Address Geocoding with ArcGIS Pro
Geocoding is the overall process of converting location descriptions (such as addresses, place names, or postal codes) into geographic coordinates (latitude and longitude) so they can be plotted on a map. Geocoding can involve different types of location data, including:
- Addresses (street addresses, ZIP codes, tec.)
- Place names (e.g. city names, landmarks)
- Coordinates (e.g. reverse geocoding from coordinates to addresses)
Geocoding encompasses:
1 - Address matching
2 - Reverse geocoding
3 - Geocoding based on place names or other location descriptors.
Applications of geocoding
Location Analysis: Geocoded data is crucial for spatial analysis, such as identifying customer locations, mapping crime incidents, study disease spread, map patient data, allocate resources based on geographic location, or planning service areas.
Routing and Navigation: Accurate geocoding enables effective routing and navigation applications, essential for logistics and emergency response.
Data Integration: Integrate geocoded data with other spatial datasets for comprehensive analysis.
Address matching is a core part of the geocoding process. The workflow focuses exclusively on matching street addresses (or parts of addresses) from a table or dataset to a spatial reference dataset, such as a road network or parcel database. The workflow involves creating an address locator, geocoding the address, and refining the results to ensure accuracy. The result is geographic coordinates for each matched address.
Steps to Perform Geocoding in ArcGIS Pro
1 - Prepare the Data
- Input Address Data: The starting point is a table of addresses. This table typically contains field like street addresses, city names, postal codes, and other location details.
- Reference Data: To match addresses to geographic locations, you need a reference dataset such as a street network, parcel data, or address point. This data will be used by the locator to find corresponding locations for each input address.
If you do not have a reference dataset (e.g. street centerline data, address points, parcel data (cadastral data), building footprints), you can use Esri's World Geocoding Service to geocode the addresses from your Excel or CSV file. This method is simple and efficient, especially when you don't have access to local reference datasets. See our article https://support.tribalgis.com/article/60-address-geocoding-with-arcgis-pro-case-study?preview=66c7a87a16e6a80d1e1d51b0 for more information.
2 - Create an Address Locator with a reference dataset
An address locator is a geoprocessing tool that defines how addresses in the input data will be matched to reference data. To create an address locator, you need a reference dataset (e.g. a feature class containing street data or a table of address points). ArcGIS Pro supports various locator styles, including Single House, Dual Range, and Point Address.
- Got to the Geoprocessing pane and search for the Create Address Locator Tool.
When you search for the Create Address Locator tool multiple tools will be displayed.
Create Locator
This is the most commonly used tool for creating an address locator based on a single reference dataset (e.g. street network, parcel dataset, or address points). It generates an address locator that can be sued to geocode addresses based on your reference data. You can use this locator to convert addresses into geographic coordinates.
Use Create Locator when you have a single source of reference data for geocoding, such as a street centerline or parcel dataset.
Create Composite Address Locator
This tool is used when you want to combine multiple locators (created from different datasets) into a single composite locator. It's helpful when you have several sources of reference data (e.g. one for the street, another for the parcels, and another for points of interest). It allows you to combine multiple locators (e.g. one for street addresses and another for place names) into one, so when you geocode, it searches across all locators.
Use Create Composite Address Locator if you want to combine several address locators from different datasets for a more comprehensive geocoding solution. For instance, if you have a dataset for streets and another dataset for POIs.
Create Feature Locator
This tool is used to create a locator for geocoding non-address features such as names of locations (e.g. parks, schools, or landmarks). This is often used for place-based searches rather than addresses. It creates a locator for feature-based attributes like place names rather than street addresses.
Use Create Feature Locator if you are geocoding by place names or features instead of addresses.
- Select the locator style (e.g. Single House, Dual Range, Point Address) and input the necessary reference data.
- Configure the locator fields: define the fields that correspond to address components in your data (e.g., street name, house number).
- Run the tool to generate the address locator. Once created, the address locator can be reused for future geocoding tasks or shared with other users.
Single House
A Single House address locator style is designed to match individual house addresses to specific points. Each house address corresponds to a unique location, usually based on a specific point dataset where each point represents a building or parcel. For example, if you input "123 School St," the locator will find the exact point of that house on the map. This style is ideal for geocoding applications where you have precise address points, such as residential properties or small-scale areas where each address has its own location.
Dual Range
A Dual Range address locator style is based on street ranges where both sides of the street (left and right) have different address ranges. This means that the locator can make the difference between odd and even addresses on either side of the street and match an address to the correct side. If you input "123 School St," the locator will use the street's address range (e.g. 100-150 on the left side, 101-150 on the right side) to find the correct side of the street and estimate the position of the house along that segment. This style is useful when working with datasets that represent streets and address ranges but lack precise point data.
Point Address
A Point Address locator style matches addresses to specific points, but these points are typically more generalized compared to Single House locators. In this case, the point represents the location of a parcel or property, rather than the exact house location. Inputting "123 School St," will return a location based on the property centroid or the most significant point within a parcel boundary. This style is effective when you have parcel-based or address points that approximate the location of properties but do not require high precision (e.g. for larger areas like rural properties).
3 - Geocode the Addresses
With the address locator ready, use the Geocode Addresses tool to start the matching process. You will input your table of addresses and specify the address locator.
- Open the Geocode Addresses tool from the Geoprocessing pane.
- Choose your address locator and input the table of addresses.
- Configure the input and output settings, such as specifying the address fields (e.g. street name, house number, etc.) and output feature class.
- Execute the tool to perform geocoding.
The tool compares each address in the table with the reference data and assigns a set of geographic coordinates (e.g. latitude and longitude) to each matched address. The result is a feature class or table containing the matched locations.
4 - Review and Refine Results
- Output Results: The tool will generate a feature class of points or a table with coordinates corresponding to the matched addresses, However, not all addresses may be matched perfectly or at all. Inspect the geocoded output for any unmatched or poorly matched addresses.
- Match Scores: Each geocoded address is assigned a match score indicating the confidence of the match (usually between 0 and 100). Higher scores indicate a better match, while lower scores indicate possible errors.
- Unmatched or Tied addresses: Some addresses may not match the reference data (unmatched) or may have multiple potential matches (tied addresses). These require further review and refinement.
- Use interactive rematch tools to manually review and correct unmatched addresses.
Interactive Rematch Tools
These tools are designed to help users manually review and correct addresses that the geocoding process could not match automatically or matched with low confidence. These tools allow for a detailed examination and adjustment of geocoding results to ensure the highest accuracy.
1 - Rematch Addresses Pane
In the Contents pane, right click on your geocoded layer, select Data and Rematch Addresses.
This pane provides a user-friendly interface for reviewing and rematching addresses. It lists all unmatched or tied addresses, allowing you to select each one and manually correct it.
2 - Unmatched Addresses
These are addresses that the locator could not match to any location. You can use the rematch pane to manually search for the correct location by entering new search criteria or selecting from a list of potential matches by the locator.
3 - Tied Addresses
Tied addresses are addresses that have multiple potential matches with similar match scores. Review the list of candidates for each tied address and select the most appropriate match based on your knowledge or additional information.
4 - Partial Matches
Partial Matches are addresses that are matched with lower confidence, often due to incomplete or ambiguous information. Evaluate the match candidates and adjust the address or match criteria to improve the match quality.
Using the Rematch Addresses Pane
- In the Contents pane, right click on your geocoded layer, select Data and Rematch Addresses. This opens the rematch pane, displaying unmatched, tied, and partially matched addresses.
- Select an address from the list. The pane shows the current match status and potential candidates.
- Use the map to visually inspect the location of each candidate, ensuring it aligns with your expectations.
- If the address is unmatched, manually enter search terms or adjust the address to find better matches.
- Each candidate match has a score indicating the confidence level. Higher scores usually represent better matches.
- Evaluate the score and spatial location of each candidate. If necessary, adjust the input address to refine the search.
- For tied addresses, choose the best candidate and confirm the match.
- For partial matches, refine the search criteria and reassess the candidates before confirming.
- After rematching, save the changes to update the geocoded feature class or table with the corrected addresses.
Both Status and Match Type (Attribute Table) provide important information when reviewing geocoding results.
Status usually refers to whether the address was matched ("M" for matched, "U" for unmatched.
Match Type further specifies how the address was matched, such as automatically (A), manually (M), or to a candidate (CM).
| Status/Match Type | Description | What it Means | What to Do |
| A | Automatic Match | The address was successfully matched automatically with high confidence | No action is usually required. The match is reliable, but you can review it if needed. |
| CM | Candidate Match | The address was matched to a candidate, but manual review or confirmation is needed. | Review the list of candidate matches and select the best one manually. |
| M | Manual Match | The address was manually matched by the user. | No further action is needed. This indicates a manual match you've already confirmed. |
| T | Tied | Multiple possible matches exist (tied candidates) that need manual selection. | Review the tied candidates and choose the correct match. |
| U | Unmatched | The address could not be matched to any location in the reference data. | Review the unmatched address, correct it, or manually place the point on the map. |
| PP | Parcel Point/Partial Parcel | The address was matched to a parcel location, often the centroid of a parcel. | Review the match for accuracy. If parcel-level precision is acceptable, keep it; otherwise manually adjust the location. |
| SP | Street Point | The address was matched to a specific point on a street. | Typically, no action is needed. This match provides good precision for street-level geocoding. |
| R | Rejected | The address was manually rejected by the user during the rematch process. | No further action is needed unless you want to revisit the rejected address. |
Reverse Geocoding
Reverse geocoding is the process of converting geographic coordinates into human-readable addresses. This can be particularly useful in applications requiring the identification of addresses from map clicks or GPS coordinates.
- Use the Reverse Geocode tool to convert coordinates back into addresses.
- This tool can be used interactively on the map or with a table of coordinates.
Overall Best Practices
Data Preparation: Ensure that the input address data is clean, standardized, and free of common errors such as misspellings or missing fields.
Reference Data Quality: Ensure that your reference data (e.g., streets or address points) is accurate and up to date to improve geocoding results. Outdated street or address data may result in unmatched addresses.
Address Standardization: before geocoding, it's useful to standardize address formats to match the conventions used by the reference dataset. For example, ensuring street types (e.g. "Ave" vs "Avenue") are consistent.
Batch Processing: For large datasets, consider batch processing geocoding task to save time.
Error Handling: Regularly review and refine geocoding results to handle unmatched addresses or errors.
Rematch Tools: Use the interactive rematch tools effectively to review low scoring, tied, or unmatched addresses and make necessary corrections.
Tips for Effective Rematching
Contextual Knowledge: Utilize local knowledge or additional datasets (e.g., aerial imagery, parcel data) to validate and improve match accuracy.
Iterative Process: Rematching can be iterative; revisit addresses of initial corrections do not yield satisfactory results.
Automation Tools: For large datasets, consider using scripting or batch processing to streamline the rematching process where possible.
We hope that this article has been helpful! If you have any feedback or questions, please feel free to send us an email or connect with us for a chat. The NTGISC team is here to assist you further!
https://resources.arcgis.com/en/help/main/10.1/index.html#//00250000000v000000
https://pro.arcgis.com/en/pro-app/latest/help/data/geocoding/fundamentals-of-creating-a-locator.htm
https://pro.arcgis.com/en/pro-app/latest/help/data/geocoding/create-a-locator.htm