Powered by the impressive Google Earth Engine (GEE) platform, C-SAR Eye (link to the tool) tries to make Sentinel-1 satellites data analysis easier by focusing on changes in the captured C-SAR radar data. The tool allows users to customise a time frame for the data to be analysed and setup thresholds so to isolate changes in it, making it easier to identify land changes, flooding, agricultural developments, new buildings, naval activity and even (some) military radar operations (more on this and about Sentinel-1 in general at this link).

A decent amount of effort has been put into making the tool as simple and intuitive as possible so most of the functionalities should “speak for themselves” but a few fundamental tips and bits of info are still needed:

  • After you change the time frame or any of the sliders, you will need to press the UPDATE button to actually visualise the changes
  • The black and white background is a cloudless composite of Landsat satellite images, the actual Sentinel-1 data is instead represented in colours.
  • By hovering the mouse over the “Layers” button in the top right corner of the screen you can change the opacity of the resulting image, useful so to have a precise idea of where and what you are looking at thanks to google maps or sat pictures.
  • VH (Vertical Horizontal) polarisation data is represented in RED
  • VV (Vertical Vertical) polarisation data is shown in GREEN
  • The interaction between the data from both polarisations can result in other other colors as yellow.

The core mechanic of the tool is that of analysing huge amounts of data (thanks to GEE), all satellite passes in the specified time frame are considered and among those, mean, maximum and minimum values are computed for each data point / pixel. The differences between this values is what the tool uses to isolate what to visualise.

For both polarisations you will notice a series of sliders :

  • Threshold (Max-Mean) : for every point of data, to the maximum value recorded the mean one (mean value, for the pixel, among all those in the selected time frame) is subtracted. If the result is above the chosen threshold, the data will be shown in the map else it will be excluded from the visualisation.
  • Threshold (Max-Min) : for every point of data, to the maximum value recorded the minimum one (lowest value, for the pixel, among all those present in the time frame) is subtracted. If the result is above the chosen threshold, the data will be shown in the map else it will be excluded from the visualisation.
  • Threshold (Mean-Min) : for every point of data, to the mean value (of all the time frame) the minimum one is subtracted. If the difference is above the chosen threshold, the data will be shown in the map else it will be excluded from the visualisation.
  • Amplifiers : for data isolated by both thresholds, amplifiers will basically amplify the resulting values so to make it more visible. Useful to bring up faint details or tone down the more intense ones.
  • Remember to press the “Update” button after having made changes to the values and allow it some time to process the data.
Area Analysis of a simple polygon

“Area Analysis” tools allow to selected either a polygon or a point on the map and have the minimum and maximum values charted across the specified time frame. Particularly useful to isolate when changes happened (example: new building constructed) or radar activity patterns. Resulting charts can then be exported in various formats for further analysis and comparisons. (Note: charts might fail to load, haven’t yet isolated the bug, but repeating the area selection usually fixes it…)

Example of Focal Blur used over the Chinese missile silo field ear the northwestern city of Yumen

“Focal Blur” if the checkbox is enabled a focal blur passed (with a circular kernel) is performed on resulting data, the slider allows to select the radius of the kernel. This is particularly useful to isolate small details making them “pop” even with low zoom levels.