Sensitivity of response of LISST-ABS and OBS to size distribution changes

0

Courtesy of Sequoia Scientific, Inc.

This Note shows how the calibration of the LISST-ABS acoustic backscatter sensor and of optical backscatter sensor (OBS) changes when water contains a broad distribution of particle sizes, and why LISST-ABS is inherently superior.

Particle size distributions in water are usually not very narrow, i.e. particles of more than a single size exist in suspension. It is common practice to consider ‘log-normal’ size distributions. A log-normal size distribution means that the distribution has a normal distribution (Gaussian form) with the log of particle sizes. The log of particle sizes is the Φ scale used by geologists, defined as:

Φ = – log2 (d);   and d is in mm.

The figure below shows 4 such distributions. Their mean sizes are chosen to be different so that the curves are separated. The width of the log-normal distribution is indicated in the legends box as ΔΦ. The mean size is indicated as a Φ value on each curve. Note, a small ΔΦ is a narrow distribution and vice versa. The amplitudes of these curves adjust to equalize total concentration (area under the curve) to unity.

Sensitivity of response of LISST-ABS and OBS to size distribution changes

Using the weighted sum method to calculate the acoustic response described on the Application Note, Understanding LISST-ABS, we show below a contrast in how calibrations would vary for the two technologies as sediment mean size and size spread changes.

In the figure below with a linear axis for sizes, which highlights the large grain sizes, the ABS calibration is seen to be very stable for all mean sizes and size spreads. In contrast, the OBS response shows a strong decrease in sensitivity  with increasing mean grain size, and for all size spreads. So, this shows that for large particles (>32 microns), the ABS is definitely superior in holding calibration.

Sensitivity of response of LISST-ABS and OBS to size distribution changes

In contrast, for small mean sizes, i.e. Φ>5 (or diameter <32 microns) the calibration sensitivity to sizes becomes clearer in the next figure. Now, the size axis is logarithmic, highlighting calibration for particles smaller than 32 microns (Φ>5).

Sensitivity of response of LISST-ABS and OBS to size distribution changes

It is seen that for particles smaller than Φ=5, both technologies show a significant variation in calibration. For the smallest particles, however, i.e. Φ>6 calibration change in OBS is smaller than ABS, but of course, this is at the cost of severely reduced sensitivity to large grains.

To summarize, for particles with mean diameter larger than about 32 microns, and irrespective of the width of size distribution, the LISST-ABS calibration is not affected significantly with a change in mean grain size or size spread. Only for limited applications, Φ>7 (<8 microns), the OBS holds calibration better.

As an example from field data, OBS type sensors don’t see larger grains that occupy the deeper part of a river column. The LISST-ABS would see this well!

Customer comments

No comments were found for Sensitivity of response of LISST-ABS and OBS to size distribution changes. Be the first to comment!