The sedimeter™ sensor invented by Dr. Ulf Erlingsson in 1985 for detecting incipient sediment motion on the sea floor. It is probably easier to show with a drawing how it works:
The sensor is anchored with the help of a holder tube that is screwed down into the bottom, thus assuring that it is not the instrument, but the bottom that changes level. We now calibrate the response in FTU (Formazin Turbidity Units), or FBU for Formazin Backscatter Units, with the sensor outside of the holder tube. The holder tube adds a significant amount of reflection, but since we are interested in the bottom level it does not matter to us.
Already from the first prototype (which actually used a power-hungry Z80 NMOS CPU!) it became clear that the sensor was able to detect incipient sediment motion: It reacted to the addition to a single grain of sand adjacent to the sensor!
Some users are also interested in the turbidity in the water column, and for that reason there is an extra turbidimeter about 11 cm higher up. In SM3 this extra turbidimeter is identical to the others, measuring NIR backscatter, whereas in SM4 there are two nephelometric turbidimeters, both an ISO style measuring 90 degree reflected NIR light, and an EPA style measuring 90 degree reflected white light.