Creating Single-Band GeoTIFFs of Sidescan Data

Chesapeake Times, Vol 7 | October 2021

What is a Vertical GeoTIFF??

GeoTIFF is a public domain metadata standard which allows geo-referencing information to be embedded within a TIFF file (Wikipedia). Normally, the GeoTIFF format is used for geo-referenced images, where each pixel in the TIFF represents a color to display at a specified geographic location. The metadata in the TIFF encodes the geospatial information needed to render the image at the appropriate location on the earth. SonarWiz offers extensive support for exporting seafloor imagery in this format. 

More recently, GeoTIFFs have been used to encode physical measurements rather than just imagery. For example, you can use a GeoTIFF to encode elevation data or seafloor amplitude information. In this format, the GeoTIFFs are serving as a raster or grid. The advantage of the GeoTIFF-as-raster format is that it is non-proprietary and the georeferencing information is well defined and embedded in the file itself. In SonarWiz, we refer to these types of GeoTIFFs as “Vertical” GeoTIFFs, to distinguish them from conventional image GeoTIFFs.

The current version of SonarWiz can convert any Grid into a “Vertical” GeoTIFF, that is, if you have a grid of bathymetry, you can convert it into a sing-band GeoTiff encoding the elevation data, or likewise, if you have a grid of sidescan amplitude data, you can convert that into a single-band GeoTIFF encoding the backscatter values.

Creating a single-band GeoTiff of sidescan data

An interesting example is to convert a sidescan mosaic into a “Vertical” GeoTIFF. This will create a single-band TIFF file where the pixel values encode the actual backscatter values instead of the mosaic colors as they are displayed in SonarWiz. At the end of the example, we will show the use of QGIS to offer more control over the bit-depth of the output TIFF.

Figure 1: Sidescan mosaic of a shipwreck

The dataset consists of two sidescan files that have been fully processed in SonarWiz (Figure 1). To create a conventional GeoTiff of this mosaic, we use the Save Project As GeoTIFF utility in the Post Processing Ribbon. This utility gives us a number of options for GeoTIFF output including setting the bit depth from 8-bit to 32-bit (Figure 2). No matter which option you choose, the GeoTIFF will always be a picture of the sonar data, not the data itself. That is, each pixel will contain the color of the data exactly as it was displayed in the main window. This is NOT a vertical GeoTIFF.

Figure 2: The Save Project As GeoImage command stores a picture of the data in the GeoTiff, not the data itself.

To create a Vertical GeoTIFF we need to first create a grid of the sidescan data. Right click on the Grids branch in the Project Explorer and select the Create a New Grid command. This opens the Gridding Setup dialog (Figure 3). Set the data source to Internal, specify an appropriate cell size for your data and pick CSF Sidescan as your Grid Surface Type. There are several different gridding algorithms available.

Figure 3: The Gridding Setup dialog used for gridding sidescan data

Once the grid is created, it will appear in the Grids Branch of the project Explorer (Figure 4). If you select the grid you can see the grid limits in the Properties window. In our case, our sidescan amplitude is the Z value of the grid and ranges from a minimum value of 0.693 to a maximum value of 3473.089.

Figure 4: Grid of sidescan data appears in the Project Explorer and Properties window

The final step is to export the grid as a Vertical GeoTIFF. Right-click on the grid and select the Export Grid As… command. There are many export formats available, we are interested in the Vertical GeoTIFF option as shown in Figure 5.

Figure 5: Export the sidescan grid as a Vertical GeoTiff

Comparing a regular GeoTIFF to Vertical GeoTIFF

You can compare the difference between a regular (image) GeoTIFF and a Vertical GeoTIFF in QGIS. When we load the tiff exported by the Save Project as GeoImage command in SonarWiz, the tiff comes in as an 8-bit, 24-bit or 32-bit color image. In this example, the TIFF was exported as a 32-bit TIFF where band 1 represents the red color, band 2 represents the green color, band 3 represents the blue color and band 4 represents the alpha channel (transparency) (Figure 4).

Notice that the image looks good out of the box using the exact color scheme we had set in SonarWiz, including all of the map decorations such as a scale bar and color legend. This is because the GeoTIFF exported by the Save Project As GeoImage command exports a picture of our mosaic, not the raw sidescan data itself.

Figure 6: Regular GeoTiff exported with Save Project as GeoImage command exports a 3-band tiff with band 1 representing Red, band 2 representing green and band 3 representing blue. Band 4, not shown has the alpha (transparency) channel

In contrast to this, if we open the Vertical GeoTIFF we exported from the Grid, the TIFF loads as a single band raster, where band 1 ranges from 0.693 to 3473.78. These are the exact amplitude values of our sidescan grid created above. In other words, the amplitude data was loaded, not a picture of the data. The image is washed out because it has not been properly rendered (yet). You will have to color the data using your GIS color settings settings, not SonarWiz. This is easily done as shown in Fig. 8.

Figure 7: Vertical GeoTiff of exported sidescan data created by exporting a grid to a Vertical GeoTiff

Figure 8: QGIS rendering of single-band GeoTiff above after modifying the color map.

Controlling the Bit-Depth of Exported Vertical GeoTIFFs
When you export a grid as a vertical GeoTIFF, SonarWiz automatically selects the bit depth to contain your data without truncation. Usually, SonarWiz selects a 16-bit storage because most sidescan data are 12-bit or 24-bit these days. If you instead need an 8-bit image, you can easily convert the data in QGIS using the Translate command found under the Raster > Conversion menu.

Figure 9: QGIS Translate command can be used to convert a 16-bit TIFF to an 8-bit TIFF

– David Finlayson, Chief Scientist