Sentinel Pipeline Reference

The data from the Sentinel Pipeline requires significantly more processing than most other scrapers. As a result, this Pipeline expects a more sophisticated set of parameters when configured for a run. Most of the parameters, such as long_left or lat_down (two out of four parameters that define the geographical bounding box) are relatively straightforward. But dataset_evalscripts is more complex, identifying the specific types of data evaluation to use in a run of the Sentinel Pipeline.

Dataset Evalscripts Format

The dataset_evalscripts parameter uses JSON format to encode all the information necessary for the Pipeline to correctly evaluate its data. Sets of square brackets [] are used to denote arrays, or simple lists of things (numbers, text, etc.). Curly brackets {} are used to capture more complicated collections of object data that require text labels (e.g. {"first_name": "Gaius", "middle_name": "Julius", "last_name": "Caesar"}). JSON elements may be nested within one another, so that it is possible to have arrays of objects, arrays as the value (but not the lable/key) of an object element, etc.

The following is an example of a value to provide the sentinel5p Pipeline, which will focus on carbon monoxide concentrations on a highlighted Sentinel-2 image:

{
  "SENTINEL5P": [
    "climate-bands",
    "climate-mask"
  ],
  "SENTINEL2-L1C": [
    "true-color"
  ]
}

Currently Available Evaluation Scripts

These are all of the evaluation scripts currently available, with descriptions to help you configure your Sentinel Pipeline according to your needs:

Satellite	Script Name	Script Description
SENTINEL2-L1C	true-color	A Sentinel-2 image highlighting areas of interest based on water, vegetation, and spectral thresholds in true color. Bands: B04, B03, B02, B08, B11, B12
SENTINEL5P	climate-bands	Carbon monoxide (CO) concentrations using a color ramp from low (blue) to high (red) and processes the image into a grid to determine dominant CO concentrations per grid cell.
SENTINEL5P	climate-mask	A mask of the carbon monoxide (CO) concentrations in the image. The mask is created by thresholding the CO concentrations in the image.
SENTINEL5P	fire-bands	Sentinel-2 image focussed on detection of wildfires, highlighting areas of interest based on vegetation (NDVI), water content (NDWI), and spectral thresholds in enhanced true color
SENTINEL5P	fire-mask	A mask of the wildfire areas in the image. The mask is created by thresholding the NDVI and NDWI values in the image.