Exploring ArcGIS REST Services¶

Data served via ArcGIS Servers support a REST API for accessing these services. This notebook provides some examples on how we can use Python to access these services using the requests package.

→ Full documentation on the ArcGIS REST API: https://developers.arcgis.com/rest/services-reference/get-started-with-the-services-directory.htm

NCOneMap's REST Endpoint¶

We'll explore NCOneMap's web services, accessed here: https://services.nconemap.gov/secure/rest/services.

• Begin by navigating to that page to familiarize yourself with the organization and listing of individual services, i.e. the Services Directory.
• You'll see a number of folders containing other services and below that a number of services listed there in the main page.
• Note the types of services listed: ImageServer, FeatureServer, MapServer. These services provide access to different types of data. Other services provide access to data processing capabilities.

• Next, click on the NC1Map_Census (MapServer) link to open the metadata for that map service.
• You'll see a number of properties for the map service including a Description, a list of Layers included in that service, Spatial Reference, and at the very bottom, a list of the Supported operations.
• We'll explore how we can use the Export Mapoperation to extract data from this service.

• Click on the 'Export Map' link at the bottom of the map service metadata page.
• Note the URL in the page that appears: This is a REST-based request. And the page that appears is a nifty form to help you edit and explore that REST-based URL.
• Change the bounding box values to: -79.3, 35, -77.9, 36.5
• Since those coordinates we just supplied are in NAD 83 geographic coordinates, set the bounding box spatial reference to the WKID of 4269

• Click Export Map (GET) to generate the URL reflecting our changes and display the results.

• Enter show:0 in the Layers: box to show only the first layer contained in this map service (1970 Census Boundary/Population). Upate the map by hitting Export Map (GET) again.

• Change the Format from HTML to Image and hit Export Map (GET) again.

• Copy the URL of the generated page and past between the quotes in code block below and run it. We've captured that same image!

Dissecting a Map Service REST request¶

Now let's dissect that URL into a service endpoint (its address) and a list of parmaters.

Below is the full url:

https://services.nconemap.gov/secure/rest/services/NC1Map_Census/MapServer/export?bbox=-79.3%2C+35%2C+-77.9%2C+36.5&bboxSR=4269&layers=show%3A0&layerDefs=&size=&imageSR=&format=png&transparent=false&dpi=&time=&layerTimeOptions=&dynamicLayers=&gdbVersion=&mapScale=&rotation=&datumTransformations=&layerParameterValues=&mapRangeValues=&layerRangeValues=&f=image

Everything up to the ? is the service end point (the server's internet address and the service location). Everything after that are parameters separated by &.

• Break those into components we can use with the requests package, i.e. a URL variable and a dictionary of parmeter name:value pairs.

• Now, let's import the requests package and execute our REST request.

Ok, it's just the same as above, but we can easily change values in our params dict and modify the output.