Calibration of the transparency tube for estimating turbidity in estuarine systems
Data files
Jun 06, 2025 version files 13.78 KB
-
Datafile_for_Dryad.csv
7.42 KB
-
README.md
6.36 KB
Abstract
Turbidity, quantified in turbidity units (nephelometric or formazin), is a common and valid measure of water quality related to transparency. A transparency tube (TT) is an economical tool developed to estimate water clarity as an alternative to the Secchi disk, but it is also frequently used to estimate turbidity. Although the relationship between TT measures and turbidity is well characterized for freshwater river and lake systems, this relationship has not been tested for estuarine waters. The objective of the current study was to empirically determine the TT-turbidity relationship for estuarine waters and compare these results with the traditional freshwater system conversions. We obtained 107 measurements of TT depth, turbidity, total suspended solids, CDOM absorbance, Secchi depth, and chlorophyll a at 22 estuarine locations in South Carolina over a one-year period. Linear regressions provide conversion equations that can be applied to estuarine waters. The TT-turbidity relationship for estuaries was compared with freshwater systems. Our results suggest that the slope of the relationship differs between systems (-1.11 vs. -1.41), resulting in different turbidity estimates for TT measurements for estuarine vs. inland waters. We propose a combined conversion table incorporating estuarine and freshwater (riverine and lacustrine) systems. A TT-turbidity conversion for estuarine waters significantly benefits current coastal water quality programs and citizen science groups by producing more accurate turbidity estimates for screening and routine monitoring efforts.
Dataset DOI: 10.5061/dryad.rxwdbrvn9
Description of the data and file structure
The primary objective of our study was to empirically determine the transparency tube (TT) to turbidity (ntu) relationship for estuarine and coastal waters and compare these results with traditional freshwater system conversions. Our aim was to evaluate the accuracy of the freshwater system conversions for estuarine systems. A second objective was to examine the correlations between TT, Secchi depth, total suspended solids, chromophoric dissolved organic matter (CDOM), and the diffuse attenuation coefficient (kd) for photosynthetically available radiation (PAR).
Estuarine locations (22 sites) were sampled for a total of 107 times in 2022 - 23, stretching from North Myrtle Beach to Port Royal Sound along the coast of South Carolina. The goal was to include a range of water types with different turbidity characteristics from tidal regions of coastal waterbodies. Site selection was determined based on accessibility, usually from a boat dock or launch ramp.
Water transparency, photosynthetically available radiation (PAR) depth profiles, salinity, and temperature were obtained on-site. Transparency was measured using a 120 cm “transparency tube” (Eisco Labs, Victor, NY, PN PH1120148120SPR). The transparency tube (TT) was filled with well-mixed water from the sample site using a 1 m bailer and a carboy. While looking through the top opening of the tube, water was slowly drained out of the bottom until the Secchi disk at the bottom of the tube was visible, and the depth of the remaining water (in cm) was recorded. The water column length in the TT was recorded as the average of two independent measurements taken by observers. The Secchi depth was determined using a 20 cm black and white Secchi disk and measured in centimeters. Both Secchi and TT measurement methodologies follow the protocol by Ohrel and Register (2007). A Li-Cor LI-250A portable light meter with a LI-COR LI-193 spherical underwater quantum sensor was lowered into the water column to measure photosynthetically available radiation (PAR, 400 – 700 nm) available throughout the water column (LI-COR, Lincoln, NE, USA). PAR values were recorded as 15-s averages at 0.25 m depth intervals, in units of µmol photons m-2 s-1. Measurements continued until the PAR equaled 0, or, if the sample depth was exceptionally shallow, until the bottom of the water column was reached (the maximum depth for the cable was 8 m). The equation used for calculating the diffuse attenuation coefficient for PAR (kd(PAR)) was:
Ez = Eoe-kz
Where z is water depth (m), Ez is the PAR at a depth (z), Eo is the PAR at the surface, and k is the diffuse attenuation coefficient (Kirk 2011). Salinity was determined using a refractometer, and temperature was measured with a digital thermometer. GPS coordinates, tidal stage, and weather conditions were recorded during sampling. The remaining sample water in the carboy was mixed well, poured into an opaque 1 l Nalgene bottle, and stored in an ice cooler for 3-5 hours before laboratory analysis.
Total suspended solids (TSS), turbidity (NTU), chromophoric dissolved organic matter (CDOM), and phytoplankton biomass (as chlorophyll a, chl a) were measured for each sample. Water for TSS measurements was filtered through pre-weighed filters (GF/F, 47 mm, 0.7 µm nominal particle retention), dried at 45 °C for 24 h, and reweighed to determine the dry weight of suspended solids. Water for chl a was vacuum filtered onto 25 mm filters (GF/F) and stored at -80 ˚C until analysis. For CDOM measurements, 50 ml of water was syringe-filtered through a dried 25 mm GF/F filter and stored at 4 ˚C for no longer than one month. A calibrated Hach 2100Q turbidimeter (PN: 2100Q01) was used to measure turbidity in NTU, following the U.S. EPA Method 180.1 (US EPA 1993). An average of at least three independent readings was recorded. A Jasco V-730 UV-visible spectrophotometer was used to measure the absorbance of CDOM at a wavelength of 355 nm. The absorbance at this wavelength represents CDOM concentration in coastal environments (Blough and Del Vecchio 2002; Kirk 2011). CDOM raw absorbance data were converted into absorption coefficients (a) using the equation (Hu et al. 2002):
a(355) = 2.303 *A(355)/0.05
Where A(l) is the raw wavelength-dependent absorbance, and 0.05 is the optical path length of the cuvette in meters. High-performance liquid chromatography (HPLC) was used to determine chl a. Samples were lyophilized for 24 h at -50 °C, placed in 90% acetone (0.750 ml), sonicated, and extracted at -20 °C for 18 - 20 h. Filtered extracts (250 µl) were injected into a Shimadzu 2050 HPLC equipped with a monomeric (Rainin Microsorb-MV, 0.46 x 10 cm, 3 µm) and a polymeric (Vydac 201TP54, 0.46 x 25 cm, 5 µm) reverse-phase C18 column in series. A nonlinear binary gradient consisting of the solvents 80% methanol: 20% 0.50 M ammonium acetate and 80% methanol:20% acetone was used for pigment separations (Pinckney et al. 2001a).
Files and variables
File: Datafile_for_Dryad.csv
Description: Data are in comma-separated form and available for future analyses and comparisons. Column headers are as follows:
Variables
Site: Name of sampling location in South Carolina, USA
Date: Date the samples were collected
Latitude and Longitude: Sampling location in digital degrees (north latitude and west longitude)
Secchi Depth: Maximum depth (m) for detecting the Secchi disk
Transparency Tube: Length of the water column (cm) for minimum detection of the disk at the bottom of the tube
Turbidity: Water turbidity in nephelometric turbidity units (ntu)
Salinity: Water salinity measured in practical salinity units (PSU)
Temperature: Water temperature in degrees Celsius (°C)
Kd(PAR): The diffuse attenuation coefficient for photosynthetically available radiation (PAR, 400 – 700 nm)
Water Depth: Water depth at the sampling location (m)
TSS: Total suspended solids (mg l-1)
A355: Absorption coefficient for water containing CDOM at 355 nm (m-1)
Chlorophyll a: Phytoplankton concentration in units of chlorophyll a (µg l-1)
Missing data are denoted by “-999”
A total of 22 estuarine locations were sampled 107 times in 2022 - 23, stretching from North Myrtle Beach to Port Royal Sound along the coast of South Carolina (Fig. 2). The goal was to include a range of water types with different turbidity characteristics from tidal regions of coastal waterbodies. Site selection was determined based on accessibility, usually from a boat dock or launch ramp.
Water transparency, photosynthetically available radiation (PAR) depth profiles, salinity, and temperature were obtained on-site. Transparency was measured using a 120 cm “transparency tube” (Eisco Labs, Victor, NY, PN PH1120148120SPR). The transparency tube (TT) was filled with well-mixed water from the sample site using a 1 m bailer and carboy. While looking through the top opening of the tube, water was slowly drained out of the bottom until the Secchi disk at the bottom of the tube was visible, and the depth of the remaining water (in cm) was recorded. The water column length in the TT was recorded as the average of two independent measurements taken by observers. The Secchi depth was determined using a 20 cm black and white Secchi disk and measured in centimeters. Both Secchi and TT measurement methodologies follow the protocol by Ohrel and Register (2007). A Li-Cor LI-250A portable light meter with a LI-COR LI-193 spherical underwater quantum sensor was lowered into the water column to measure photosynthetically available radiation (PAR, 400 – 700 nm) available throughout the water column (LI-COR, Lincoln, NE, USA). PAR values were recorded as 15 s averages at 0.25 m depth intervals, in units of µmol photons m-2 s-1. Measurements continued until the PAR equaled 0, or, if the sample depth was exceptionally shallow, until the bottom of the water column was reached (the maximum depth for the cable was 8 m). The equation used for calculating the diffuse attenuation coefficient for PAR (kd(PAR)) was:
Ez = Eoe-kz
Where z is water depth (m), Ez is the PAR at a depth (z), Eo is the PAR at the surface, and k is the diffuse attenuation coefficient (Kirk 2011). Salinity was determined using a refractometer, and temperature was measured with a digital thermometer. GPS coordinates, tidal stage, and weather conditions were recorded during sampling. The remaining sample water in the carboy was mixed well, poured into an opaque 1 l Nalgene bottle, and stored in an ice cooler for 3-5 hours before laboratory analysis.
Total suspended solids (TSS), turbidity (NTU), chromophoric dissolved organic matter (CDOM), and phytoplankton biomass (as chlorophyll a, chl a) were measured for each sample. Water for TSS measurements was filtered through pre-weighed filters (GF/F, 47 mm, 0.7 µm nominal particle retention), dried at 45 °C for 24 h, and reweighed to determine the dry weight of suspended solids. Water for chl a was vacuum filtered onto 25 mm filters (GF/F) and stored at -80 ˚C until analysis. For CDOM measurements, 50 ml of water was syringe-filtered through a dried 25 mm GF/F filter and stored at 4 ˚C for no longer than one month. A calibrated Hach 2100Q turbidimeter (PN: 2100Q01) was used to measure turbidity in NTU, following the U.S. EPA Method 180.1 (US EPA 1993). An average of at least three independent readings was recorded. A Jasco V-730 UV-visible spectrophotometer was used to measure the absorbance of CDOM at a wavelength of 355 nm. The absorbance at this wavelength represents CDOM concentration in coastal environments (Blough and Del Vecchio 2002; Kirk 2011). CDOM raw absorbance data were converted into absorption coefficients (a) using the equation (Hu et al. 2002):
a(355) = 2.303 *A(355)/0.05
Where A(l) is the raw wavelength-dependent absorbance, and 0.05 is the optical path length of the cuvette in meters. High-performance liquid chromatography (HPLC) was used to determine chl a.. Samples were lyophilized for 24 h at -50 °C, placed in 90% acetone (0.750 ml), sonicated, and extracted at -20 °C for 18 - 20 h. Filtered extracts (250 µl) were injected into a Shimadzu 2050 HPLC equipped with a monomeric (Rainin Microsorb-MV, 0.46 x 10 cm, 3 µm) and a polymeric (Vydac 201TP54, 0.46 x 25 cm, 5 µm) reverse-phase C18 column in series. A nonlinear binary gradient consisting of the solvents 80% methanol: 20% 0.50 M ammonium acetate and 80% methanol:20% acetone was used for pigment separations (Pinckney et al. 2001a).
Statistics
All data were log10 transformed before analysis. IBM SPSS Statistics (v. 29.0) was used for the correlations and linear regressions, and slope comparisons were made using OriginPro 2025.