Create a MongoDB to store information about runs/datasets

[carlosgjs]

Need to define:

• What documents will be stored here and their schema
• When/how are new documents inserted, by the noisepy cli? from notebooks?
• Who has permissions to read/write this DB?
• Is there support for multiple databases? If so , how is this configured?

[mdenolle]

Update: we keep the cross correlation themselves into S3 as a supported noisepy data stack store. But we should write to a documentDB the data products:

• Output dv/v, d-attenuation time series into a mongoDB document: we must create a data schema. this should also be output as a CSV file.
• Ouput dispersion curves from the imaging package should also be output 1) raster image for the dispersion image, or 2) tabular data in CSV/JSON in mongoDB.