A device for sediment concentration measurement with fixed volume overflow sampling

Sediment concentration is one of the most crucial physical parameters for the soil and water loss and soil erosion. However, it is still lacking on the accurate and rapid measurement on the sediment concentration at present. In this study, the novel device was designed with a constant volume sampling bottle. Three subsystems included the fixed-volume overflow sampling, weighing measurement and control subsystem. The developed device was small and easy to carry for the sediment concentration measurement. According to the volume-mass conversion, sediment concentration was calculated to accurately measure the volume and mass of the water-sediment mixture. Specifically, the total mass of runoff sample was the sum of sediment and water mass, and the total volume was equal to the sediment volume plus water volume. In order to avoid the impact of sediment particle size on measurement, the total mass of runoff sample was determined by a weighing sensor; the total volume of runoff sample was equal to the sampling bottle volume. The sampling bottle volume was precisely determined to fill the chamber with the distilled water at different temperatures. The results showed that the sampling bottle volume was averaged 1.01 L, and the volume varied from 0.96 to 1.05 L, with an average Coefficient of Variation of 3.46%. After that, the experiment was conducted on the samples with different sediment concentrations, in order to test the sensitivity and accuracy of the device. The result showed that the concentrations of 20 g/L were measured for 40 times, and the average concentration was 19.94 g/L. The histograms of the two target sediment concentrations showed the normal distributions, indicating a high precision of the device. Then, a linear regression was fitted to measure the concentrations with a slope of 0.981 and R² of 0.999. Additionally, the sediment concentrations were measured with the averaged relative error of 3.64% and the highest precision of 99.75%. Specifically, the averaged relative errors were 0.81% for the high sediment concentration group in the range from 8 to 800 g/L. While the sediment concentration of 1 and 2 g/L was resulted in the lager errors of 16.00% and 15.67%, indicating that this device was limited to the sediment concentrations less than 2 g/L. The better confidence was achieved in the high repeatability and accuracy of the device. The frequency distribution of relative errors showed that the small value was observed in the most relative errors, where the samples with the relative errors ＜5% were accounted for 74.36% of the total. Moreover, the developed device was used to collect and measure the instantaneous sediment content of Yarlung Zangbo River from Lhasa to Shigatse and some reaches of Lhasa River and Nianchu River. Meanwhile, the oven drying and the device were used to carry out synchronous sampling and measurement. The results show that the sediment concentration measured by the developed device was much higher than that by oven drying, with the average relative error of 6.13%. Furthermore, the average instantaneous sediment concentration was 9.82 g/L measured by the development device. The average instantaneous sediment content of Yarlung Zangbo River was significantly higher than that of Lhasa River and Nianchu River, which were 12.95, 7.19 and 8.91 g/L, respectively. Filed experiment showed that the instantaneous measurement on sediment concentrations in different rivers was extremely higher than the rest from previous studies, indicating the intensified soil and water loss in this region. The device can be expected to measure the sediment concentration with the high accuracy in near real-time. The finding can provide a strong reference for the regional soil erosion as an alternative way with the excellent potential.