Geoscience Reference
In-Depth Information
Figure 4.2. (a) GO-FLO water sampler. (b) Niskin bottle.
recommended as particles can interfere with CDOM measurements, for example, scatter-
ing of light (Blough et al.,
1993
; Chen and Gardner,
2004
). There is no consensus on the fil-
ter cutoff size to define DOM operationally, and studies have ranged in the filter cutoff size
used in the aquatic sciences from 0.1 to 1.2 μm, dependent largely on the field of research
and the goals of the study. The vast majority of studies though have used filters in the
0.2-0.7 μm cutoff range. The requirements of an ideal filter have been previously described
by Kremling and Brugmann (
1999
); modified for DOM analysis an ideal filter should have
a uniform and reproducible pore size, a high filtration rate and not clog quickly, not adsorb
any of the DOM to be determined or contain any constituents to be determined, have rea-
sonable mechanical strength, not shed fibers, and in addition it should be easy to clean.
Filters fall into two main categories; depth filters or sieve filters (Kremling and Brugmann,
1999
). With respect to depth filters the pore sizes are poorly defined and separation of parti-
cles from the solution is dependent on physical trapping and the amount of surface contact.
Such filters have a nominal pore size (i.e., in reality there is a range of pore sizes around
the reported cutoff), although this can often be very close to the nominal cut-off (e.g., 0.6
µm-0.8 µm for Whatman GF/F 0.7 µm filters) and are typically made of cellulose, metal
