Water Observatory Backend example


#1

While exploring example "example-optical-water-level-extraction " I have some doubts:-

  1. Here one number is used in ‘extract_surface_water_area_per_frame’.What that number refer to and from where that number is taken?

  2. I also want to know Bluedot Water Observatory’s API link used for calculating nominal outline


#2

Let me explain BlueDot Water Observatory’s currently very simple RESTful API using an example water body from Spain (https://water.blue-dot-observatory.com/19853/2017-10-08). The number 19853 is water body’s unique identifier and it will appear in all requests.

  • all measured surface water levels can be obtained via: https://water.blue-dot-observatory.com/api/waterbodies/19853/index.html, which returns a JSON with the following content:
    • measurements: list of available measurements (note that more measurements have been performed, but were filtered due to too large cloud coverage, failure of algorithm, …), where each measurement consists of:

      • cc: cloud coverage in percent (threshold is set to 2%)
      • date: date of S2 acquisition
      • level: surface water level as defined above
      • min/max_level: obsolete. Set to 0.0. Will be removed in future.
    • nominal_outline: Polygon of water body’s nominal water extent (taken from OSM in most cases)

    • properties: other properties of water body

      • country, name
      • lat, long in WGS84
      • id: unique identifier
    • the measured water extent can be obtained for each available date via: https://water.blue-dot-observatory.com/api/waterbodies/19853/measurements/2016-10-08/index.html, where 2016-10-08 represents date of a measurement (from the above json file). The result of the above call is geo-json file with a Polygon and nothing else.

In summary, all you need to know is the water body’s unique identifier (can be obtained from the url in the dashboard when selecting one) or from our “database”, available at https://www.blue-dot-observatory.com/aboutwaterobservatory.

I hope this is enough details for now. Let me know, if you have any other questions.

PS: Number in question 1 represents the ware body ID. In the example code this is actually not needed since the example doesn’t make any calls to water observatory’s API.


#3

Though I have list of queries but at present I need help on this part.When I was trying to understand the further working I came across file “geom_utils.py”.Within this there is function named “apply_DEM_veto”.Can you explain logical working of this function?I am not able to understand why logical operators and threshold value is used here?I am not from remote sensing background, so not able to understand theoretical concepts.
Overall do you have any documentation of theoretical working of BlueDot Water Monitoring Observatory ?


#4

The purpose of this step is to reduce the false positive water detection for water bodies near mountainous terrain due to topographic shadow. For more in depth understanding of the problem and algorithm I suggest you to check this PhD thesis. The algorithm that is used in Water Observatory is very similar to the one described in the thesis. One difference for example is that we use DEM instead of HAND (height above nearest drainage), since we don’t have access to the latter.