Technical note—Relative variability of selected turbidity standards and sensors in use by the U.S. Geological Survey

Abstract

The challenges associated with field measurements of turbidity are well known and result primarily from differences in reported values that depend on instrument design and the resulting need for reporting units that are specific to those designs. A critical challenge for making comparable turbidity measurements is the selection and use of appropriate turbidity standards for sensor calibration. The accepted primary standards for turbidity measurements use formazin made from scratch; all others should relate back to readings obtained using standard formazin. However, because turbidity is a qualitative property of water, comparing standards is not as simple as it is for many chemical measurements. The U.S. Geological Survey “National Field Manual for the Collection of Water-Quality Data” currently allows for the use of two standards, formazin and polymer beads, for the calibration of field turbidimeters. Another challenge for making comparable turbidity measurements is selection of turbidity sensors. A turbidity sensor commonly used in the U.S. Geological Survey, the Yellow Springs Instruments (YSI) 6136, has been replaced by the manufacturer with the YSI EXO turbidity sensor. Both sensors operate on the same principles but have slight design differences that result in readings that are not directly comparable on a 1:1 basis. Differences in calibration standards and sensors are a cause of concern in ongoing studies that require switching calibration standards or sensor types, and for comparisons of data collected with sensors calibrated by using different calibration standards, different sensor types, or both. The objectives of this study were to evaluate the response of two YSI turbidity sensors in both formazin-based standards (StablCal) and polymer turbidity standards (in this case YSI brand; however, other brands are available) and to compare the performance of the YSI EXO and YSI 6136 turbidity sensors under similar laboratory and environmental (field) conditions. To quantify these differences, a series of laboratory and field side-by-side comparisons were conducted. Nine field comparisons of YSI EXO and YSI 6136 sensors were performed at site locations in Kansas and Virginia. Two field comparisons of StablCal and polymer calibration standards were performed in Kansas, both using YSI EXO turbidity sensors. Five laboratory comparisons between the YSI EXO and YSI 6136 turbidity sensors were performed, and seven laboratory comparisons between StablCal and polymer turbidity standards were performed using YSI EXO turbidity sensors. The results can help the USGS and others better understand how turbidity data can differ depending on the sensors and calibration standards used. Key findings and conclusions include the following— Regardless of the comparison, strong linear associations were typically found across all measures of turbidity under field and laboratory conditions, but linear associations were not necessarily 1:1 and varied by type of standard, type of sensor, and field and laboratory conditions. The mean relative percentage differences for all but a few comparisons were greater than 10 percent but less than 30 percent. However, differences were inconsistent across the laboratory and field conditions measured in this study, precluding the ability to formulate definitive statements of consistent directional bias depending on the type of standard or sensor used. Across all tests and a range of 0 to 1,000 formazin nephelometric units (FNU), no consistent bias between the YSI EXO and YSI 6136 turbidity sensors was observed, but either regression relations were near 1:1 or the YSI EXO turbidity readings were lower than those measured with the YSI 6136. Relative percentage differences typically exceeded the 10- and 30-percent benchmarks. However, most (about 95 percent) of the absolute differences between turbidity values measured with the YSI EXO and those measured with the YSI 6136 sensors were less than 20 FNU. In laboratory experiments, when turbidity was “constant,” the absolute difference between YSI EXO and YSI 6136-measured turbidity values ranged from near 0 FNU to nearly 400 FNU (over a range of 0 to 1,000 FNU). Substantial variability in turbidity measurements makes comparison between standards and sensor types challenging. Given the inherent variability in turbidity measurements and the lack of consistent bias between calibration standards or sensors, changing methods during an ongoing study would compromise the comparability of the data. The effect of changing methods ultimately depends on study objectives. If method changes are required, laboratory- and field-based comparisons across a range of conditions ideally would be conducted to determine whether site-specific biases can be identified.