This code calculates the locations of 1D edges and ridges in 2D atmospheric data.
An example is given that calculates the locations of the Congo Air Boundary (CAB) and the Kalahari Discontinuity (KD) in era-5 reanalysis
as per https://doi.org/10.1175/JCLI-D-19-0437.1.
It is envisaged that this example may be used as a starting point when generalising to other regions, variables and datasets. 

canny_mod.py             contains base edge/ridge detection algorithms
drylines.py              contains useful routines and functions to apply algorithms
get_input.py             downloads era5_nearsurface_2016.nc 
identify_cab_kd.py       finds CAB and KD in 1 month of ERA-5 reanalysis

To get started with the example case, set up the copernicus web api (https://cds.climate.copernicus.eu/api-how-to) if you haven't already.
You'll need to register to do this. 
Install dependencies, including iris, cartopy, cmocean and skimage. 
(of these, only skimage is truly required. You can fairly easily remove the former three packages by editing identify_cab_kd.py)
Then you can run get_input.py to download some sample data. 
Running identify_cab_kd.py will then identify drylines and convergence lines in the sample data. 
It does so by importing code from drylines.py, which in turn imports from canny_mod.py, so make sure these are in the same directory. 
You won't need to run canny_mod.py or drylines.py directly.


Note: the Kalahari Discontinuity (KD) was previously referred to as the Trans-Kalahari Discontinuity (TKD) before it was pointed out to 
me that not every acronym needs three letters. Some instances of tkd may still exist in the code and/or doc.