342
filter will substantially improve results in oligotrophic waters dominated by small-sized
particles.
Second, special caution must be exercised to correct for the baseline and minimize baseline
shifts. Because the optical density of blank filters, even from a specified lot of filters, can
vary by as much as 0.05, the baseline should be measured separately for each filter prior to
filtration. Also, it is not sufficient to rely on the automatic baseline correction which is built
in to most of the commercial spectrophotometers. The spectrum of the baseline should be
saved to a file, and then subtracted from the sample spectrum when processing the data. This
procedure requires the use of a spectrophotometer on board the ship which is consistent with
another recommendation that favors the immediate analysis after sampling rather then
long-term storage of samples and post-cruise analysis. In addition, the filters should be
positioned exactly in the same way for the baseline and sample scans (a small marker on the
edge of the filter is helpful).
Third, it may be advantageous to measure the optical density on the filter against air rather
than against the blank filter whenever a geometrical configuration of the instrument allows
such a measurement. In this case, the reference beam is kept exactly constant, so the possible
shifts of the baseline associated with the blank filter (e.g. changes in water saturation) are
avoided.
Finally, the commonly used null correction of absorption spectra at 750 nm does not appear
to be justified for total particulates. A number of measurements made at sea with careful
determinations of baseline for each filter separately (D. Stramski, oligotrophic Pacific, unpubl.
data), showed that the absorption by total particulates at 750 nm is significantly different from
zero. Only one of 35 measurements, which had been corrected for baseline, exhibited negative
value of optical density at 750 nm, the remaining were positive. The significance of this
observation is supported by the fact that the optical density on a filter at 750 nm differs
insignificantly from zero when measured with the same care and precision with healthy
phytoplankton cultures free from other particles.
filter will substantially improve results in oligotrophic waters dominated by small-sized
particles.
Second, special caution must be exercised to correct for the baseline and minimize baseline
shifts. Because the optical density of blank filters, even from a specified lot of filters, can
vary by as much as 0.05, the baseline should be measured separately for each filter prior to
filtration. Also, it is not sufficient to rely on the automatic baseline correction which is built
in to most of the commercial spectrophotometers. The spectrum of the baseline should be
saved to a file, and then subtracted from the sample spectrum when processing the data. This
procedure requires the use of a spectrophotometer on board the ship which is consistent with
another recommendation that favors the immediate analysis after sampling rather then
long-term storage of samples and post-cruise analysis. In addition, the filters should be
positioned exactly in the same way for the baseline and sample scans (a small marker on the
edge of the filter is helpful).
Third, it may be advantageous to measure the optical density on the filter against air rather
than against the blank filter whenever a geometrical configuration of the instrument allows
such a measurement. In this case, the reference beam is kept exactly constant, so the possible
shifts of the baseline associated with the blank filter (e.g. changes in water saturation) are
avoided.
Finally, the commonly used null correction of absorption spectra at 750 nm does not appear
to be justified for total particulates. A number of measurements made at sea with careful
determinations of baseline for each filter separately (D. Stramski, oligotrophic Pacific, unpubl.
data), showed that the absorption by total particulates at 750 nm is significantly different from
zero. Only one of 35 measurements, which had been corrected for baseline, exhibited negative
value of optical density at 750 nm, the remaining were positive. The significance of this
observation is supported by the fact that the optical density on a filter at 750 nm differs
insignificantly from zero when measured with the same care and precision with healthy
phytoplankton cultures free from other particles.