Skip To Content

Publishing your imagery

Publishing elevation services

Publishing elevation services is similar to publishing all other image services. Find more information about using ArcMap and ArcGIS Server to publish image services. If you're using server raster functions , these should be defined during the publishing process.


The creation of a cache is not appropriate for services based on numeric values (elevation, slope, aspect), since the cache will be an 8-bit (grayscale single band or RGB color) visualization.

A cache may be appropriate for visualizations of elevation data (hillshade, for example, or color maps of slope or aspect). However, for representations such as a slope map or an aspect map, using image services is generally sufficient.

If you want to create a static cache of a hillshade or other cartographic representation to incorporate into other maps, the best way to do this is to (1) use the server caching, or (2) use the Manage Tile Cache tool. Once the cache is created, it can be directly used as a raster dataset or incorporated into a map cache.

For users who want to fully optimize the performance of elevation access, one potential option is to create a service that serves elevation but has Hillshade (for example) as the default raster function. To do this:

  1. Publish this service with caching set to ON and set the schema to ArcGIS Online with suitable levels.
  2. Set caching to Create tiles on demand and Update cache manually.
  3. After publishing the service, use Manage Tiles to pre-generate the cache for smaller scales and areas that users are expected to visit more frequently. When users access this service through ArcMap, they will be able to get fast access to the cached hillshade (or similar service).

    If users zoom to areas that have not cached, the appropriate areas will be cached on demand. Additionally, users will still be able to configure ArcMap not to use the cache, and will be able to use selectable raster functions to see the other defined functions and get data values.