Work with image services in ArcGIS Pro

In ArcGIS Pro, you can visualize and analyze image services directly in your project. You can use functionality that is built into the image service, such as processing templates or temporal querying, or you can perform analysis using raster functions. In this guided tutorial, you will explore some of the many ways to work with image services in ArcGIS Pro using data from the ArcGIS Living Atlas of the World.

First, you will add global Sentinel-2 and Landsat data to a map from the ArcGIS Living Atlas of the World and examine the layer properties.

Next, you will explore built-in image service functionality by observing some interesting places around the world.

Finally, you will run additional processing on the data to determine changes in vegetation cover between two dates.

Add imagery to your project

You can visualize, analyze, and complete imagery tasks with image services in an ArcGIS Pro project. If you have access to image services from a specific server, you can follow along with the workflow below using your own data. Otherwise, you will work with two image services hosted on the ArcGIS Living Atlas of the World portal.

1. In ArcGIS Pro, create a project using the Map template and sign in to your ArcGIS Online account.
2. Click the Catalog pane in ArcGIS Proand click the Portal tab.

   On the Portal tab, you have four options: My Content, Groups, All Portal, and Living Atlas.

3. Click the Living Atlas button. In the search bar, type sentinel and press Enter.
   
4. Drag the Sentinel-2 Views image service to your map.
5. In the Contents pane, right-click the Sentinel-2 Views image layer and choose Properties.
6. Click the Metadata tab and read the description.
The description contains information that was published with the data. The Sentinel-2 data is multitemporal and multispectral, and by default it is displayed with the most recent cloud-free imagery on top. You can also read the descriptions of each band, and you can see that the imagery is analysis ready with top of atmosphere (TOA) correction applied.
7. Click the Mosaic tab.
The mosaic method, order field, order base value, and mosaic operator all determine how the imagery is displayed. By default, the Sentinel-2 imagery is mosaicked according to an attribute field named Best, which is a measure of quality and time, and it displays in ascending order where a lower value indicates higher quality and a more recent date. Click the drop-down menu to choose other attribute fields to sort the imagery.
8. Click the Processing Templates tab and click the Processing Template drop-down menu to see all the processing templates available with this image service.
A processing template is a raster function template that is published along with the image data to allow you to quickly apply the processing to the service in real time.

9. Click Cancel to close the Properties window.
10. In the Catalog pane, from the Living Atlas portal, search for the Multispectral Landsat image service and add it to your map.
11. In the Contents pane, right-click the Multispectral Landsat image layer and choose Properties. Explore the Metadata, Mosaic, and Processing Templates menus.
12. Click the Time menu.

    By default, the Time property for this image service is defined by a field named AcquisitionDate. This is the date and time when the Landsat image was captured. The time extent begins in 1972, the beginning of the Landsat program, and ends with the most recent capture date, which is updated daily.

13. Click Cancel to close the Properties window.

Work with Living Atlas content using built-in functionality

As you saw in the layer properties, image services are published with built-in functionality such as processing templates, temporal properties, and definition queries. If you are working with your own image service, you will see functionality that is different from what you work with on the ArcGIS Living Atlas of the World layers. Also note that all ArcGIS Living Atlas of the World layers have different built-in functionality based on the data that is available and typical use cases for the imagery.

1. Uncheck the Multispectral Landsat image layer in the Contents pane and click the Sentinel-2 Views image layer to select it.
2. On the Map tab, under the Inquiry group, click Locate and paste the following coordinates in the Locate search bar: 105.9999893°E 13.9254059°N.
3. From the scale list at the bottom of the map display, choose 1:100,000. With the Sentinel-2 Views image layer checked in the Contents pane, you see an image of the Mekong River at the border of Cambodia and Laos.
   
4. With the Sentinel-2 Views image layer selected in the Contents pane, a contextual tab set named Image Service Layer is visible on the ribbon, along with two subtabs named Appearance and Data. Click the Data tab.
5. Click the Processing Templates drop-down menu. The default display is Natural Color with Dynamic Range Adjustment (DRA). DRA automatically adjusts the image stretch based on the pixel values in the current display. Change the template to Agriculture with DRA and read the description. The display changes to one where bright green represents healthy vegetation, and brown represents bare earth. Change the template to Agriculture (without DRA) to see the difference the DRA can make on image display.
   
6. Click the Processing Templates drop-down menu again and change the template to each option in turn. Notice how you can visually identify different features in the display depending on the template you are viewing. You may need to apply stretching on the image so you can visualize the contrast more easily in the extent in which you are viewing. To do this, click the Appearance tab and click the Dynamic Range Adjustment button.
7. Turn off the Sentinel-2 Views image service layer and turn on the Multispectral Landsat layer in the Contents pane.
8. Use the Locate tool again and paste the following coordinates into the Locate search bar: 10°34′06″S 26°11′44″E. Zoom out to a scale of 1:100,000.
Here you can see the Tenke Fungurume Mine in the southeastern region of the Democratic Republic of Congo (DRC). It is one of the world's largest known copper and cobalt resources, and the largest copper mine in the DRC. The mine opened in 2009, so you will take a look at how the landscape has changed since then.
9. Make sure the Multispectral Landsat image service is selected in the Contents pane.
10. On the Data tab, in the Image Service Layer tab set, under the Selection group, click to open the Explore Raster Items pane. You can use this tool to explore all the available Landsat images in the image service.
    1. On the Select tab, set the By Area of Interest option to Display Extent to view all items within the current map extent.
    2. Check the Exclude Overviews box so you only look for source images.
    3. Set the By Attribute setting to Field, and set the field to AcquisitionDate.
    4. Leave the start date as it is—it will default to 5/11/2002, because that is the earliest date available for source imagery in this region. For the end date, click to open the calendar button and choose today as the end date.
       
    5. Click Apply.

       A list of results appears in the pane. Make sure the pane is wide enough for you to view the information fully. Since you are looking for data to see change over time, the most important parameters to note are the Acquisition Date and Cloud Cover values. The images listed are all available in the Multispectral Landsat image service, and they are listed in ascending order from oldest to newest.

    6. Scroll down until you see the image acquired on 4/30/2004 (it should be the second image in the list). The OBJECTID is 535417, and it was collected by Landsat 7 ETM+. Click the Add to Current Map button .

    In 2004, the Tenke Fungurume Mine had not yet begun production. The land is almost untouched, and there is no sign of a mining project.

11. In the Contents pane, rename the 2004 image to Fungurume 2004.
12. Now click the Sort in Descending Order button and scroll down until you see the image with OBJECTID 821460. This is an image from May 2018, a much more recent representation of the Tenke Fungurume Mine and during a time of year that is similar to your 2004 image.
13. Add this image to your map and turn the image on and off to compare with the 2004 image. Click anywhere in the image to see the pop-up showing metadata from the image.
14. Rename the May 2018 image to Fungurume 2018.

Perform additional processing on the image service

Not only can you view the different processing templates for an image service, but you can also create your own processing templates or even use image services as inputs to tools for more complex analysis. Take a look at the difference in vegetation cover from 2004 to 2018 in the area of the Tenke Fungurume Mine. To do this, you will use the Normalized Difference Vegetation Index (NDVI), which indicates the presence and health of green vegetation.

1. Change the Processing Template for both layers to None.

   This is important because a processing template can limit the number of bands available to you in a raster function. With no processing template set, all the bands from the sensor are available to be used in processing.

2. On the Imagery tab, click the Raster Functions pane. Click the NDVI function from the Analysis group.
3. Enter Fungurume 2018 as the input raster. Under Visible Band ID, choose 4 for the red band. For Infrared Band ID, choose 5 for the near-infrared band. Check the Scientific Output box. Click Create New Layer.
4. Repeat the NDVI processing for the Fungurume 2004 layer.

You now have two raster layers showing NDVI on a scale of -1 to 1 in the years 2004 and 2018. Take a look at the difference between the two by using the Minus function.
5. In the Raster Functions pane, search for and click the Minus function.
6. Choose NDVI_Fungurume 2018 as the first raster and choose NDVI_Fungurume 2004 for Raster2. Click Create new layer.

   The result is a single-band raster representing the change in NDVI from 2004 to 2018 in the Tenke Fungurume Mine region. Dark areas represent places where NDVI has decreased, indicating a loss in healthy green vegetation.

   Bright white areas should represent areas where NDVI has increased. However, by referencing your 2004 image, you can see that many of these areas are actually the locations of clouds when the image was captured in 2004.

   

In this tutorial, you added image services directly from the ArcGIS Living Atlas of the World portal, explored the built-in functionality associated with the image services, and performed additional processing on the image services using raster functions. For more information on these topics, see the resources below:

• ArcGIS Living Atlas of the World
• Key concepts for image services
• Raster functions
• Get started with Landsat imagery