160
For convenience, fluorescence excitation spectra are determined for
the same samples as absorption.
Since a relatively large volume of seawater must be filtered for a determination of a p ' measurements of fluorescence on the same sample are subject to serious quenching by reabsorption and attenuation of the light field to which the particles are exposed.
The correction algorithm for these two phenomena were determined by filtering from 1 to 50 mls. of an exponentially growing culture of Q. tertiolecta onto GF/C filters.
Fluorescence and absorption spectra were determined for each sample. Also, determination of the optical density of a
blank single layer GF/C filter, relative to air, was carefully made by
placing the filter directly on the diffusing plastic resulting in a distance of only about ,3 mm. from the surface of the filter to the photocathode film on the photomultiplier entrance window. This ensured that
all forward scattered light was accepted by the photomultiplier.
III. RESULTS
A. Volume absorption coefficient algorithm
For our routine measurements of absorption for a double layer of a
sample filter which has been cut in half and stacked, the first correction which must be applied to the raw data is to generate a true spectrum of the equivalent material on a single layer. According to Beers
Law for absorbing materials in pure solution, one expects a doubling of
the optical density when the geometric path length is doubled.
However,
since the glass fiber filters used in this analysis are highly diffusing,
the concepts for pure solutions are not applicable.
In fact, because
of the large scattering cross-section of the filters, the actual optical
pathlength of the sample is greater than the geometric pathlength. Butler
(1962) defined the term B as the ratio of the optical to geometric pathlength. Stavn (1981) defined this parameter as the mean pathlength and
used it to describe absorption of light in scattering media. For our
measurements on glass fiber filters, we have divided the correction into
two parts: the correction for BCA,l) for a single layer relative to a
minimally scattering suspension, and B*(A, 2) for a second layer relative
to a single layer. B*(A,2) is not a true B in the sense of Butler(1962)
because both the single and double layer measurements are concerned with
diffuse transmittance, while Butler was concerned with the amplification
for a diffuse transmittance relative to a collimated transmittance.
For ES and E B , the measured transmitted irradiance for the sample
and blank, respectively, the optical density for the sample is defined
as:
For convenience, fluorescence excitation spectra are determined for
the same samples as absorption.
Since a relatively large volume of seawater must be filtered for a determination of a p ' measurements of fluorescence on the same sample are subject to serious quenching by reabsorption and attenuation of the light field to which the particles are exposed.
The correction algorithm for these two phenomena were determined by filtering from 1 to 50 mls. of an exponentially growing culture of Q. tertiolecta onto GF/C filters.
Fluorescence and absorption spectra were determined for each sample. Also, determination of the optical density of a
blank single layer GF/C filter, relative to air, was carefully made by
placing the filter directly on the diffusing plastic resulting in a distance of only about ,3 mm. from the surface of the filter to the photocathode film on the photomultiplier entrance window. This ensured that
all forward scattered light was accepted by the photomultiplier.
III. RESULTS
A. Volume absorption coefficient algorithm
For our routine measurements of absorption for a double layer of a
sample filter which has been cut in half and stacked, the first correction which must be applied to the raw data is to generate a true spectrum of the equivalent material on a single layer. According to Beers
Law for absorbing materials in pure solution, one expects a doubling of
the optical density when the geometric path length is doubled.
However,
since the glass fiber filters used in this analysis are highly diffusing,
the concepts for pure solutions are not applicable.
In fact, because
of the large scattering cross-section of the filters, the actual optical
pathlength of the sample is greater than the geometric pathlength. Butler
(1962) defined the term B as the ratio of the optical to geometric pathlength. Stavn (1981) defined this parameter as the mean pathlength and
used it to describe absorption of light in scattering media. For our
measurements on glass fiber filters, we have divided the correction into
two parts: the correction for BCA,l) for a single layer relative to a
minimally scattering suspension, and B*(A, 2) for a second layer relative
to a single layer. B*(A,2) is not a true B in the sense of Butler(1962)
because both the single and double layer measurements are concerned with
diffuse transmittance, while Butler was concerned with the amplification
for a diffuse transmittance relative to a collimated transmittance.
For ES and E B , the measured transmitted irradiance for the sample
and blank, respectively, the optical density for the sample is defined
as: