Support for GFS and other weather models about rocketpy

[giovaniceotto]

A possible solution to this issue would be to add a translator dictionary to the Environment class which would convert standard variable names to those used in specific forecasts or reanalysis, such as GFS and ERA 5.

This dictionary could be an input parameter to the __init__ method of the Environment class.

[giovaniceotto]

To solve this issue in a more general form, the Environment class has suffered some big changes.

Now, when initializing the class , no atmospheric data is given, only rail length, possibly gravity, and spacetime data:

Env = Environment(railLength=5,
                                latitude=10,
                                longitude=10,
                                elevation=100,
                                date=(2019, 2, 6, 12))

Then, the atmospheric model can be set by using:

Env.setAtmosphericModel(type='Forecast', file='GFS')

As the example shows, not only are other forecast models supported by they can be easily specified my a short name.

The Environment.setAtmosphericModel method supports several atmospheric model types:

• International Standard Atmosphere
  • As defined by ISO 2533, with no wind.
• Custom Atmosphere
  • Which allows the user to input custom pressure, temperature, and wind profiles. In case pressure and/or temperature are not given, the International Standard Atmosphere is used.
• Wyoming Upper Air Soundings
  • Sounding data given at http://weather.uwyo.edu/upperair/sounding.html
• NOAA Ruc Soundings
  • Sounding data given at https://rucsoundings.noaa.gov/
• Forecasts
  • Forecast files in netCDF or OPeNDAP formats. Default ones such as GFS, FV3, RAP and NAM have shortcuts.
• Reanalysis
  • Reanalysis files in netCDF or OPeNDAP formats.
• Ensembles
  • Ensemble forecasts and reanalysis in netCDF or OPeNDAP formats. Default ones such as GEFS and CMC ensembles also have shortcuts.

The use of dictionaries needed for Forecasts, Reanalysis and Ensembles is well described in the method documentation and two dictionaries are available through shortcuts: ECMWF and NOAA.

Overall, this has greatly increased the capabilities of the Environment class relative to atmospheric models. It feels pretty complete by now, even though some few and small bugs might pop up in the future and need to be fixed.