Post-processing tools for ocean model outputs.
galene uses Iris as it's
data model. Iris is easiest to install with Anaconda Python. These instructions have been tested for Anaconda version Anaconda3-2019.03.
It recommended to create a specific conda environment for Iris, for example:
conda create --name iris python=3The environment must be activated prior usage:
conda activate irisThen we can install Iris. See the Iris website for up-to-date installation instructions.
conda install -c conda-forge irisOnce Iris is installed, and the iris environment is active, install
galene with
pip install -e /path/to/galene/- time series plots
- vertical profile plots
- comparison of datasets, interpolation on common grid
- computation of statistics
- Taylor and target diagrams
- Geographical plots (maps)
galene data model uses Iris Cube objects to represent data
(see Iris documentation).
In addition, two metadata entries are required:
cube.attributes['dataset_id']: A string that identifies the data set, e.g.mynemorun1orobservationscube.attributes['location_name']: A string that identifies the spatial location of the data, e.g. station or transect identifier.
Currently these kinds of geospatial data are supported (dimension coordinates in parentheses):
- timeseries: (time) + auxiliary scalar depth
- surfacetrack: (time) + auxiliary latitude, longitude, scalar depth
- profile: (depth) + auxiliary scalar latitude, longitude
- timeprofile: (time, depth)
- transect: (depth, index) + auxiliary latitude, longitude, depth
- timetransect: (time, depth, index)
All data types are stored as a Cube objects. The function get_cube_datatype(cube) returns the data type as a string.
galene can read netCDF files that contain sufficient metadata.
It is recommended to first generate Iris Cube objects and then store them to
disk.
To support reading in various model output files, some metadata editing may be
necessary. See galene/nemo_reader.py.