0.12

Changes in 0.12

• Updated list of kerchunk datasets
• Added xarray Data Trees as fourth data model. Data Trees provide a tree-like
  hierarchical structure of datasets, where single nodes can be accessed by an
  identifier. This is a preferable approach to splitting datasets into sub-datasets,
  as. e.g., "esacci.FIRE.mon.L3S.BA.MSI-(Sentinel-2).Sentinel-2A.MSI.2-0.pixel~h41v14-fv2.0-JD"
  The solution has been made available for FIRE, LC, and VEGETATION datasets.
  A full list of available datasets can be retrieved with the store method
  get_data_ids(data_type="datatree").
  To open a data tree with a subset of datasets, pass place_names as parameter
  to the store method open_params(). You can get a list of available place names
  from get_open_data_params_schema().
• In concordance with the change above, the support for the dataset
  "esacci.RD.satellite-orbit-frequency.L3S.WL.multi-sensor.multi-platform.MERGED.v1-1.r1"
  has been modified so it is no longer split up into single datasets as, e.g.
  "esacci.RD.satellite-orbit-frequency.L3S.WL.multi-sensor.multi-platform.MERGED.v1-1.r1~AMAZON_NEGRO_SAO-FELIPE".
  To open the dataset with a subset of places, pass place_names as parameter
  to the store method open_params(). You can get a list of available place names
  from get_open_data_params_schema().
• We fixed an issue with geodataframes which could cause that not all data were read in.
• Fixed issues that data could not be accessed from ICESHEETS datasets with multiple time steps in base files.
  Solves #69