What’s New#

release date PyPI

pypi dwn

Coming up next#

the set of parameters. (#29) by G. Maze.

v0.4.0 (2 Feb. 2024)#

Note

New version compatible with Parcels versions >= 3.0.0

  • In order to stay compatible with last Parcels development, we had to update VirtualFleet. This is a major internal refactoring that does not impact the user API. All details here. (#24) by G. Maze.

v0.3.1 (22 Nov. 2023)#

Note

This is the last version compatible with Parcels versions < 3.0.0

New features

  • Mission parameters can now be set for each floats of the deployment plan. This is useful to limit the number of simulations to explore a set of configuration parameters. (#22) by K. Balem.

For instance:

# Number of floats
nfloats = 5

# Define space/time locations of deployments:
lat = np.linspace(40, 41, nfloats)
lon = np.full_like(lat, 5)
tim = np.array(['2020-01-16' for i in range(nfloats)], dtype='datetime64')
my_plan = {'lat': lat, 'lon': lon, 'time': tim}

mission = [
    FloatConfiguration('default').update('parking_depth', 100),
    FloatConfiguration('default').update('parking_depth', 200),
    FloatConfiguration('default').update('parking_depth', 500),
    FloatConfiguration('default').update('parking_depth', 1000),
    FloatConfiguration('default').update('parking_depth', 1500)
]

VFleet = VirtualFleet(plan=my_plan, fieldset=VELfield, mission=mission)

v0.3.0 (25 Jan. 2023)#

By G. Maze and K. Balem.

This last release is a major one. It introduces new features and breaking changes in the API.

New features

  • New Argo float configuration manager. It was designed to make easier the access, management and backup of the virtual floats mission configuration parameters. All details are available on the API page FloatConfiguration and the documentation page “Argo floats mission parameters”.

cfg = FloatConfiguration('default')  # Internally defined
cfg = FloatConfiguration('cfg_file.json')  # From json file
cfg = FloatConfiguration([6902919, 132])  # From Euro-Argo Fleet API
cfg.update('parking_depth', 500)  # Update one parameter value
cfg.params  # Return the list of parameters
cfg.mission # Return the configuration as a dictionary, to be pass on a VirtualFleet instance
cfg.to_json("cfg_file.json") # Save to file for later re-use

  • New Argo virtual floats type: this new float type can change their mission parameters when they enter a specific geographic area (a rectangular domain). In order to use these floats, you can load a FloatConfiguration instance with the local-change name, like this:

cfg = FloatConfiguration('local-change')
cfg.update('area_cycle_duration', 120)  # Update default parameters for your own experiment

where you will note the added properties area_*:

<FloatConfiguration><local-change>
      - area_cycle_duration (Maximum length of float complete cycle in AREA): 120.0 [hours]
      - area_parking_depth (Drifting depth in AREA): 1000.0 [m]
      - area_xmax (AREA Eastern bound): -48.0 [deg_longitude]
      - area_xmin (AREA Western bound): -75.0 [deg_longitude]
      - area_ymax (AREA Northern bound): 45.5 [deg_latitude]
      - area_ymin (AREA Southern bound): 33.0 [deg_latitude]
      - cycle_duration (Maximum length of float complete cycle): 240.0 [hours]
      - life_expectancy (Maximum number of completed cycle): 200 [cycle]
      - parking_depth (Drifting depth): 1000.0 [m]
      - profile_depth (Maximum profile depth): 2000.0 [m]
      - vertical_speed (Vertical profiling speed): 0.09 [m/s]

Passing this specific FloatConfiguration instance to a VirtualFleet will automatically select the appropriate Argo float parcel kernels (app_parcels.ArgoFloatKernel_exp). This new float type was developed for the EA-RISE WP2.3 Gulf-Stream experiment.

  • All Argo float types (default and local-change) now come with a proper cycle number property. This makes much easier the tracking of the float profiles.

Utilities:

  • utilities.simu2index, utilities.simu2csv: An Argo profile index extractor from the simulation netcdf output. It is not trivial to extract the position of virtual float profiles from the trajectory file of the simulation output. We made this easier with these functions. It also comes bundled with the VirtualFleet.to_index method.

  • utilities.set_WMO: A function to identify virtual floats with their real WMO from the deployment plan. This could be handful if the deployment plan is actually based on real floats with WMO.

  • utilities.get_float_config: A function to retrieve Argo float cycle configuration using the Euro-Argo meta-data API: .

Breaking changes

  • Huge internal refactoring, with proper submodule assignment !

  • The former VelocityField function to work with known velocity fields is now Velocity(). The new VelocityField refers to the class used to manage a velocity field.

  • Options in VirtualFleet:

    • instantiation argument vfield has been replaced by fieldset and now must take a parcels.fieldset.FieldSet or a VelocityField instance.

    • the VirtualFleet.simulate() method has been refactored to use more explicit arguments and now takes datetime.timedelta as values, instead of mixed/confusing integer units.

v0.2.0 (30 Aug. 2021)#

By K. Balem

# Mission parameters
parking_depth = 1000. #in m
profile_depth = 2000.
vertical_speed = 0.09 #in m/s
cycle_duration = 10. # in days

mission = {'parking_depth':parking_depth, 'profile_depth':profile_depth, 'vertical_speed':vertical_speed, 'cycle_duration':cycle_duration}
VFleet = vaf.virtualfleet(lat=lat, lon=lon, depth=dpt, time=tim, vfield=VELfield, mission=mission)

v0.1.0 (29 Jun. 2020)#

By K. Balem

This is the first release of Virtual Fleet with a single kernel (type of virtual Argo float) available and all its parameters are set internally.