Fractal floc properties in the surface waters of a partially-mixed estuary: insights from video settling, LISST, and pump sampling

This study focuses on observations collected along the York River Estuary, major tidal tributary of the Chesapeake Bay, USA. Although microtidal in terms of tidal range, the partially-mixed York River Estuary experiences near-surface tidal currents of up to 1 m/s at spring tide (Friedrichs, 2009). Observations of particle properties were collected using a profiling system that includes a Laser InSitu Scattering and Transmissometry (LISST) 100X Type C instrument, a high-definition Particle Imaging Camera System (PICS) incorporating a video settling tube, and a high-speed pump sampler. At each sampling station, the profiler was lowered to a depth of 1-3m below the surface, and kept there while the suite of instruments sampled for 2-5min. With the profiler located above the pycnocline, observed particles were more likely to be dominated by slowly settling flocs, without major influence of heavier non-flocculated particles, such as the compacted mud pellets commonly resuspended in the lower water column of the York River Estuary. Results reported here were collected at a total of 45 samples stations on nine cruises conducted in the York between September and December over the course three years, 2014-2016. Samples were collected in the fall and early winter to avoid phytoplankton blooms, which are most likely to occur during spring and summer in this system. An aim of this study is to examine properties of flocculated particles without observations being significantly confounded by the presence of relatively large, intact algal cells. The PICS video camera system collected in-situ measurements of particle settling velocity and particle diameter (df ), which were then combined with standard settling relationships to estimate floc particle excess density, Δρ, as a function of df (Smith and Friedrichs, 2015). Independent particle size distribution (PSDs) measured by the LISST and PICS were merged to capture a potentially larger size range of particles than could be observed by either instrument alone (2.5- 1000µm), and to help account for some of the known limitations associated with both instruments. Specifically, LISST results were used for smaller flocs known to be under-represented by the 10µm pixel resolution of the PICS, and PICS results were used for larger particles for which erroneous rising distribution tails are commonly produced by the LISST. The LISST scattering property kernel matrix for random shaped particles was used for the inverting the LISST data because the inversion matrix for spherical particles was found to produce unrealistic rising distribution tails toward the - 2 - small end of the PSD. Water samples collected by the profiler’s high-speed pump were analyzed for total suspended solids (TSS) and organic content via standard gravimetric analysis. Bulk estimates of apparent density (ρa_bulk) for each sample was then estimated by dividing TSS by total volume concentration (VCT ), with VCT determined from the merged LISST and PICS PSD