PySARTS is a Python package I developed during my undergraduate dissertation while working with interferometric synthetic aperture radar (InSAR) data. The package implements methods for working with timeseries of interferograms and making atmospheric corrections. There's also a variety of plotting methods.
The atmospheric corrections were the focus of my dissertation. I investigated the use of weather radar for atmospheric corrections. Two of the implemented corrections were developed during my dissertation: a correction for the liquid delay using radar data and a correction for the wet and dry delays using the ERA-interim weather model and radar data. An additional "pure" weather model correction is also implemented with an implementation based off TRAIN (with some differences).
I don't expect anyone to use the project for anything professional in its current form. The documentation is seriously lacking, there's no formal tests and there's lots of hacks put in place to meet deadlines. If I get time, I'll try to document the package more thoroughly but otherwise, I don't foresee any further development.
Clone the repository and put it with your other Python packages. Or update your
PYTHONPATH environment variable to point at another location where you've
stored the repository.
The file requirements.txt lists the project's requirements. You can run pip install -r requirements.txt from the repository to install everything. The main
requirements are the SciPy stack, Numba and netCDF
bindings. PySARTS has only been tested with Python 3 but might work with Python
2 too.
PySARTS works with a project directory. This is a directory containing a file
called config.yml that tells PySARTS:
- The study region.
- The resolution to work at.
- The location of your data (if it isn't in the project directory).
- The master study date.
Check out the example_config.yml file in the repository to see what the
configuration looks like. The data paths in the example file are the default
paths. If you put those directories/files in the project directory, you don't
need to configure anything.
As a bare minimum, you'll need the following files:
- A directory containing unwrapped interferograms named
SLAVEDATE_MASTERDATE.ncwhere dates are formatted as%Y%m%d. - A text file containing the perpendicular baselines for the interferograms. This should have three columns: the slave date, master date and the perpendicular baseline.
To do anything with weather radar data you'll need
- A directory containing weather radar images (Cartesian, not radial) named
%Y%m%d%H%M.nc.
For any atmospheric corrections, you're going to need
- A DEM of the target region.
- A directory containing weather model output named
%Y%m%d%H%M.nc.
PySARTS works with NetCDF 4 files. Everything in the previous section needs to be in NetCDF format (except the baselines file). 2D data (interferograms, radar images) needs three variables in the NetCDF:
- Two 1D arrays containing longitudes and latitudes.
- A 2D array containing the data.
These variables can be called either x, y and z OR lon, lat and
Band1. If those variables aren't found, you'll get some cryptic error from
NetCDF.
Weather model output should contain 7 variables. Three of those variables should be the modelled temperature, relative humidity and geopotential. The variable names should be the same as those produced by ERA-interim. If you download output from ERA-interim containing the aforementioned modelled variables, you should be good to go. Note: Only the first time in the model output will be read. You should store separate model times in separate files.
PySARTS exposes a command line interface from the pysarts and pysarts.plot
modules. Assuming your PYTHONPATH is set correctly, you can run
python -m pysarts -h
python -m pysarts.plot -hto get help with the command line interfaces. The help isn't great but it should be enough to get you going. For more details on how things work, check out the documentation strings in the source.
All modules under pysarts except the workflow, plot and config modules
are designed to be usable from other programs or scripts. They make no
assumptions about the existence of a project directory or any files. You should
read the docstrings of the methods to see how to use them.
MIT. Go wild.