Reclassify RTC Tool

 This tutorial is one part of a larger workflow using On Demand RTC products from the Alaska Satellite Facility. 

Calm surface water is generally characterized by very low radiometric returns in Synthetic Aperture Radar (SAR) imagery. This tendency can be used to identify surface water in a Radiometric Terrain Corrected (RTC) SAR product.

The RTC products used in this demonstration were generated using ASF's  On Demand Processing  functionality at  Data Search - Vertex . Click the button below to search for the same two source granules used in the demonstration.

In Michigan, heavy rainfall in May 2020 resulted in flooding in Shiawassee National Wildlife Refuge. Because of the greater differentiation among dark pixels that is visible in dB scale, we will use RTC images in dB scale to determine surface water extent after flooding.

If you did not order your On Demand RTC products in dB scale, refer to the  tutorial for the Scale Conversion Tool  to learn how to convert RTC products from amplitude or power to dB scale.

There is not one constant threshold value that can be applied to any dB image to identify surface water. The value will depend on the local site conditions at the time the image was required. To determine the appropriate value to apply, start by simply applying a classified symbology to the dB image.

In this example, we will use the co-polarized (VV in this case) image to generate the water mask. On a calm day, surface water will return very little VV backscatter to the sensor, because the surface of the water is smooth and water has a high dielectric constant. As a result, most of the incoming radar signal bounces off the surface and off in the opposite direction of the sensor.

On a windy day, however, the water surface can become rough enough to return some of the signal to the sensor, and it becomes much more difficult to distinguish between the water and surrounding terrain. If you must classify the water using an image collected on a windy day, you may be better off using the cross-pol (VH or HV) dataset, or a combination of the VV and VH values.

Step through the slideshow below to learn how to determine the threshold value to use for classification of surface water.

You can check your extent map against a variety of other datasets. It can be helpful to compare the classified layer against the original dB layer in grayscale stretch. In this example, the water mask is applied to all pixels with values less than -18 dB.

Another potential resource to use for comparison is an  RGB Decomposition  image, which combines the co- and cross-pol datasets into a false color image. Pixels with low backscatter in both co- and cross-pol datasets appear as dark blue in the decomposition. There is an RGB Decomposition Tool included in the ASF_Tools ArcGIS Toolbox. In this example, the water mask is applied to all pixels with values less than -19 dB.

Once you determine the threshold value that will work best for your specific application, you could just save the newly symbolized raster as a layer and use it for a visual overlay. In some cases, however, you may want a reclassified raster indicating surface water extent to use for analysis with other data layers.

The RTC Reclassify Tool allows you to generate a new mask raster that only includes the pixels that fall below the threshold value.

The RTC Reclassify Tool in the ASF_Tools ArcGIS Toolbox generates a raster mask based on a threshold value. Any pixels in the input raster with a value below the threshold are classified with a value of 1, and all other pixels are set to No Data. It was designed to be used to generate water masks, based on a dB scale threshold value, but will work for any reclassification where the values of the pixels of interest are below a threshold value.

You can  download the toolbox  from ASF's website. Extract the zip file to the directory of your choice on your computer.

Step through the slideshow below to learn how to use the tool in ArcGIS Pro. This tool can also be used in ArcMap; the process is generally the same, but the interface looks different.

When using the tool in ArcGIS Pro, hover over any of the blue information icons in the Geoprocessing dialog for more information about the parameters or the tool in general. The full tool information can be viewed by right-clicking the tool in the Catalog window and selecting View Metadata.

When using the tool in ArcMap, click the Show Help button in the tool dialog box to see the information about specific parameters or the tool in general in the Help pane. The full tool information can be viewed by right-clicking the tool in the Catalog window and selecting Item Description.

It is important to be aware that not all surfaces with very low backscatter are water. While this approach can work well in many situations, make sure to consider the context of the image before relying on the masks.

In particular, both calm surface water and dry desert sand tend to have very low backscatter, but for opposite reasons. The high dielectric constant and smooth surface of the water result in the signal reflecting off the surface, essentially bouncing off in the opposite direction from the incoming signal.

Dry sand, on the other hand, has very low dielectric properties, allowing the signal to penetrate deeply into desert landscapes. By the time the signal is scattered, it lacks the energy necessary to return back through the medium. In both cases, there are low radiometric returns, but one is because the signal was bounced away, while the other is because the signal was absorbed by the medium.