Sprint 2 - Explore data sources to compute required metrics

[baogianghoangvu]

We need to find data sources that can programmatically provide necessary information (data type) to compute the metrics as described in #7.

Per our discussion so far:

• Data sources:
  • Google Earth Engine
  • Earth explorer - USGS
• Data types:
  • Multispectral image (or hyperspectral) are tentatively the data type of choice.
  • Otherwise, standard 3-channel images can be used instead.

Please:

• continue investigating these data sources or others you may find, keep a list of them so we have more options to fall back on, depending on how @ntmt2903 proceeds with investigating metrics,
• then implement Python solutions to extract/download and process data from these sources based on any customized requirement (e.g. may input any timeframe, location coordinates, frequency, etc.). As a first step, you may want to create functions like so:

  def extract_1day_from_usgs(date, coordinates):
      # todo: connect to USGS' public API with appropriate parameters and retrieve data
      return downloaded_file

• ideally, the solutions should be able to process a whole timeframe (start_date to end_date) with custom frequency (e.g. daily, weekly, etc.)
• Document your approach:
  • which library you use,
  • the steps/instructions such as authentication to extract data
• Last resort is static data (i.e. manually downloaded data)

[baogianghoangvu]

Please create a new branch named datasource to differentiate from the existing _dev branches.