In situ particle size distributions and volume concentrations from a LISST-100 laser particle sizer and a digital floc camera

• Published in: Continental shelf research 2005-10, Vol.25 (16), p.1959-1978
• Main Authors: Mikkelsen, Ole A., Hill, Paul S., Milligan, Timothy G., Chant, Robert J.
• Format: Article
• Language: English
• Subjects: 12°28′E; 14°16′E; 40°45′N; 42°28′N; 44°48′N; 74°W; Adriatic sea; Apennine margin; Flocculation; Hudson river; Italy; Lasers; Marine; New Jersey; Newark Bay; Particle scattering; Sediment transport; Suspended particulate matter; Underwater photography; USA
• ISSN: 0278-4343; 1873-6955
• DOI: 10.1016/j.csr.2005.07.001

A LISST-100 in situ laser particle sizer was deployed together with a digital floc camera during field work in the Newark Bay area (USA) and along the Apennine margin (the Adriatic Sea, Italy). The purpose of these simultaneous deployments was to investigate how well in situ particle (floc) sizes and volume concentrations from the two different instruments compared. In the Adriatic Sea the two instruments displayed the same temporal variation, but the LISST provided lower estimates of floc size by a factor of 2–3, compared to the DFC. In the Newark Bay area, the LISST provided higher values of floc size by up to a factor of 2. When floc size was computed using only the overlapping size bins from the two instruments the discrepancy disappeared. The reason for the discrepancy in size was found to be related to several issues: First, the LISST measured particles in the 2.5–500 μm range, whereas the camera measured particles in the 135–9900 μm range, so generally the LISST should provide lower estimates of floc size, as it measures the smaller particles. Second, in the Newark Bay area scattering from particles >500 μm generally caused the LISST to overestimate the volume of particles in its largest size bin, thereby increasing apparent floc size. Relative to the camera, the LISST generally provided estimates of total floc volume that were lower by a factor of 3. Factors that could explain this discrepancy are errors arising from the accuracy of the LISST volume conversion coefficient and image processing. Regardless of these discrepancies, the shapes of the size spectra from the instruments were similar in the regions of overlap and could be matched by multiplying with an appropriate correction coefficient. This facilitated merging of the size spectra from the LISST and the DFC, yielding size spectra in the 2.5–9900 μm range. The merged size spectra generally had one or more peaks in the coarse end of the spectrum, presumably due to the presence of flocs. The fine end (