It might be beneficial to move the map code, velocity files, and other datasets to a supporting notebook.

I finally got arrows plotted (kind of) -  they are offset, and I am not sure why. It could be related to the projection in the svgDrawVectorField function.

I now have a feeling the vectors are centered on the station locations - I need to change it so that the arrow starts at the station location.

I removed the metadata info before the header for each file. To see full description, download the files from UNAVCO (I've linked them above). I also converted the velocities and errors to mm/yr. 

Each station has multiple entries as the velocities and positions are updated through time. I believe going with the last entry in time for each station is a good way to go for plotting the vectors.

Note that the pbo velocity vector files available on the existing VV site seem to be filtered already. So, I do not believe it is plotting different datasets for the pbo filtered and unfiltered versions, but I would like to check with someone because it is difficult to verify by viewing the map. Currently, I am only using what I believe are the filtered files.

I have asked to make sure these files on Gitlab are OK for use

If they are, the structure of these files, from the gitlab source, appears to be 

{'stat_id': l[7], 'latitude': float(l[1]), 'longitude': float(l[0]),
                 'altitude': 0.0, 'vel_east': float(l[2]), 'vel_north': float(l[3]), 'vel_up': 0.0,
                 'stddev_north': float(l[5]), 'stddev_east': float(l[4]), 'stddev_up': 0.0, 
                 'ne_corr': float(l[6])}

I added headers to these files so that they follow the same naming conventions as the UNAVCO files, are in csv format.

I need to inquire about the North America and NNR specification here. NNR in my mind is necessarily global, but it could be that there are two steps, going from NNR --> NA

GPS_NNR.csv

cwu.final_nam14_cut.csv

GPS_AF.gmt.csv

GPS_AN.gmt.csv

GPS_AR.gmt.csv

GPS_AU.gmt.csv

GPS_CA.gmt.csv