There are two different types of spikes that may happen in CTD data:
- Bad data from electrical failures. Usually it manifest in all the measured variables (Temperature, Conductivity, Oxygen, etc);
- Salinity spikes due to a wrong temperature value used when computing it from conductivity. This can happen due to a a bad Alignment of the temperature and conductivity sensors, and/or poor Cell thermal mass correction (Lueck 1990).
How to detect the second case? For a start they do not manifest in the raw data. This kind of spike is a product of bad pre-processing parameters (either Align and/or Cell Thermal Mass). For more information on this check the SBE Software documentation and the Go-SHIP manual
import gsw
from ctd import DataFrame
before = DataFrame.from_cnv('./data/CTD-before.cnv.gz', compression='gzip').split()[0]
after = DataFrame.from_cnv('./data/CTD-after.cnv.gz', compression='gzip').split()[0]
# Compute salinity before.
p = before.index.values.astype(float)
before['SP'] = gsw.SP_from_C(before['c0S/m'].values * 10, before['t090C'].values, p)
# Compute salinity after.
p = after.index.values.astype(float)
after['SP'] = gsw.SP_from_C(after['c0S/m'].values * 10, after['t090C'].values, p)
The file CTD-before.cnv.gz is just the raw data acquired be the CTD. The
file CTD-after.cnv.gz had both the Align (-0.01 s and 0.07 s for the
primary and secondary conductivity sensors respectively as provided by the
calibration lab) and the Cell Thermal Mass ($\alpha = 0.03$ for
thermal anomaly amplitude and $\tau = 7$ s for the time constant) corrections
applied.
Here is the result of those "corrections":
kw_before = dict(linestyle='-', color='#339933', linewidth=3, label=r"Practical Salinity Before Align and CellTM")
kw_after = dict(linestyle='-', color='#ff3333', alpha=0.65, linewidth=3, label=r"Practical Salinity After Align and CellTM")
fig, ax = before['SP'].plot(**kw_before)
ax.plot(after['SP'], after.index, **kw_after)
ax.set_xlabel("Salinity [g kg$^{-1}$]")
ax.set_ylabel("Pressure [dbar]")
l = ax.legend(loc="lower right")
fig.set_size_inches((5, 7))
Unfortunately most people apply those corrections using the defaults on the SBE Software and then "correct" the bad corrections with averaging, smoothing, and despiking techniques. A much better solution is to estimate these parameters for your CTD system and avoid them altogether.
HTML(html)
This post was written as an IPython notebook. It is available for download or as a static html.
python4oceanographers by Filipe Fernandes is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Based on a work at https://ocefpaf.github.io/.