Evaluate Sentinelsat and Sentinelhub libraries to access satellite data

[tnigon]

Sentinelsat docs page

Use cases

1. For each ground truth point, retrieve the representative reflectance for each sentinel band at that location for the date closest in time (also save the days from image date).
2. For each CSS field, retrieve the images as they become available (in the future), crop image by field bounds, and use Training.predict() to make a prediction on each image pixel.

Evaluation criteria

1. Ease of retrieving data for each of the use cases above
2. Cost
3. Reliability
4. Auxiliary information that may be helpful in current or future predictions

To do

1. Create an account with Copernicus
2. Install sentinelsat conda install -c bilelomrani sentinelsat
3. Show a proof of concept that we can get latest satellite imagery for any location on demand

[tnigon]

Closing issue because was addressed in spatial repository (now combined with db).