This code package performs piecewise PV inversion as described in detail in Teubler and Riemer 2016 in a latitudinal belt on the northern hemisphere. The PV inversion code itself is based on Davis 1992. PV anomalies are defines as upper-level anomalies and low-level anomalies defined as deviations from a 30-day background field.
The development of the PV inversion was part of the PhD thesis of Franziska Teubler under the supervision of Michael Riemer.
- file with instantaneous fields
- file with background information (e.g. temporal mean) to calculate anomalies
The input files can be grib or netcdf and have to contain u,v,T and Geopotential on pressure levels. The programm will automatically interpolate on 17 pressure levels needed for PVInversion. The more pressure levels are available the better.
The output file will contain u and v for the balanced, the background, the upper-level, and low-level flow component (Geopotential can be added easily, if required).
To run the code the following python packages need to be installed:
$ conda -c conda-forge numpy numba netcdf4 h5netcdf xarray eccodes scipy petsc petsc4py
The main file is run_PVI.py. Execute the PVinversion by simply running
$ python run_PVI.py
- I recommend to calculate PV on your own (as it is done here) and do not use PV available in datasets
- Do not try to use a higher resosution than 1°x 1°; for higher resolution there will be no convergence (nonlinear balance can not be reached)
- I recommend a maximal meridional extend from 20°N to 85°N (better 80°N). Further north and south balance condition will not be reached.