Twenty-five years with OBS sensors: The good, the bad, and the ugly

Since their development for investigating sand transport in the surf zone more than 25 years ago, optical backscatter sensors (OBSs) have been in continual use for a wide variety of scientific, engineering, and environmental-monitoring applications. Many OBS designs have evolved in that time but they all work in the same simple way. A water sample is illuminated by a light source and photodetectors in the sensor convert the light scattered from the sample to photocurrent. The amount of photocurrent depends mainly on the illuminated area of the particles, and because particle volume is proportional to area, the photocurrent provides an indirect estimate of suspended-sediment concentration. The relationship between OBS signal and sediment concentration is almost linear for many suspended sediments and this is the most important “good” feature for OBS users. Many other factors influence OBS measurements, including: (1) particle size, shape, composition, and aggregation/flocculation; (2) dissolved light-absorbing matter; (3) bubbles; and (4) chemical and biological fouling. Some of these factors can be troublesome and can result in poor data quality or ambiguous results; these are “bad” ones. And a few of the same factors can have truly “ugly” consequences when their effects are ignored. Unknown time variation in particle size or aggregation/flocculation and fouling will nearly always result in inaccurate OBSs data.