Economical DIY optical backscatter sensors for measurements of turbidity and sediment concentrations

Abstract:

Fluvial and coastal sediment is a critical currency in the transport of nutrients and pollutants, and helps aggrade marshes and deltas which serve as buffers against sea-level rise. Tracking sediment fluxes through aquatic systems is key to understanding storage and dispersal, but flux measurements often require the deployment of multiple costly sensors, or reliance on remote-sensing products which require robust, labor-intensive calibration datasets from in situ measurements at diverse locations. Commercial turbidity sensors (typically optical backscatter sensors, or OBSs) are widely available for these applications, but autonomous versions cost ~$3000. The advent of microcontrollers like Arduino and Raspberry Pi have allowed rapid developments in low-cost, “do-it-yourself” sensor construction, even for rugged field deployments. In this project, we leverage the Arduino platform and a custom-designed circuit to create an OBS that can be constructed using readily available components for ~$60. The circuit uses a near-infrared diode emitter and receiver to capture a voltage signal which is linearly related to the turbidity of a fluid solution. The signal is shifted and amplified using an operational amplifier, and recorded using a high-resolution analogue-to-digital converter. Different resistance values are used to optimize the signal gain for different fluid turbidities. A laboratory comparison was performed between the DIY sensor and a commercial OBS, using mixtures of natural sediments (diameter <0.063 mm) in concentrations ranging from 2 to 1500 mg/L. The commercial OBS yielded measurements with 10-30% standard deviation from the mean for turbidities <15 NTU, and 3–5% standard deviation from the mean for turbidities of 30–770 NTU. The new DIY OBS yielded measurements with standard deviation of 2-4% for the lower range and 2–3% for the upper range, representing comparable results for most turbidities measured and better results with less noise for the lower values. The commercial and DIY OBS concentration measurements were linearly related through C_DIY (volts) = 0.749 x C_OBS (NTU/1000) + 0.12, with R²=0.998. This new sensor thus offers an effective, low-cost advancement in sensing capability for diverse water-quality monitoring, geologic investigation, and remote-sensing calibration needs.