NortekSupport/pynortek

Python toolbox to read and process data files recorded with acoustic current profilers from Nortek, e.g. Vector ADV, Aquadopp

Install

The package was developed using python 3.5+, it might work with earlier versions, but its not supported.

User

Install as a user

python setup.py install --user

Uninstall as a user

pip uninstall pynortek

Developer

Install as a developer

python setup.py develop --user

Uninstall as a user

pip uninstall pynortek

FEATURES

Pynorteks main functionality are the binary reading of ADV Vector (.vec) files and storing in the netCDF format, which is designed to handle huge datasets. The original Nortek software converts the data into text files, which become unhandy when the dataset increases in size. The main feature during the conversion process is to add a timestamp to all velocity measurements. The raw binary velocity data does not include time stamps. To do so packages with timestamp (sys, velocity header) in the vicinity of the velocity package have to be found and, depending on the samplingrate, timestamp of the velocity package are calculated.

EXAMPLES

The main function is pynortek_vec2nc, arguments are the input file and the netcdf filename, i.e.:

pynortek_vec2nc advfile.vec advfile.nc

Which results in a netcdf file called advfile.nc. One can also merge several split datafiles into one netcdf file,

pynortek_vec2nc advfile1.vec advfile2.vec advfile3.vec advfile.nc

The conversion can also be done within a python script.

import pynortek
pynortek.bin2nc(advfile.vec,advfile.nc)

Plotting netCDF4 files

Assuming that pynortek_vec2nc created a netCDF file called advfile.nc. The following small program plots the first velocity component.

import pylab as pl
import netCDF4
import matplotlib.dates as mdates

nc = netCDF4.Dataset('advfile.nc')
tsys = nc.groups['sys'].variables['time'][:] # Unix time
tvel = nc.groups['vel'].variables['time'][:] # Unix time
v1 = nc.groups['vel'].variables['v1'][:] # The v1 velocity
burst = nc.groups['vel'].variables['burst'][:] # The burst sample

pl.figure(1)
pl.clf()
pl.subplot(2,1,1)
pl.plot(tvel,v1)
pl.xlabel('unix time [s]')
pl.ylabel('u [m/s]')

pl.subplot(2,1,2)
for i in range(max(burst)):
ind = burst == i
pl.plot_date(pl.epoch2num(tvel[ind]),v1[ind],'-')

pl.xlabel('Date')
pl.ylabel('u [m/s]')
pl.draw()
pl.show()

NortekSupport / pynortek