Code for SuperDARN Borealis 3.5s integration into lompe, with plotting functions, otherwise known as the Fast Borealis Ionosphere (FBI). Can also be used to read in SuperDARN FitACF level data and put it into a format that Lompe accepts.

FBI data can be obtained from superdarn.ca, which can be read in with this package.

Installation

Install with the following command:

pip install git+https://github.com/billetd/FBI.git

All pre-requisites will be installed, if they are not already.

Important note: A special forked branch of Lompe is installed, with some small changes to make FBI work properly. If you already have Lompe installed in the enviroment you are using FBI, uninstall before installing FBI. I recommend a fresh virtual enviroment.

Creating an FBI output file

See process_example.py for an example of creating an FBI HDF5 file from SuperDARN data. The code was written to work with Borealis wide-beam data, but should work with any SuperDARN operating mode.

The HDF5 file created contains the Lompe fits to the SuperDARN data used as input. See readwrite.py for how this is formatted.

Code for making plots is also included. See plotting_example.py for an example of reading in an FBI HDF5 file and making plots.

Basic usage for reading FitACF into a Lompe format

If all you want to do is read in fitacf SuperDARN files and incorporate them into lompe, along with data from other sources, your code should look something like this:

import datetime as dt
import apexpy
import lompe
from FBI import fitacf
from FBI.process import prepare_lompe_inputs


all_data = fitacf.read_fitacfs(fitacfs, cores=5)  # This reads in data from a list of fitacf files you make
time = dt.datetime(2016, 5, 6, 2)  # Time of interest. Change accordingly.
apex = apexpy.Apex(time, refh=300)  # Make an apexpy object for coordinate transforms 
superdarn_data, _ = prepare_lompe_inputs(apex, all_data, time, 120, True)  # Make the lompe data object

# Pass to Lompe and make a fit. This assumes you’ve already defined conductances and have your own grid set up (See Lompe github for more details).
model = lompe.Emodel(grid, Hall_Pedersen_conductance=(SH, SP)) 
model.add_data(superdarn_data)  # Add other data objects as appropriate
model.run_inversion(l1=10, l2=0.1, lapack_driver='gelsy') # Adjust regularisation as appropriate