Geocoding is the process of converting a textual representation of place into spatial coordinates. If you think of your home address, it consists of a house number, streetname, city, state, and zip code. Each of these pieces are important for the delivery of goods and services; or for the arrival of new friends to your door. We simply type the information into our phone or GPS and are presented with turn-by-turn directions to arrive at our destination. This process is routine and appears simplistic but relies on a series of spatial operations that start with geocoding. Our familiar addresses are not immediately translatable into a unique point when we want to represent locations on a map or use some network analysis to determine a route. Geocoding performs that translation and allows us to convert single addresses or even long lists of them (batch geocoding) into points in a spatial dataset that are useful for mapping or determining distance or travel time.
Geocoding requires a list of one or more addresses and an “address locator” that will convert that information into point locations. Address locators operate at levels of complexity ranging from a simple lookup table to a more complex analysis that interpolates the location of a house number along a linear street segment. Lookup tables are lists of address information with an associated set of coordinates representing the point to which the address refers. These lists can link Zip Codes to the centroid of that area or link full addresses to a point location based on the position along the roadway or the centroid of an actual property parcel of building footprint. Interpolation involves using linear street features that also contain attributes storing the starting and ending house numbers; the locator finds the correct street feature and then estimates the location of the target address number along the segment assuming even distribution of values. The level of detail (accuracy) available in an address locator is directly related to the cost of creating such a resource. Some locators apply a cascading model whereby the most accurate match is sought but if that is not successful, matches at decreasing levels of detail (i.e. zip code, city, etc.) are attempted until a match is obtained.
There are several resources available to the Washington University community for geocoding. See the Address Locators tab for descriptions and access instructions. The Geocoding Workflows tab presents step-by-step instructions for performing geocoding in several different software tools. For further details on geocoding in general and on specific tools, see the Documentation and Training tab.
Please contact Bill Winston or Data Services if you have any questions.
These tools convert textual addresses into point locations at varying levels of detail. WashU has several options available to accomplish this process; see below for descriptions and access instructions. Some locators are accessed via an API or network service and would require address information to be sent over the network; if your project requires HIPPA protection, this type of locator may not be appropriate. Other locators can be installed on researcher systems for direct access that does not require any network traffic; this is a suitable option for those projects that must protect confidential personal information.
Reference the “Geocoding Workflows” for instructions describing how to use these locators to perform individual or batch address geocoding.
Streetmap Premium - New online service available for Fall 2022
This product includes an address locator that is updated several times per year and is the most current 'in-house' address locator available to the WashU community. We have added an online geocoding service that accesses our most recent version of the Streetmap Premium dataset. This service is available to WashU community members who are logged into the secure campus network directly or via VPN. See this page for details.
The locator can also be accessed directly on one of the Research Studio computers or in some cases can be downloaded and installed on a researcher's lab or personal computer.
ArcGIS World Service (ArcGIS, ArcGIS Pro, and ArcGIS Online only)
This cloud based option is provided by ESRI as the default locator in ArcGIS Pro, ArcGIS Online, and ArcGIS Desktop as long as the user is logged into their ArcGIS Online account in the application and has sufficient credits to cover the 'cost' of the geocoding operation. Individual addresses can be located at no cost using the appropriate search function in any of the ESRI suite of software. Batch geocoding will require credits and this locator should not be used for projects requiring more than 4000 records.
Geolytics (2009 and 2016)
These locators use the Zip+4 format and are only able to match to the Zip Code level. The database was purchased from Geolytics in 2009 and again in 2016 and was used to create locators for each of those years. It would be an appropriate option for projects focused in or around that time. The locators can be downloaded and used without license from the Library's Geospatial Data Collection at this path...geocoding\zip_plus_4\Geolytics_20XX_Zip4.loc
Streetmap NA (2007)
This collection was provided with the ESRI Data and Maps collection in 2008 and includes locators that return results at several levels of accuracy ranging from address points to the city/state level. The locators can be downloaded and used without license from the Library's Geospatial Data Collection at this path...geocoding\
Online Geocoding Services
Several options exist for skilled users who wish to take advantage of an API for programmatic control of the geocoding process. Review provider documentation for details on these sources.
There are several applications that will perform the geocoding operation. A popular solution is to use one of the Esri GIS applications licensed by the University and the workflows listed below. See the Software Installation Guide for instructions to download and install available applications. It is also possible to geocode using QGIS or an online API but this requires more technical skill and hands-on attention.
Prior to conducting a geocoding operation, you should make an effort to clean your input data table to:
See below for step-by-step workflows or video demonstrations for some of the most common applications.
Intro to Geocoding (video - ESRI User's Conference 2019 - entry level)
Individual address can be found using the Locate - Search function and the default locator.
Batch processing relies on access to an address locator and WashU has several available; see the Address Locators tab for details and links.
Geocoding Workflow (ESRI online help)
Individual addresses can be found using the "Find" tool and the default locator.
Batch processing relies on access to an address locator and WU has several available; see the Address Locators tab for details and links.
AGOL will automatically geocode a table of addresses (or other locational information) when the table is added to a map page. Users can drag and drop comma separated value (csv) files into an AGOL map and point features will be added based on those records.The table must have columns containing locational information; coordinates, street addresses, city, state, or zipcode. You will be prompted to select the geocoding type and to specify the columns to use for the operation. This process does consume credits (40 credits per 1000 geocoded records) so this is not an appropriate solution for more than 4000 records. Contact Data Services if you have a large project and we will help you find the best geocoding solution.
Links to useful courses and videos: