There are two sample Geocoded Covariance (GCOV) products available, one for each of the acquisitions used in other sample products that require imagery pairs:

NISAR_L2_PR_GCOV_001_030_A_019_2000_SHNA_A_20081012T060910_20081012T060926_D00402_N_F_J_001.h5

NISAR_L2_PR_GCOV_002_030_A_019_2000_SHNA_A_20081127T060959_20081127T061015_D00402_N_F_J_001.h5

The covariance variables are essentially equivalent to Normalized Radar Backscatter (NRB) or Radiometrically Terrain Corrected (RTC) products. They include additional information that make them suitable for additional analysis workflows beyond what a standard RTC product can be used for, but can also simply be used in place of an RTC.

The illustrations displayed in this section use the acquisition from October 12, 2008. Note that the data used to create these sample products are from JAXA’s PALSAR sensor, which also operates in the L band, but the product has a smaller footprint than the NISAR acquisitions.

Add GCOV Data

You can use any of the standard methods to add GCOV products to your ArcGIS Pro map. If you drag and drop, the HH-HH covariance variable will be displayed by default.

When using the Add Multidimensional Data dialog, any of the variables in the grids path of the file can be added to the map, but most users will be interested in viewing the Covariance layers.

Note that the sample data only includes one polarization (HH). Most NISAR data will include at least two polarizations (over land, most acquisitions will include at least HH and HV), and each will be included as a separate variable in the h5 file.

Visualize GCOV Products

You can adjust the symbology of the NISAR sample data layers just like you would any other raster layer. Right-click the layer in the Contents pane, and select Symbology. Adjust the stretch as desired. If you are accustomed to looking at RTC products in grayscale, set the color scheme to Black to White, and make sure that you uncheck the box to invert the color direction.

Common approaches for backscatter stretch settings are to use a Minimum-Maximum stretch with fixed values applied (i.e. 0 to 0.25), or to use a Standard Deviation stretch, choosing an n value that works well for your particular image.

If you’ve used the Multivariate approach to add multiple variables with different value ranges to a single Multivariate layer, recall that the same symbology will be automatically applied to each of the layers when you select a different variable. In this case, you may want to use a Standard Deviation stretch rather than a Min-Max, as it will accommodate a wider range of values. Consider also using the DRA (Dynamic Range Adjustment) setting for statistics rather than the dataset statistics.

The following series of screen shots displays all the different GCOV sample data grid variables using the same stretch setting.

Multiband Raster Symbology

While this is not a particularly compelling approach for the current sample GCOV data, which only includes one polarization for covariance, it is possible to add multiple covariance variables from a single HDF5 file to your project as a Multiband raster. Once data files are available that include multiple polarizations, you will be able to use this feature to generate RGB images by assigning different polarizations to different bands (i.e. set the Green band to use the HV variable and the Red and Blue bands to use the HH variable).

To use this functionality, set the Output Configuration to Multiband Raster when selecting the variables to add to the map. This adds a single layer to the map, with the individual variables available as bands in the Symbology interface.

In the example below, the HH backscatter is assigned to the red channel, the number of looks to the green channel, and the rtc gamma to sigma factor to the blue channel. It doesn’t make for a very compelling view, but if you had multiple covariance variables with different polarizations available in the HDF5 file, this approach would be a quick way to display a simple RGB decomposition of the available polarizations.

The sample Geocoded Unwrapped Interferogram (GUNW) data file is:

NISAR_L2_PR_GUNW_001_030_A_019_002_2000_SH_20081012T060912_20081012T060924_20081127T061001_20081127T061013_D00402_M_P_J_001.h5

While the product name calls out that the interferograms have been unwrapped, both the wrapped and unwrapped phase difference products are included as variables.

Most of the variables in this file have float or integer pixel types, but the wrapped phase variable has a complex pixel type. ArcGIS can display all of the variables, but refer to the Wrapped Phase section for details on how to appropriately visualize complex datasets. This table describes the pixel types for each variable included in the grids group of the h5 file:

To learn more about InSAR products in general, including the difference between wrapped and unwrapped phase and the meaning of coherence, refer to ASF’s Exploring Sentinel-1 InSAR tutorial.

Add GUNW Data – Float Format

Any of the variables that are not 64-bit complex can be added to the ArcGIS Pro project using the standard methods. The wrapped interferogram (and any other complex variable) must be added as a Multiband Raster in order to make use of the tools available for extracting phase measurements from the dataset. Refer to the Add Complex Data section for more information.

If you simply drag and drop the GUNW HDF5 file from the Catalog pane into a Map in ArcGIS Pro, it will display the alongTrackOffset variable by default.

The GUNW products include quite a few gridded variables, as listed in the table above. There are also many metadata variables from the /science/LSAR/GUNW/metadata/radarGrid/ path that can be visualized in the map, as well.

The most common variables to visualize are the unwrapped phase, the wrapped interferogram, and the coherence magnitude (note that coherence variables are available based on both the wrapped and unwrapped interferograms).

Note that there are limitations for working with the wrapped interferogram if it is added as a multidimensional raster layer. It will display the Amplitude values by default, but you will not be able to use the Complex raster function to access the Phase values. Refer to the Add Complex Data and Wrapped Phase sections for more information.

Add Complex Data

For variables with 64-bit complex pixel values, such as the wrapped interferogram, you must add the data as a Multiband Raster layer in order to allow for the correct extraction of the phase measurements. The Complex raster function, described in the Wrapped Phase section does not function properly if the variable was added as a Multidimensional Raster layer.

Use the same Add Multidimensional Data workflow to navigate to the HDF5 file, but after selecting the complex variable, select Multiband Raster from the dropdown menu for the Output Configuration.

By default, the pixel values displayed when the image is added will indicate Amplitude. If you add the variable using the Multidimensional Raster output configuration, Pro will also display the Amplitude values by default, but you will not be able to extract the Phase values.

Note in the image below that the pixel values are the same in the default visualization of the wrapped interferogram, regardless of whether it was added as a Multiband Raster or Multidimensional Raster layer.

By default, Multiband Raster layers are displayed in grayscale and Multidimensional Raster layers use a color scheme.

Visualize GUNW Products

You can adjust the symbology of most NISAR GUNW variables as you would any other raster, with the majority of these adjustments involving the stretch characteristics. There are some additional considerations to keep in mind when visualizing the wrapped phase, however, so we’ll start with that.

Wrapped Phase

The wrapped phase variable has a complex pixel type, containing both real and imaginary components that can be used to calculate either the amplitude or the phase values for the dataset. When looking at an interferogram to detect and quantify surface deformation, we’re interested in differences in phase rather than in amplitude.

ArcGIS Pro provides a ‘Complex’ raster function, which can be used to display the phase values instead of amplitude values. It is included in the SAR section of the raster functions panel in ArcGIS Pro. This raster function requires that the variable be added as a Multiband Raster rather than a Multidimensional Raster layer.

1. Add the wrapped interferogram variable to the map, ensuring that you set the Output Configuration to Multiband Raster, as described in the Add Complex Data section.
2. From the Imagery menu, select Raster Functions to open the Raster Functions panel
3. Type Complex in the search bar at the top of the Raster Functional panel, and click on the Complex raster function
4. Select the wrapped interferogram layer in both the Raster and the Imaginary Raster fields
5. Select Phase from the Value Type drop-down menu
6. Click the Create New Layer button to display a layer showing the phase difference values of the wrapped interferogram (or use the arrow next to the Create New Layer text to select the option to save the output as a stand-alone raster).

The values displayed in the image now range from -pi to pi, which is what we would expect for phase measurements in a wrapped interferogram.

Note that the phase layer is a virtual raster that is displayed on-the-fly in your project, but you can export the layer as a stand-alone raster if you would like to save it for future use in other applications.

It may be helpful to adjust the symbology so that the stretch displays the full range of values (Minimum-Maximum). Applying a color scheme where 0 values have a neutral color may also help to indicate the sign and magnitude of the phase differences between the two acquisitions.

Unwrapped Phase

Phase unwrapping converts the wrapped phase values, which are constrained to a range of -3.14 to 3.14 radians, to a continuous range of consecutive multiples of 2π radians. The output values are no longer complex values, but are 32-bit float values with units of radians. You can treat the unwrapped phase variable like any other multidimensional raster. Note that the pop-up indicates that the pixel values are in radians.

Adjust the symbology as desired to highlight the information you want to see. To get a sense of the overall direction of phase difference, try setting custom Minimum-Maximum values that are equal in magnitude in the positive and negative directions and apply a color scheme with a neutral color in the middle.

Coherence Products

There are two different coherence variables included in the GUNW product: wrapped and unwrapped coherence. Both variables have pixel values that range between 0 and 1, with values closer to 0 indicating low coherence and values closer to 1 indicating high coherence.

The key difference between these two variables is that there is multilooking applied to the unwrapped coherence product. While this decreases the resolution of the data, it can make for a more useful visualization of the overall coherence patterns.

In the comparison below, the wrapped and unwrapped coherence products are displayed both zoomed in to the level where individual pixels are visible (top images) and zoomed out to nearly the full extent of the image.

While the wrapped coherence image may provide a more detailed representation, especially when looking at very small individual areas, the unwrapped coherence gives a stronger indication of areas with very low coherence (such as lakes and agricultural fields) and very high coherence (urban areas and regions with little vegetation).

The sample Geocoded Offset (GOFF) data file is:

NISAR_L2_PR_GOFF_001_030_A_019_002_2000_SH_20081012T060912_20081012T060924_20081127T061001_20081127T061013_D00402_M_P_J_001.h5

The approach used in the GOFF products is often called Pixel Tracking.

There are 21 variables included in the grid group, seven variables for each of three different layers, generated with different window sizes. They are all 32-bit float datasets. Variables from this product can be added using any of the standard methods, and visualized as you would any other raster.

If you drag and drop the GOFF HDF5 product from the Catalog pane to the map, it will display the /science/LSAR/GOFF/grids/frequencyA/pixelOffsets/HH/layer1/alongTrackOffset variable.

There is one sample Soil Moisture (SME2) product:

NISAR_L3_PR_SME2_001_008_D_070_4000_QPNA_A_20190829T180759_20190829T180809_P01101_M_P_J_001.h5

Note that the product linked on the JPL site does not appear to be the most recent iteration of the test data. The key issues are:

• The variable descriptions are blank
• NoData pixels are assigned a value of -9999, which complicates visualization of the data

You can still use any of the standard methods to add the variables to your project, however.

If you use the drag-and-drop method to add the file, the /Algorithm/DSG/Algorithm_Param_Beta variable will be added by default, which is not generally of interest to most users.

There are three different variables containing soil moisture data: DSG (Disaggregated), TSR (Time Series Rates), and PMI (Physical Model Inversion). Use any of the standard methods to add the desired variables to your map.

Visualizing SME2 Dta

The sample product currently available from the JPL website uses the value -9999 to indicate NoData.

This requires some adjustments to the symbology settings before you can display the data with a meaningful stretch applied.

To adjust the visualization to exclude the -9999 values from the grayscale and treat them as background pixels:

1. Right-click on the soil moisture layer, and select Symbology
2. Click the Mask tab, check the box to Display background value
3. By default, the background is transparent
   • If you want to see the extent of the padding pixels, you can select a color (the illustration used pink)
   • Once a color has been selected, click on the color swatch again and select ‘Color Properties’, where you can apply a transparency value if desired (the illustration used 90%)

4. Change the Stretch type to Minimum-Maximum
5. Select the Statistics tab
6. Use the drop-down menu to set the Statistics field to Custom
7. Change the value in the Minimum field from -9999 to 0 and press the tab button to update the scalebar label

There are two sample Geocoded Single-Look Complex (GSLC) products available, one for each of the acquisitions used in other sample products that require imagery pairs:

NISAR_L2_PR_GSLC_001_030_A_019_2000_SHNA_A_20081012T060910_20081012T060926_D00402_F_N_J_001.h5

NISAR_L2_PR_GSLC_002_030_A_019_2000_SHNA_A_20081127T060959_20081127T061015_D00402_F_N_J_001.h5

The illustrations in this section use the acquisition from October 12, 2008. Note that the data used to create these sample products are from JAXA’s PALSAR sensor, which also operates in the L band, but the product has a smaller footprint than the NISAR acquisitions.

Add GSLC Data

As with the wrapped phase variable in the GUNW products, the SLC variables in the GSLC are in complex data format. While you can add these variables using the Drag and Drop or Add Multidimensional Raster approaches, these methods will display the Amplitude measurements by default, and you will not have access to the phase measurements.

To extract the phase values, you will need to add the SLC variables as Multiband Rasters, so we’ll start with that method, then go over the rest of the options.

Add Multiband Rasters

In the Map tab, click the down arrow next to the Add Data button and select Add Multidimensional Data. Navigate to the GSLC HDF5 file, and select it to open the list of variables. Select the focused SLC variable. Click the drop-down menu under Output Configuration and select Multiband Raster, then click OK to add the variable.

Note that the sample data only includes one polarization for the SLC products, but actual NISAR data will often have multiple polarizations available.

By default, ArcGIS Pro displays the Amplitude values when a complex raster is added to the map.

Drag and Drop

If you drag and drop the GSLC HDF5 file from the Catalog pane into a Map in ArcGIS Pro, the first variable in the HDF5 file is displayed. The sample GSLC files only have two gridded frequency variables, and the first is the focused SLC image in HH polarization. The values in the SLC variables are in 64-bit complex format, so if you do use the drag-and-drop method, you will only be able to access the Amplitude measurements (which are displayed by default).

Note that the software will try to calculate the statistics for the dataset, and it will keep trying for a very long time. You can cancel this effort to view the image with estimated statistics.

Because the statistics were not calculated, the range of values in the legend will likely be different from the layer added as a Multiband raster. If you click on a pixel, however, both layers return the same pixel value.

Add Multidimensional Data – Individual Variables

You can also add variables from the GSLC files as multidimensional rasters. Again, you will only be able to access the amplitude measurements in the SLC variables, but there are other variables that are better suited to this approach.

The other gridded frequency variable is the validity mask, but the variables in the /metadata/radarGrid path can also be visualized in ArcGIS Pro. These variables include along-track unit vectors, elevation angles, ground track velocity, incidence angles, LOS unit vectors, slant range, and Zero Doppler Azimuth time.

To add individual variables, click the drop-down arrow on the right side of the Add Data button, and select Add Multidimensional Data. Navigate to the GSLC HDF5 file and select it to view the available variables.

Click one or more of the variables listed, and leave the Output Configuration set to the default of Multidimensional Raster. Click the OK button to add the variable(s) to the map. If you select multiple variables, they will each be added as individual layers.

Add Multidimensional Data – Multivariate

It is also possible to add multiple variables to the project as a single layer. Use the same approach as when adding single variables to browse to the HDF5 file and view the list of variables to add. Select the desired variables, but instead of leaving the Output Configuration set to the default of Multidimensional Raster, use the drop-down Output Configuration menu to select Multidimensional Multivariate Raster.

This adds a single layer to the map, labeled with the filename only (no variables indicated). The different variables that were selected can be accessed using the Multidimensional menu available when you click on the layer in the Contents pane.

Keep in mind that the same constraints apply to visualizing the complex-valued SLC variables; you will only be able to view the amplitude values.

Visualizing GSLC Products

The focused SLC variables in the GSLC products have complex pixel values, which include both real and imaginary components. ArcGIS Pro offers a Raster Function that allows you to extract either the Amplitude or the Phase measurements from the complex numbers, but the variables must be added using the Multiband Raster output configuration in order to use this functionality. You can then adjust the visualization of the output raster as desired.

Any of the other variables in the GSLC product (the validity mask or any of the variables in the radarGrid path) can be visualized as you would any other raster.

Apply Complex Raster Function

Click on the Raster Functions button under the Imagery menu to open the Raster Functions window. Either type Complex into the search bar, or expand the SAR menu and click the Complex button.

This opens the dialog window for the Complex raster function. Select the same layer (focused SLC variable) for both the real and imaginary inputs. The layer must have been added as a Multiband Raster.

For the value type, select either Amplitude or Phase, then click the button to Create new layer. This will add a virtual layer to your map project. Alternatively, select Save As from the drop-down menu instead of Create new layer if you want to save the amplitude or phase output a new file for use in other projects.

If you select Amplitude, the output product will have the same values as the default visualization of the product. Again, the range of values in the legend may be different due to the calculation/estimation of statistics, but if you click on pixels, the values will be the same.

If you select Phase, the output product will have values ranging from -3.14 to 3.14. Generally the phase is not very informative visually, with what looks like a random pattern.

Note that the output layer is labeled with a Complex prefix in the Contents pane, regardless of whether you selected Amplitude or Phase. You may want to edit the layer name if you want to add both to your map.

When you click on a pixel, however, the value displayed in the pop-up window includes an indication of whether you’re looking at amplitude or phase, even though the layer names are the same.

Symbology for Amplitude Measurements

You can use the same visualization approaches as with the GCOV products when viewing the samplitude component of the dataset. Right-click the layer in the Contents pane, and select Symbology. You can adjust the color scheme, the stretch type, and the range of values used for the stretch.

Because the amplitude values aren’t calibrated backscatter, the value ranges that work well may be different from what you would use with the GCOV products. For example, here is the GSLC amplitude image with a Minimum-Maximum value range set to 0 and 0.25. The visualization has much different contrast characteristics than what those same settings rendered for the GCOV product.

Symbology for Phase Measurements

There’s generally not much to explore visually with GSLC phase measurements. If you click the Histogram icon next to the Stretch type in the Symbology pane, you can see that the pixel values are fairly evenly distributed across all values. As such, it doesn’t make much difference visually if you apply a Minimum-Maximum, Percent Clip, Standard Deviation, or Histogram Equalize stretch type.