viewers that let you look at parcel boundaries and get info for specific addresses, but no way to download the whole dataset.
• Sometimes there may be a separate website with downloadable data. But other times, the county may only sell the data or may not make it available at all!
• Map viewer example: https://mapping-tompkins-tompkinscounty.hub.arcgis.com/
• Data download example: https://tcdata-tompkinscounty.opendata.arcgis.com/
• Statewide parcels: https://gis.ny.gov/parcels

OpenStreetMap

• Most complete street dataset of the world (better than Google, Apple, Bing), but also has data for many other types of features.
• Preview the data at https://www.openstreetmap.org/
• Country-based shapefile downloads: https://download.geofabrik.de/
• Search for specific features: https://overpass-turbo.eu/
• Thematic extracts for city-sized areas: https://export.hotosm.org/
• OSM can sometimes be a good source for many of the types of data listed below. What data it contains is usually very good quality, but it is not necessarily complete for a given area.

Building footprints

• OpenStreetMap has a lot of buildings, but can be incomplete for many areas
• Microsoft and Google have been using AI to extract building polygons from imagery
• Microsoft: https://github.com/microsoft/GlobalMLBuildingFootprints/#will-there-be-more-data-coming-for-other-geographies
• Google: https://sites.research.google/open-buildings/
• Some places may have their own building inventories.

Boundaries

• GeoBoundaries has collected the best publically-available data for each country: https://www.geoboundaries.org/
• US Census Shapefiles: https://www.census.gov/cgi-bin/geo/shapefiles/

Transportation

• Streets are often maintained at the state level
• Complete layer vs major roads + minor roads
• OpenStreetMap is the best choice for many countries

Landcover

• US NLCD (30m): https://www.mrlc.gov/data
• US Cropscape (30m): https://nassgeodata.gmu.edu/CropScape/
• GlobCover (300m): http://due.esrin.esa.int/page_globcover.php
• ESA WorldCover (10m): https://viewer.esa-worldcover.org/worldcover/

Elevation

• US NED 30m and 10m: https://cugir.library.cornell.edu/?q=ned
• Global SRTM 90m: https://srtm.csi.cgiar.org/srtmdata/
• Global SRTM 30m: https://dwtkns.com/srtm30m/ (NASA Earthdata login required)
• Higher-resolution data (1m to 5m) is occasionally available for some localities. Many of these, plus global elevation datasets, can be found via: https://opentopography.org/

Hydrography

• US NHD has streams, waterbodies, watersheds, etc: https://apps.nationalmap.gov/downloader/#/

Business Points

• AllThePlaces has location data for specific restaurant/store chains, scraped from company websites: https://www.alltheplaces.xyz/
• To get point locations for all businesses in a given location, there are a couple websites that Cornell Library subscribes to:
  • Simply Analytics: http://resolver.library.cornell.edu/misc/6168667
  • Mergent Intellect: http://resolver.library.cornell.edu/misc/8058953
  • PolicyMap: http://resolver.library.cornell.edu/misc/8068928 (demographic, economics, health data too)

Infrastructure

• HIFLD has schools, hospitals, radio towers, rail crossings, public venues, etc.: https://hifld-geoplatform.hub.arcgis.com/
• Check HIFLD data for quality – sometimes there are aggregation errors, as they compile this data from many different sources. Sometimes local sources are better or more recent.

ArcGIS Online

• ArcGIS Online hosts many commonly-used datasets, as well as more unique datasets. Use the search at https://arcgis.com/
• Many datasets are tagged as “Authoritative” (typically from government sources) or “Living Atlas” (curated by Esri)
• Other datasets may be from random ArcGIS Online users – be more cautious about these, and try to assess the legitimacy of the source.

General advice

• For other types of data, try Geolode, then try Google.
• Try searching for shp instead of GIS data
• Try including the term download
• For countries where English is not the main language, search in the local language (example, Spanish SIG datos) and be sure to include any accents (Sénégal)
• Specialized data (like NYC street/sidewalk curbs) that you’ve seen for one location may not be available for other locations.
• Keep notes of where you download various files – it’s not always possible to figure out the source from the downloaded files.
• Whenever possible, download any documentation or metadata at the same time you download the data.
• Often there are multiple sources for the same data – put it all on a map and compare specific locations, in terms of geometries as well as attributes. Pick the data source that seems best for your project.
• Keep looking! But also know when to give up – the data may not exist.
• Feel free to ask for help if you get stuck: https://guides.library.cornell.edu/gis/help