Nice question indeed. I have no answer, but the first results we get from an ADCP in the lower part of the Rhone river indicate that it seems to be very difficult. This is especially right for such environment variable in few hours during a flood. All our samples collected for SPM calibration of ADCP are also analyzed for grain-size. We hope to obtain a database sufficient to work on it.

Yes, nice question. The relationship between suspended agregates and their optical signature depends on concentration, refractive index and particle size distribution. The main radiative transfer models used in the ocean use by default a particle size distribution (PSD) of Junge's type (power law), but coastal turbid waters are far from this hypothesis, and you can not properly estimate one parameter (C, mr or PSD) without knowing the others. There is a second main application for modelling purpose: particle size distribution (PSD) can be used to estimate the settling velocity distribution to calibrate/validate a transport model - of course, the settling velocity distribution is more appreciable in this way than the PSD-. I also think from readings that small particles are more precisely measured using optics and coarse particles using acoustics, that's why PSD matters.

Totally agree with Sylvain.

The OBS measurements is firstly driven by concentration, and then by size and refractive index. The respective impact of these two parameters slightly depend of the kind of OBS you have also (different angles), and strongly depends on the acceptance angle (a big difference in terms of concentration can be observed depending on the acceptance angle, which as been published with transmissometers). Spectral information of such measurements would help to get information about size of your particualte assemblage. Instead of OBS you can use scattering and backscattering meters simultaneously to get rough information on size, and refractive index, and accurate information on concentration.

The Power-law distribution does not apply to aggregate. The sedimentology literature is full of example of in-situ PSDs that have a large hump in the floc fraction. High frequency acoustics is a great tool to look at large grains of sand. Very inefficient for flocs due to what seems like sound absorption (work that Clemantina Russo and I have tried to publish for the last 3yrs and has been blocked every time by interested reviewers). Optical parameters seem to conserve optical property/mass ratio through the process of aggregation (see latest paper by Slade in 2011 in Optics Express). Combining LISST and a camera seems to provide the most consistent in-situ PSDs of suspended sediments (works by Paul Hill and colleagues).

Particles size matters for many things - from settling velocities to vision (you can't see thorugh fog but if you package the same mass of droplets in large rain drops you can). Surprisingly, for optics we see no change when we create large fluffy aggregate (you will expect change in properties/mass if you formed solid particles). For acoustics the change is very important and we observe it in both lab and field.

One way to obtain a reasonable estimate of the ratio between sortable silt particles and aggregates is to measure grain-size distributions in situ by means of a LISST particle sizer while, at the same time, extracting suspended matter by pump sampling at the same location and height above the bed. The latter samples are then dispersed and analysed, e.g. by Sedigraph. The difference between the two curves gives an indication of the two components. See for example: Chang et al. (2006), Sedimentary Geology 202: 425-435. See also the attached figure.