Santee Experimental Forest, Watershed 79: streamflow and water chemistry data

Metadata:

Data_Quality_Information:

Attribute_Accuracy:

Attribute_Accuracy_Report:

HISTORIC DAILY FLOW:

Stage was measured at 15-minute intervals using an analog-to-digital water level recording (ADR) device accurate to the nearest 0.01 foot (ft). Submerged flow (if any) was measured using an FW-1 recorder. Daily mean flow values were calculated by first translating the ADR tape measurements to obtain stage values, which were used with stage-discharge relationship equations for the weir and box culverts on either side in a Fortran program to compute the corresponding 15-minute flow rates. These 15-min flow rates for each day were then averaged to obtain the daily mean flow rates. A staff gauge was mounted on the face of the blockhouse to allow direct comparison of measured stage data to manual readings (as a quality assurance / quality control [qaqc] check).


DAILY FLOW/HOURLY AND 10-MIN FLOW:

A Global Water GL300 logger linked to a WL300 pressure transducer with an accuracy of 0.01 ft was installed in the WS79 blockhouse on 1/24/2002. On 7/17/2003 a Teledyne-ISCO 3210 flowmeter was added. The ISCO 3210 had an ultrasonic sensor that could be calibrated to the level of water above or below the v-notch in the weir and also measured stage with an accuracy of 0.01 ft.

Stage data above and below the v-notch for the period between 1/24/2002 and 7/17/2003 were predicted using the following regression equation:

ISCO 3210 = 1.00*GL300 - 1.66, R2 = 0.99

Stage data for this period were hourly, but starting on 7/17/2003 the ISCO 3210 flowmeter recorded data every 10 minutes. The above regression equation was developed in November 2016 using ISCO 3210 and GL300 data collected at the same timepoints between 8/26/2003 and 9/12/2003 (regressing ISCO 3210 data on the GL300 data), and it replaced an older, invalid equation. The stage and flow values for the period have been recalculated and replaced in the data portal.

A staff gauge was mounted on the face of the blockhouse to allow direct comparison of measured stage data to manual readings (as a qaqc check).

A WL15 logger/pressure transducer was installed on 2/4/2004 as a backup to the ISCO 3210 flowmeter (replacing the GL300/WL300 unit).

Weir plate replacement work was initiated on 10/12/2007 at about 9:30 AM (stage level was lowered dramatically as the plate was removed). This work continued into January 2008, as the first efforts resulted in leakage around the new plate. Stage data during and shortly after the replacement work were therefore likely underestimates.

Stage values from 9:40 to 12:20 on 10/25/2007 were predicted using a regression relationship developed between the ISCO 3210 flowmeter data and the WL15 logger/pressure transducer data:

ISCO 3210 = 0.98*WL15 - 1.63, R2 = 0.99

ISCO 3210 data collected from 10/25/2007 12:30 to 10/31/2007 23:50 were regressed on WL15 data collected from the same period to generate this equation (stage data from 10/25/2007 12:30 to 10/31/2007 10:20 were first adjusted to reflect the new calibration of the v-notch by adding 0.055 ft to the measured stage heights).

During the download process on 12/19/2007, the ISCO 3210 flowmeter date was somehow set to the year 2052, and data between then and 1/8/2008 could not be retrieved by normal procedures.

The ISCO 3210 flowmeter was replaced by a Teledyne-ISCO 4210 flowmeter on 2/11/2008.

The WL15 logger/pressure transducer was replaced by a similar WL16 unit on 9/3/2008.

Because the ultrasonic sensor had been submerged, ISCO 4210 stage data from 10/24/2008 19:10 to 10/25/2008 8:50 (during an extremely large flow event caused by a tropical storm) were predicted using data from the new WL16 backup logger/pressure transducer and the following regression equation:

ISCO 4210 = 0.99*WL16 - 1.60, R2 = 0.99

This equation was derived by regressing ISCO 4210 stage data on WL16 stage data for the remainder of the month of October 2008.

Moderate-sized log removed from weir around 10:00 on 12/31/2013.

ISCO 4210 stage data missing from 12/7/2014 6:00 to 12/8/2014 11:10 because of a power outage were predicted using data from the WL16 backup logger/pressure transducer and the following regression equation:

ISCO 4210 = 0.99*WL16 - 1.44, R2 = 0.98

This equation was derived by regressing ISCO 4210 stage data on WL16 stage data for the first week of the month of December 2014.

Nearby lightning strike resulted in loss of all data between 7/6/2015 9:50 and 7/14/2015 8:50.

Extreme rainfall event October 3-4, 2015 (as an indirect effect of Hurricane Joaquin) and also large amounts of rain in the days before and after) led to record flow levels on the Santee Experimental Forest. The ultrasonic sensor was briefly submerged during this event, but data from a backup WL16 logger and pressure transducer was available to help predict the missing stage values. Data from 10/1/2015 0:00 through 10/2/2015 22:00 and then 10/6/2015 0:00 through 10/9/2015 22:00 was used to generate the following regression equation:

ISCO 4210 = 1.00*WL16 - 1.06, R2 = 0.99

Stage values around the peak of the highest flows exceeded the range of the stage-discharge equation for the WS79 weir (0 to 5.14 ft [ft] above the v-notch). See Amatya et al. (2016) for peak discharge approximations.

The ultrasonic sensor was briefly submerged from 8/10/2016 13:40 to 8/10/2016 19:00 during a large flow event on the SEF as a result of Hurricane Matthew. Data from a backup WL16 logger and pressure transducer was available to help predict the missing stage values. Data from both instruments over the remainder of the month was used to generate the following regression equation:

ISCO 4210 = 0.99*WL16 - 1.04, R2 = 0.99

Hurricane Matthew October 7-8, 2016 caused very high flow levels on the Santee Experimental Forest. The ultrasonic sensor was briefly submerged during this event, but data from a backup WL16 logger and pressure transducer was available to help predict the missing stage values. Data from 10/7/2016 6:30 to 10/8/2016 0:50 was used to generate the following regression equation for use in estimating stage values during the rising limb of the event:

ISCO 4210 = 0.97*WL16 - 1.02, R2 = 0.99

Data from 10/9/2016 4:30 to 10/11/2016 9:00 was used to generate the following regression equation for use in estimating stage values during the falling limb:

ISCO 4210 = 1.01*WL16 - 1.08, R2 = 0.99

Stage values around the peak of the highest flows again exceeded the range of the stage-discharge equation for the WS79 weir (0 to 5.14 ft above the v-notch).

There were several periods during April, July and September 2017 when WS79 4210 stage values were again predicted using regression equations developed between the 4210 ultrasonic sensor and the backup WL16 pressure transducer (because water levels rose during large rain events - e.g., Tropical Storm Irma in September - into the 4210 "dead zone").

In April 2017, rising limb values from 4/24/2017 20:30 to 4/24/2017 21:50 were predicted using the following equation:

ISCO 4210 = 0.99*WL16 - 1.06, R2 = 0.99

Falling limb values from 4/24/2017 22:00 to 4/25/2017 0:00 were predicted using the following equation:

ISCO 4210 = 1.00*WL16 - 1.08, R2 = 0.99

In July 2017, rising limb values from 7/10/2017 6:50 to 7/10/2017 7:30 were predicted using the following equation:

ISCO 4210 = 0.99*WL16 - 1.01, R2 = 0.99

Falling limb values from 7/10/2017 7:40 to 7/10/2017 8:20 were predicted using the following equation:

ISCO 4210 = 1.00*WL16 - 1.05, R2 = 0.99

We also used linear interpolation to smooth the peak at the 7/10/2017 7:40 and 7:50 timepoints.

In September 2017 (during Tropical Storm Irma), rising limb values from 9/11/2017 18:30 to 9/11/2017 22:00 were predicted using the following equation:

ISCO 4210 = 1.01*WL16 - 1.05, R2 = 0.99

Falling limb values from 9/11/2017 22:10 to 9/12/2017 5:40 were predicted using the following equation:

ISCO 4210 = 1.00*WL16 -1.04, R2 = 0.99

We also used linear interpolation to smooth the peak at the 9/11/2017 22:10 and 22:20 timepoints.

A moderately large log was removed from the weir on 2/20/2018, and the affected flow data between 2/20/2018 0:30 to 2/21/2018 14:00 were omitted.

Stage measurements were affected by a lengthy period of beaver activity from April to October 2018. The beaver was finally captured on 10/21/18, and selected flow data have been omitted from 4/2/2018 to 10/21/2018.

During a large storm event on 12/14/2018 the ultrasonic sensor was briefly submerged from 12:30 to 19:10. We predicted the stage data between these timepoints by using stage data from the WL16 backup pressure transducer to generate the following regression equation:

ISCO 4210 = 1.01*WL16 -1.09, R2 = 0.99

Spike in flow on 1/14/2019 was caused by release of debris dam (not beaver-related) at WS80 weir contributing to this watershed. Used linear interpolation to correct and replace values between 1/14/2019 12:30 and 1/15/2019 18:20.

Small leak detected at base of weir plate (left side as looking upstream) on 7/1/2019. Therefore, all flow estimates from this point onward until and shortly after repair work was completed (8/15/2019) are underestimates.

Weir plate repair conducted 8/14/2019 through 8/15/2019 by Joey Walters and son Joey (Banter Construction). This involved lowering the pool level, removing the plate and drilling new holes/filling old ones, cutting off old, worn bolts and replacing them with more, smaller stainless-steel bolts and nuts, then fitting new rubber gasket and sealing with marine caulk and hydraulic cement.

Ultrasonic sensor submerged during flow event starting 9/5/2019 during Hurricane Dorian (submerged from 9/5/2019 10:00 to 9/6/2019 1:40). Used combination of linear regression (4210 stage on WL16 stage) and approximated curve smoothing to estimate stage during the period of submergence.

Debris dam removed on 10/28/2019 required stage corrections (10/27/2019 through 10/28/2019) using linear interpolation.

Ultrasonic sensor submerged during flow event starting 12/23/2019 during very large rain event (submerged from 12/23/2019 16:20 to 12/24/2019 3:10). Used combination of linear regression (4210 stage on WL16 stage) and approximated curve smoothing to estimate stage during the period of submergence:

Rising limb: ISCO 4210 = 0.97*WL16 - 1.01, R2 = 0.99
Falling limb: ISCO 4210 = 1.01*WL16 - 1.08, R2 = 0.99

Large log removed from WS80 weir on 1/21/2020 required WS79 stage correction using linear interpolation because of subsequent small surge in level.

Sampler placed in standby mode on 3/23/2020.

Water level rose into the sensor "dead zone" during the large event that started on 8/5/2020. Used combination of linear regression (4210 stage on WL16 stage) and approximated curve smoothing to estimate missing stage data:

Rising limb: ISCO 4210 = 0.99*WL16 - 1.04, R2 = 0.99
Falling limb: ISCO 4210 = 0.99*WL16 - 1.01, R2 = 0.99

Water level rose into the sensor "dead zone" during the large event that started on 8/24/2020. Used combination of linear regression (4210 stage on WL16 stage) and approximated curve smoothing to estimate missing stage data:

Rising limb: ISCO 4210 = 0.99*WL16 -1.01, R2 = 0.99
Falling limb: ISCO 4210 = 0.99*WL16 - 1.00, R2 = 0.99

Removal of debris dam on 11/30/2020 may have led to a small stage rise for 2-3 hours.

Unusual drop in stage observed 3/28/2021; no sign of beaver activity detected.
Linear interpolation was used to correct data between 3/28/2021 6:10 and 3/28/2021 23:30.

Unauthorized pumping apparatus observed at weir pool on 6/30/2021, and while CAH was present employees of RW Harris, Inc. arrived to pump more water for use in construction; after being asked to remove the equipment and cease removing water from SEF streams they complied. Linear interpolation was used to correct data between 6/28/2021 7:10 and 6/29/2021 0:30.

Large log removed from weir on 7/27/2021, and this required stage correction by linear interpolation between 7/26/2021 14:30 and 7/27/2021 18:30.

Medium sized log removed from weir on 8/7/2023; corrected resulting brief dip in stage data by using linear interpolation between 8/6/2023 21:50 and 8/7/2023 19:40.


HISTORIC WATER QUALITY:

Laboratory analyses were performed in the USDA Forest Service analytical laboratory at its facility in Charleston, SC. These analyses were carried out according to protocols specified in the following volume:

Standard Methods for the Examination of Water and Wastewater. Ed. Clesceri, L.S.; Greenberg, A.E.; Eaton, A.D. 1989. American Public Health Association. Washington, DC.


WATER QUALITY STARTING IN 2006:

Laboratory analyses from September 2006 through February 2007 followed standards described in SEF_Laboratory_Manual_01-09-14.pdf, a summary of sample collection, handling, transporting, processing, preservation and laboratory analysis protocols compiled by Ms. Lara Matthews, Laboratory Manager at the Center for Forested Wetlands Research in October 2006. Analytical instruments were calibrated with known standards before every sample run, and laboratory analyses of sample sets always included at least 10% duplicate, blank and spiked samples for quality assurance.

The analytical laboratory at the Center for Forested Wetlands Research in Charleston was shut down in the spring of 2007 (all personnel were in the process of moving out to the Santee Experimental Forest offices and labs in Cordesville), and after March 2007 water quality sampling was discontinued temporarily and did not resume until December 2007. Instrumental analyses for samples collected from March 2007 forward were conducted at the USDA Forest Service Coweeta Hydrologic Laboratory analytical laboratory and followed standards described in "Procedures for Chemical Analysis," (wetlab-cookbook_revised-2017-08-08.pdf) a laboratory manual compiled by Mr. James M. Deal and revised by Ms. Cindi Brown and others that summarizes protocols employed at the Coweeta Hydrologic Laboratory, 3160 Coweeta Lab Road, Otto NC 28763.

Calibration and field use of the Manta multiprobe was performed as recommended in the "Manta Water Quality Multiprobe Startup Guide," version dated June 4, 2004 (Eureka Environmental Engineering, 2113 Wells Branch Parkway Suite 4400, Austin TX 78728).

Problems with the pH sensor in the Manta multiprobe started to develop in March 2016, and we switched to measuring pH in the field with the Oakton PC 300 Waterproof Hand-held pH/Conductivity/TDS/Temperature Meter on October 28, 2016. Calibration and field use of the PC 300 Meter were performed as recommended in the Instruction Manual for the PC 300, rev. March 5, 2008 (Oakton Instruments, P.O. Box 5136, Vernon Hills IL 60061).

In September 2017 we started using the Hanna HI98194 multiprobe for measurement of pH, temperature, specific conductance, salinity and dissolved oxygen. Calibration and field use of the HI98194 multiprobe were performed as recommended in the Instruction Manual for the HI98194, HI98195 and HI98196 Multiparameter Meters (Hanna Instruments Inc., Highland Industrial Park, 584 Park East Drive, Woonsocket RI 02895).

During the COVID19 pandemic, the automatic sampler was turned off in May 2020 (because our laboratory assistant was resigning) and we started taking weekly grab samples during periods of active flow. Collection of grab samples ceased in July 2021 (and the automatic sampler remained off).

The Hanna multiprobe broke during early December 2018. After replacing the unit, bi-weekly Hanna multiprobe measurements resumed in June 2020 and continued through December 2020, but they ceased during the SRS COVID lockdown in January 2021 and did not restart until October 2021. However, the new instrument broke in January 2022. The Hanna was replaced by a YSI ProQuatro multiprobe, and measurements began again in April 2023.


The accuracy and resolution of the parameters measured by the Manta are listed below:

pH: accuracy = +/- 0.2 units; resolution = 0.01 units;
temperature: accuracy (over the range of -5 to 50 degrees Celsius [degrees C]) = +/- 0.08 degrees C; resolution = 0.01 degrees C;
specific conductance: accuracy (over the ranges of 0 - 5, 5 - 25 and 25 - 112 microS per cm) = +/- 1% reading +/- 0.001, +/- 1% reading, and +/- 1% reading, respectively; resolution = to 4 digits; and
salinity: accuracy (over the range 0 - 70 PSS) = +/- 1% reading +/- 1 count; resolution = 0.01 PSS;
and dissolved oxygen: accuracy (over the range 0 - 50 milligrams per liter [mg per L]) = +/- 0.2 mg per L (for less than or equal to 20 mg per L) and +/- 0.6 mg per L (for greater than 20 mg per L).


The accuracy and resolution of pH measurements made using the PC 300 Meter are listed below:

pH: accuracy = +/- 0.01 units; resolution = 0.01 units.


The accuracy and resolution of the parameters measured by the Hanna are listed below:

pH: accuracy = +/- 0.02 units; resolution = 0.01 units;
temperature: accuracy = +/- 0.15 degrees C; resolution = 0.01 degrees C;
specific conductance: accuracy = +/- 1% of reading; resolution = 0.001 mS/cm from 0.000 to 9.999 mS/cm;
salinity: accuracy = +/- 2% of reading; resolution = 0.01 PSU; and
dissolved oxygen: accuracy (over the range 0.00 to 30.00 mg/L) = +/- 1.5% of the reading or +/- 0.10 mg/L, whichever is greater (w.i.g.); resolution = 0.01 mg/L.


The accuracy and resolution of the parameters measured by the ProQuatro are listed below:

pH: accuracy = +/- 0.2 units; resolution = 0.01 units;
temperature: accuracy = 0.2 degrees C; resolution = 0.1 degrees C;
specific conductance: accuracy = 0.5% of reading or 0.100 mS/cm, w.i.g.; resolution = range dependent.
salinity: accuracy = +/- 1% of reading or 0.1 ppt, w.i.g.; resolution = 0.01 ppt;
dissolved oxygen: accuracy (over the range 0 to 20 mg/L) = +/- 2% of the reading or 0.2 mg/L, w.i.g.

Logical_Consistency_Report:

HISTORIC DAILY FLOW:

Visual inspection and comparison with flow data from other SEF flow gauging stations (as well as daily rainfall data) were used to expose anomalous data.


DAILY FLOW/HOURLY AND 10-MIN FLOW:

Visual inspection and comparison with flow data from other SEF flow gauging stations (as well as daily rainfall data from its nested watersheds WS77 and WS80) were used to expose anomalous data.


HISTORIC WATER QUALITY:

Laboratory analyses were performed in the USDA Forest Service analytical laboratory at its facility in Charleston, SC. These analyses were carried out according to protocols specified in the following volume:

Standard Methods for the Examination of Water and Wastewater. Ed. Clesceri, L.S.; Greenberg, A.E.; Eaton, A.D. 1989. American Public Health Association. Washington, DC.


WATER QUALITY STARTING IN 2006:

Visual inspection of electronically plotted water quality data and calculation of concentration means and standard deviations were used to identify possible outliers. Any values outside 5 standard deviations of the mean were considered outliers and omitted from the data set (but noted in the metadata).

Samples collected on 10/3/2016, as well as frozen filtered samples collected and processed on or after 9/16/2016, were exposed to room temperature sometime after 10/5/2016 due to power outage resulting from Hurricane Matthew.

Samples collected on 10/11/2016 were stored in an outbuilding to keep cool until 10/11/2016 13:30 after power had been restored.

Because of problems with the ultrapure water system, filters used to process samples collected on or after 12/19/2016 were pre-soaked in DI, not ultrapure, water prior to use.

Completeness_Report:

All missing data are indicated by blank cells.


HISTORIC DAILY FLOW:

Flow monitoring was discontinued starting 11/1/1973 but was resumed on 11/1/1989 following the devastation caused by Hurricane Hugo on 9/21/1989. However, reliable historic data are only available until 10/31/1990 for this site.


DAILY FLOW/HOURLY AND 10-MIN FLOW:

No flow data were collected at the WS79 weir from 11/1/1990 through 1/24/2002. However, South Carolina was experiencing a severe drought from 1999 to 2002 (South Carolina Department of Natural Resources [SC DNR] declared that the drought had ended in Berkeley County in November 2002) and there was little or no flow at the gauging station during the drought. Since then there have been brief periods of missing data, largely as a result of equipment failures.

Data from 12/19/2007 to 1/8/2008 were not available because of equipment malfunction.

Data from 12/2/2013 to 12/9/2013 were not available because of equipment malfunction.

Data from 9/28/2014 to 9/29/2014 were not available because of equipment malfunction.

Data from 7/6/2015 to 7/14/2015 were not available because of nearby lightning strike.

Some data on both 10/3/2015 and 10/4/2015 are not available because stage values during massive flow event exceeded the ceiling of the box culverts (greater than 5.14 ft above the v-notch), resulting in submerged flow.

Data on 10/8/2016 were not available because some stage values during Hurricane Matthew exceeded the ceiling of the box culverts (greater than 5.14 ft above the v-notch), resulting in submerged flow.

A moderately large log was removed from the weir on 2/20/2018, and the affected flow data between 2/20/2018 0:30 to 2/21/2018 14:00 were omitted.

Stage measurements were affected by a lengthy period of beaver activity from April to October 2018. The beaver was finally captured on 10/21/2018, and selected flow data have been omitted from 4/2/2018 to 10/21/2018.


HISTORIC WATER QUALITY:

Approximately weekly water samples were collected upstream of the WS79 weir between 10/13/1989 and 12/29/1994.


WATER QUALITY STARTING IN 2006:

Stream water samples were collected as weekly grab samples and (starting August 30, 2010) by a Teledyne-ISCO 3700 automated sampler during periods of active flow on a flow-proportional basis; during periods without flow no samples were collected. While the Center for Forested Wetlands Research was in the process of relocating its offices and labs from Charleston to the Santee Experimental Forest in Cordesville, collection of water quality samples at the WS79 gauging station was discontinued temporarily beginning in April 2007 and did not resume until March 2008.

During the COVID19 pandemic, the automatic sampler was turned off in May 2020 (because our laboratory assistant was resigning) and we started taking weekly grab samples during periods of active flow. Collection of grab samples was discontinued in July 2021 (and the automatic sampler remained off).


Missing Data (2006 - 2007):

None of the potential reasons for a number of possible outliers can be confirmed, and so these values have been retained in the data set and it is at the discretion of the user to exclude them.

Manta data missing for following periods:
1/30/2007 (DO values removed; at or above 100%)

All pH, conductivity and salinity data for 2006 and 2007 removed because of questions about instrument calibration during the period.

No analyses were conducted for SiO2 from 2006 - 2007.


Outliers and Missing Data (2008 - 2023):

There are a number of possible outliers, including the following dates and analytes:

08/27/2011 11:57 - TDN (2.802), NH4 (1.466), K (6.11), DOC (70.67), SO4 (16.326), PO4 (0.631)
06/12/2012 01:57 - TDN (1.775)
08/22/2013 23:23 - TDN (2.654), NH4 (1.111), NO3_NO2 (0.284), K (4.16), PO4 (0.639)
09/22/2014 20:01 - NO3_NO2 (0.195), K (2.76), PO4 (0.949)
01/25/2015 18:03 - Cl (75.136)
07/20/2015 04:09 - NH4 (0.69)
08/06/2015 17:18 - NH4 (0.793)
06/08/2016 18:47 - SO4 (18.356)
07/21/2018 09:30 - TDN (2.873), NH4 (2.194), PO4 (0.993)
08/14/2018 19:46 - Cl (40.577)
08/21/2018 19:35 - TDP (0.631), K (8.67)
12/02/2018 04:41 - Cl (47.293)
12/04/2018 19:29 - Cl (50.377)
12/18/2018 11:43 - Cl (41.073)
02/01/2019 06:49 - Ca (25.87)
06/10/2019 05:37 - SO4 (16.668)
03/02/2021 11:07 - Ca (21.24), Na (12.38)
03/29/2021 10:50 - Ca (19.22)
04/12/2021 10:50 - Ca (21.47), Na (11.21)
06/30/2021 10:36 - NO3_NO2 (0.103)

Concentration values outside five standard deviations of the mean were considered possible outliers.

Because none of the potential reasons for the extreme values can be confirmed, these possible outliers have been retained in the data set and it is at the discretion of the user to exclude them.

DOC data from samples collected during the June 11, 2012 to April 16, 2013 period have been removed as suspect because of concerns raised at the Coweeta Hydrologic Laboratory (their instrument may have been measuring Total Carbon instead of Dissolved Organic Carbon during this period).

Manta data missing for following periods:

All pH, conductivity and salinity data for 2008 removed because of questions about instrument calibration during the period.

8/16/2013 (specific conductivity and salinity values removed as outliers).

Manta data not available from November 2015 through early February 2016 because of problem with cable, and also not available from early March 2016 through early June 2016 because of sensor problems.

Suspect pH data for 3/4/2016 removed.

Only pH data available for 10/28/2016.

We only conducted analyses for SiO2 briefly in 2012.

The Hanna multiprobe broke during early December 2018, and there were no new data until the unit was replaced. Bi-weekly Hanna multiprobe measurements resumed in June 2020 and continued through December 2020, but ceased during the SRS COVID lockdown in January 2021 and did not restart until October 2021. However, the new instrument broke in January 2022. The Hanna was replaced by a YSI ProQuatro multiprobe, and measurements began again in April 2023.

Lineage:

Methodology:

Methodology_Type: Field

Methodology_Description:

HISTORIC DAILY FLOW:

In 1966 the Forest Service constructed a compound v-notch weir with a rectangular box culvert on either side (see Amatya et al. 2022 for their detail dimensions) and associated blockhouse (to house monitoring instruments) at the WS79 outlet on Lotti Road. Stage data were recorded on tapes at 15-minute intervals by analog-to-digital water level recording (ADR) equipment. Data collection ceased on 11/1/1990.

Methodology:

Methodology_Type: Field

Methodology_Description:

DAILY FLOW/HOURLY AND 10-MIN FLOW:

Between 1/24/2002 and 7/17/2003, hourly stage data were recorded by a Global Water GL300 logger (and associated WL300 pressure transducer) and downloaded directly to a computer in the field using Global Logger software.

A Teledyne-ISCO 3210 flowmeter with an ultrasonic sensor was installed on 7/17/2003. 10-minute stage data from the ISCO 3210 flowmeter were downloaded directly to a computer in the field using Flowlink 3 software. Manual measurements of stage level above or below the v-notch in the weir were performed and compared to the instrument reading, and if they differed by 0.01 ft or more the instrument was recalibrated.

A WL15 logger/pressure transducer was installed on 2/4/2004 as a backup to the ISCO 3210 flowmeter (replacing the GL300/WL300 unit). 10-minute stage data from the WL15 logger were downloaded directly to a computer in the field using Global Logger software.

The ISCO 3210 flowmeter was replaced by a Teledyne-ISCO 4210 flowmeter on 2/11/2008. From that point onward 10-minute stage data were downloaded directly to a computer in the field using Flowlink 4 (and later Flowlink 5) software.

The WL15 logger/pressure transducer was replaced by a WL16 unit on 9/3/2008. 10-minute stage data from the WL16 logger were downloaded directly to a computer in the field using Global Logger II software.

Trainum Brothers Co. installed new wooden platforms and safety railing at the site on 6/12/2017 to 6/13/2017.

Weir plate repair conducted 8/14/2019 through 8/15/2019 by Joey Walters and son Joey (Banter Construction). This involved lowering the pool level, removing the plate and drilling new holes/filling old ones, cutting off old, worn bolts and replacing them with more, smaller stainless-steel bolts and nuts, then fitting new rubber gasket and sealing with marine caulk and hydraulic cement.

Methodology:

Methodology_Type: Field

Methodology_Description:

HISTORIC WATER QUALITY:

Between 10/13/1989 and 12/29/1994, approximately weekly water samples were collected upstream of the WS79 weir in clean plastic bottles and returned to the laboratory on ice. At the time of sample collection, stream pH was measured using a hand-held logger and probe.

Methodology:

Methodology_Type: Lab

Methodology_Description:

HISTORIC WATER QUALITY:

Samples were brought to the Charleston laboratory and refrigerated until processing could begin. The samples were filtered (details of filtration procedure are not known at this time) and then subjected to analysis according to standard protocols (Clesceri et al. 1989).

Most analyses (NO3/NO2-N, NH4-N, PO4-P, Cl, SO4, SiO3, TKN and TDN) were performed using the Technicon TRAACS-800 system. Cation analyses (K, Na, Ca, and Mg) were performed using a Perkin-Elmer Model 3100 Atomic Absorption Spectrometer.

Finally, HCO3 was measured by the titration method using an automatic titrimeter.

Methodology_Citation:

Citation_Information:

Originator: Clesceri, L.S. (ed.)

Originator: Greenberg, A.E. (ed.)

Originator: Eaton, A.D. (ed.)

Publication_Date: 1989

Title:

Standard Methods for the Examination of Water and Wastewater

Edition: 17

Geospatial_Data_Presentation_Form: document

Publication_Information:

Publication_Place: Washington, DC

Publisher: American Public Health Association

Methodology:

Methodology_Type: Field

Methodology_Description:

WATER QUALITY STARTING IN 2006:

From September 2006 to April 2007, weekly grab samples were collected upstream of the WS79 weir. The sample bottles were taken to the laboratory in coolers and either filtered immediately (through 0.45 micron, 47 millimeter [mm] WCN type filters) or frozen until filtration could begin.

A Teledyne-ISCO 3700 automated sampler was installed in the blockhouse at the gauging station on August 30, 2010. During periods of active streamflow a Teledyne-ISCO 4210 flow logger/ultrasonic sensor, calibrated to the level of water above or below the v-notch in the WS79 weir, signaled the automated sampler to take a 200-milliliter (ml) water sample after a known amount of water had passed over the weir (1802 cubic meters; amount based on the median volume of historic flow events). Such samples were collected using 3/8 inch vinyl suction tubing attached to the sampler and extending to the center of the stream, where the intake strainer was installed about 1 ft below the bottom of the v-notch in the weir. Samples were composited (4 samples per 1000 ml bottle), and approximately once a week filled bottles were removed from the autosampler and replaced with clean, acid-washed bottles. The sample bottles were taken to the laboratory in coolers and either filtered immediately (through 0.45 micron, 47 mm WCN type filters) or stored at 4 degrees C and filtered as soon as possible.

About once a week the Manta multiprobe (linked to the Amphibian data logger) was taken to the WS79 gauging station and used in-situ to measure pH, temperature, specific conductance, salinity and dissolved oxygen in the pool behind the weir plate (if enough water was present to submerge the probe). In 2009 the Manta measurement interval became approximately bi-weekly.

Manta measurements were discontinued from November 2015 through early February 2016 because of problems with the cable, and then again from early March 2016 through early June 2016 because of sensor problems.

Beginning on October 28, 2016 measurements of pH were performed in the field using an Oakton PC 300 Waterproof Hand-held pH/Conductivity/TDS/Temperature Meter.

We replaced the hard shell bottles in the Teledyne ISCO 3700 automated water quality sampler with Teledyne ISCO ProPak frames and EPA-approved LDPE plastic inserts on August 30, 2017.

In September 2017 we started using the Hanna HI98194 multiprobe for measurement of pH, temperature, specific conductance, salinity and dissolved oxygen.

The Hanna multiprobe broke during early December 2018. After replacing the unit, bi-weekly Hanna multiprobe measurements resumed in June 2020 and continued through December 2020, but they ceased during the SRS COVID lockdown in January 2021 and did not restart until October 2021. However, the new instrument broke in January 2022. The Hanna was replaced by a YSI ProQuatro multiprobe, and measurements began again in April 2023.

During the COVID19 pandemic, the automatic sampler was turned off in May 2020 (because our laboratory assistant was resigning) and we started taking weekly grab samples during periods of active flow. Collection of grab samples was discontinued in July 2021 (and the automatic sampler remained off).

Methodology:

Methodology_Type: Lab

Methodology_Description:

WATER QUALITY STARTING IN 2006:

Laboratory methods from September 2006 to February 2007 were as described in SEF_Laboratory_Manual_01-09-14.pdf, a summary of sample collection, handling, transporting, processing, preservation and laboratory analysis protocols compiled by Ms. Lara Matthews, Laboratory Manager at the Center for Forested Wetlands Research in October 2006.

Briefly, the laboratory protocols involved the following:

After filtration through 0.45 micron, 47 mm WCN type filters, each sample was aliquoted to three labelled 125 ml HDPE sample bottles and one labelled 40 ml amber VOA vial for preservation and subsequent analyses. 140 microliters of a concentrated acid was added to each of three aliquots: H2SO4 in the 125 ml bottle for NH4, NO3/NO2, total dissolved N and total dissolved P analyses; HNO3 in the 125 ml bottle for cation analyses; and H3PO4 in the 40 ml vial for DOC analysis. The remaining 125 ml bottle (for Cl analysis) received no preservative. All sample aliquots were refrigerated until analyses could be performed.

The analyses, instruments used and method detection limits (mdl) during this period were the following:

TDN: performed using QuikChem© Method 10-107-04-3, In-Line Digestion followed by FIA on a LachatQuickChem FIA+ Autoanalyzer; mdl = 0.5 mg per L before March 2004, 0.3 mg per L (March 2004 to July 2005), and 0.1 mg per L after July 2005;

TDP: performed using QuikChem© Method 10-115-01-3, FIA Colorimetry (In-Line Persulfate Digestion Method) on a LachatQuickChem FIA+ Autoanalyzer; mdl = 0.1 mg per L before July 2005 and 0.01 mg per L after July 2005;

NO3/NO2-N: performed using QuikChem© Method 10-107-04-1, Flow Injection Analysis (FIA) on a LachatQuickChem FIA+ Autoanalyzer; mdl = 0.2 mg per L before July 2005 and 0.02 mg per L after July 2005;

NH4-N: performed using QuikChem© Method 10-107-06-1, FIA Colorimetry on a LachatQuickChem FIA+ Autoanalyzer; mdl = 0.1 mg per L before July 2005 and 0.02 mg per L after July 2005;

Cl: performed using QuikChem® Method 10-117-07-1-C, FIA Colorimetry on a LachatQuickChem FIA+ Autoanalyzer; mdl = 0.1 mg per L;

Ca, K, Mg, Na, P: performed using a Jobin Yvon Ultima II Inductively Coupled Plasma Spectrometer; mdl = 0.05 mg/L; and

DOC: performed using a Tekmar Phoenix 8000 Total Organic Carbon Analyzer; mdl = 0.1 mg per L.

Starting in March 2007 instrumental methods were as described in "Procedures for Chemical Analysis," (wetlab-cookbook_revised-2017-08-08.pdf) a laboratory manual compiled by Mr. James M. Deal and revised by Ms. Cindi Brown and others that summarizes protocols employed at the Coweeta Hydrologic Laboratory, 3160 Coweeta Lab Road, Otto NC 28763.

As mentioned in the section on field methodology, starting in August 2010 samples collected in the field from the newly-installed ISCO 3700 autosampler were no longer frozen before processing but instead were refrigerated and filtered (using 0.45 micron, 47 mm WCN type filters) as soon as possible. The filtrate was aliquoted to three labeled plastic 50-ml tubes and all were frozen prior to overnight shipment to the Coweeta Hydrologic Laboratory for analysis.

In June 2016 a new processing method for DOC samples and blanks was adopted: beginning on 6/20/2016, sample tubes designated for DOC analysis were pre-rinsed in 10% HCL followed by a triple-rinse in DI water.

Samples collected on 10/3/16, as well as frozen filtered samples collected and processed on or after 9/16/2016, were exposed to room temperature sometime after 10/5/2016 due to power outage resulting from Hurricane Matthew. These samples were moved to an outbuilding on 10/11/2016 to keep cool until 10/12/2016 13:30, after power had been restored.

Samples collected on 10/11/2016 were stored in an outbuilding to keep cool until 10/12/2016 13:30, after power had been restored.

Because of problems with the ultrapure water system, filters used to process samples collected on or after 12/19/2016 were pre-soaked in DI, not ultrapure, water prior to use.

The analyses, instruments used and average method detection limits (mdl) during the March 2007 through August 2021 period (no automated or grab samples were collected after July 2021) were the following:

TDN: determined by luminescence using a Shimadzu DOC-VCPH TN analyzer until November 2016, and then by luminescence using a Shimadzu DOC-VCPH TNM-1 analyzer starting in November 2016; average mdl = 0.015 mg per L (range = 0.006-0.038 mg per L) through November 2016; average mdl = 0.014 mg per L for December 2016; and average mdl = 0.013 mg per L for 2017 through August 2021 (range = 0.007-0.029 mg per L).

TDP: performed using persulfate in line UV digestion on a LachatQuickChem FIA+ Autoanalyzer until April 2011, when the Lachat results became unreliable, and then determined by optical emission using a Jobin Yvon Ultima II Inductively Coupled Plasma Spectrometer (first acidifying the sample with 0.3% HNO3) through November 2012, and by optical emission using a Thermo Fisher iCAP 6300 starting in December 2012; average mdl = 0.003 mg per L (range = 0.001-0.005 mg per L) through April 2011, average mdl = 0.024 mg per L for the remainder of 2011 through 2012 (range = 0.002-0.046 mg per L), and average mdl = 0.005 mg per L from 2013 to August 2021 (range = 0.001-0.017 mg per L);

NO3/NO2-N: determined by Micro-membrane Suppressed Ion Chromatography, using an AS 18 column, on a Dionex Ion Chromatograph ICS2500; average mdl = 0.002 mg per L (range = 0.0002-0.006 mg per L);

NH4-N: performed using the automated Phenate method on an Astoria 2 Autoanalyzer through November 2015, and using the automated Phenate method on a new Astoria 2 Autoanalyzer from December 2015 through 2021; average mdl = 0.004 mg per L (range = 0.001-0.009 mg per L) through November 2015, and average mdl = 0.001 mg per L from December 2015 through August 2021 (range = 0.001 to 0.002 mg per L);

Cl: determined by Micro-membrane Suppressed Ion Chromatography, using an AS 18 column, on a Dionex Ion Chromatograph ICS2500; average mdl = 0.013 mg per L (range = 0.001-0.065 mg per L);

Ca: determined by optical emission using a Jobin Yvon Ultima II Inductively Coupled Plasma Spectrometer through November 2012, and by optical emission using a Thermo Fisher iCAP 6300 starting in December 2012; average mdl = 0.043 mg per L (range = 0.017-0.095 mg per L) through 2012, and average mdl = 0.019 mg per L from 2013 to August 2021 (range = 0.005-0.056 mg per L);

K: determined by optical emission using a Jobin Yvon Ultima II Inductively Coupled Plasma Spectrometer through November 2012, and by optical emission using a Thermo Fisher iCAP 6300 starting in December 2012; average mdl = 0.024 mg per L (range = 0.009-0.045 mg per L) through 2012, and average mdl = 0.007 mg per L from 2013 to August 2021 (range = 0.001-0.029 mg per L);

Mg: determined by optical emission using a Jobin Yvon Ultima II Inductively Coupled Plasma Spectrometer through November 2012, and by optical emission using a Thermo Fisher iCAP 6300 starting in December 2012; average mdl = 0.018 mg per L (range = 0.003-0.053 mg per L) through 2012, and average mdl = 0.005 mg per L from 2013 to August 2021 (range = 0.001-0.008 mg per L);

Na: determined by optical emission using a Jobin Yvon Ultima II Inductively Coupled Plasma Spectrometer through November 2012, and by optical emission using a Thermo Fisher iCAP 6300 starting in December 2012; average mdl = 0.028 mg per L (range = 0.005-0.106 mg per L) through 2012, and average mdl = 0.009 mg per L from 2013 to August 2021 (range = 0.001-0.026 mg per L);

DOC: performed using the catalytically-aided platinum 680°C combustion technique for sample oxidation on a Shimadzu DOC-VCPH TN analyzer through October 2016, and by using the catalytically-aided platinum 680°C combustion technique for sample oxidation on a Shimadzu DOC-VCPH TNM-1 analyzer starting in November 2016; average mdl = 0.036 mg per L (range = 0.011-0.096 mg per L) through October 2016, and average mdl = 0.044 mg/L from November 2016 through August 2021 (range = 0.030-0.073 mg per L);

Br: determined by Micro-membrane Suppressed Ion Chromatography, using an AS 18 column, on a Dionex Ion Chromatograph ICS2500; average mdl = 0.003 mg per L (range = 0.001-0.008 mg per L);

SO4: determined by Micro-membrane Suppressed Ion Chromatography, using an AS 18 column, on a Dionex Ion Chromatograph ICS2500; average mdl = 0.009 mg per L (range = 0.001-0.062 mg per L);

PO4: determined by Micro-membrane Suppressed Ion Chromatography, using an AS 18 column, on a Dionex Ion Chromatograph ICS2500; average mdl = 0.007 mg per L (range = 0.002-0.023 mg per L); and

SiO2: determined by Ammonium Molybdate reaction and reduction with Ascorbic Acid, using an Astoria 2 Autoanalyzer; 2012 mdl = 0.002 mg per L.

Process_Step:

Process_Description:

HISTORIC DAILY FLOW:

The ADR tape data were read (picked by hand or translated to ASCII code using a photoelectric reader) and 15-minute stage values were converted to the corresponding flow values using stage-discharge equations for the compound weir/dual box culvert in a Fortran program. Daily mean flow rates were generated by averaging the 15-minute flow rates.

Process_Date: Unknown

Process_Step:

Process_Description:

DAILY FLOW/HOURLY AND 10-MIN FLOW:

Downloaded stage data were first post-processed (removing spurious data, such as download artifacts), and then hourly or 10-minute flow rates were obtained by using the theoretical hydraulic equations for a compound v-notch weir with associated box culverts on each side of the weir. Total daily flow data were obtained by integrating the hourly or 10-minute flow rates over the course of each day.

Specific information about regression equations used to convert stage data to streamflow for different time periods and devices, as well as other specifics can be found in the attribute accuracy section.

Process_Date: Unknown

Process_Step:

Process_Description:

HISTORIC WATER QUALITY:

Data collected in the field or during laboratory analyses were recorded into electronic spreadsheets and checked for missing values and/or outliers.

For the sake of consistency, data reported for SO4 during the 1989-1994 period have been adjusted to reflect the concentration of SO4-S.

Process_Date: Unknown

Process_Step:

Process_Description:

WATER QUALITY STARTING IN 2006:

Data (in the form of concentrations obtained from the laboratory) were entered into Microsoft Excel spreadsheets and checked for missing values and/or outliers.

For the purposes of this database, analyte concentrations registering below the method detection limit (mdl) prior to March 2007 were assigned the value of the respective mdl.

Starting in March 2007, analyte concentrations were posted as received from the Coweeta Hydrologic Laboratory except in the case of obvious contamination or negative values. Occasional negative values (sometimes observed for analytes such as Total Phosphorus or Mg) were assigned the value "0.001" according to the established practice at the Coweeta Hydrologic Laboratory.

Process_Date: Unknown

Entity_and_Attribute_Information:

Overview_Description:

Entity_and_Attribute_Overview:

FILES DESCRIBED BELOW ARE AVAILABLE THROUGH FULL DATA PUBLICATION DOWNLOAD.

Data are available as comma-separated values (CSV) files. Data are also available as tab-delimited files through the online query tool which also provides the ability to calculate summary statistics, but be aware that those calculations are based on user selections and should be carefully interpreted. All missing data are indicated by blank cells.


VARIABLE DESCRIPTION FILE

_variable_descriptions.csv: List and description of the variables provided in each data file.
Filename = name of data file
Variable = name of variable
Units = units (if applicable)
Precision = precision (if applicable)
Description = description of variable


\DATA\FLOW\

WS79_dailyflow_2002_2023.csv: WS79 daily mean flow rate and total daily flow volume data starting in 2002.
Location = watershed (WS79 = Watershed 79)
Instr_ID = gauging station name (Weir79 = water level recorder used in conjunction with WS79 weir)
Date_ = date associated with stage data (date format = mm/dd/yyyy)
Avg_FlowRa = daily mean flow rate in L/s
DailyFlow_ = total daily flow volume in cubic meters (m³)

WS79_historical_data_1966-1990.csv: CSV file containing WS79 daily mean flow rate data from 1966 to 1990 (no data collected November 1973 to October 1989).
Location = watershed name (WS79 = Watershed 79)
Instr_ID = gauging station name (Weir79 = water level recorder used in conjunction with WS79 gauging station)
Date_ = date associated with stage data (format MM/DD/YYYY)
Dailyflow_ = daily mean flow rate in liters per second (L/s)

WS79_hourly_10min_2002_2023.csv: WS79 hourly and 10-minute flow rate data starting in 2002.
Location = watershed name (WS79 = Watershed 79)
Instr_ID = gauging station name (Weir79 = water level recorder used in conjunction with WS79 gauging station)
Date_time = Instantaneous date and time stage data recorded (format mm/dd/yyyy hh/mm)
Date_ = date associated with stage data (format MM/DD/YYYY)
Time24_Flo = nearest minute associated with stage data (0 to 1430)
Flow_liter = flow rate (calculated from stage data), in liters per second (L/s)


\DATA\WATER_QUALITY\

WS79_historical_chemistry_1989-1994.csv: Historic WS79 water quality data from 1989 to 1994.
Location = watershed (WS79 = Watershed 79)
Instr_ID = gauging station (Weir79 = WS79 weir)
Date_ = date associated with data (date format = mm/dd/yyyy)
pH = stream pH
NO3_NO2_N_mgL = nitrogen concentration in the form of nitrate/nitrite in milligrams/liter (mg/L)
NH4_N_mgL = nitrogen concentration in the form of ammonium (mg/L)
PO4_P_mgL = phosphorus concentration in the form of inorganic phosphate (mg/L)
Cl_mgL = chloride concentration (mg/L)
K_mgL = potassium concentration (mg/L)
Na_mgL = sodium concentration (mg/L)
Ca_mgL = calcium concentration (mg/L)
Mg_mgL = magnesium concentration (mg/L)
SO4_S_mgL = sulfur concentration in the form of sulfate (mg/L)
TKN_mgL = total Kjehldal nitrogen concentration (mg/L)
SiO3_mgL = silicate concentration (mg/L)
HCO3_mgL = bicarbonate concentration (mg/L)
TN_mgL = total nitrogen concentration (mg/L)

WS79_water_quality_2006-2023.csv: water quality and physical property data recorded from WS79 gauging station in starting in 2006.
Location = watershed (WS79 = Watershed 79)
Instr_ID = gauging station (Weir79 = WS79 weir)
Date_time = instantaneous date and time sample collected (format = mm/dd/yyyy hh/mm)
Date_ = date associated with data (format = mm/dd/yyyy)
Sample_typ = sample type (grab, automated or Manta/Hanna multiprobe measurement)
TDN_mgL = total dissolved nitrogen concentration in milligrams per liter (mg/L)
TDP_mgL = total dissolved phosphorus concentration in mg/L
NH4_N_mgL = Nitrogen concentration in the form of ammonium in mg/L
NO3_NO2_N_mgL = Nitrogen concentration in the form of nitrate/nitrite in mg/L
Cl_mgL = Chloride concentration in mg/L
Ca_mgL = Calcium concentration in mg/L
K_mgL = Potassium concentration in mg/L
Mg_mgL = Magnesium concentration in mg/L
Na_mgL = Sodium concentration in mg/L
DOC_mgL = Dissolved organic carbon concentration in mg/L
Br_mgL = Bromide concentration in mg/L
SO4_mgL = Sulfate concentration in mg/L
PO4_mgL = Phosphate concentration in mg/L
SiO2_mgL = Silicate concentration in mg/L
Temp_C = Stream temperature in degrees Celsius (C)
pH = Stream pH
Conductivi = Stream conductivity in microSiemens per centimeter (microS/cm)
Salinity_P = Stream salinity in none (Practical Salinity Scale)
DO_mgL = Dissolved oxygen in stream in mg/L
DO_per_sat = Percent saturation dissolved oxygen in stream in %


\SUPPLEMENTS\

SEF_Laboratory_Manual_01-09-14.pdf: Portable Document Format (PDF) file containing "Santee Experimental Forest: Laboratory Methods" edited by Julie Arnold on 01/09/2014.

wetlab-cookbook_revised-2017-08-08.pdf: PDF file containing "Procedures for Chemical Analysis" by Coweeta Hydrologic Laboratory (obtained on 11/01/2019 from https://www.srs.fs.usda.gov/coweeta/tools-and-data/wetlab-cookbook_revised-2017-08-08.pdf).

Entity_and_Attribute_Detail_Citation:

none provided

Metadata_Reference_Information:

Metadata_Date: 20240815

Metadata_Contact:

Contact_Information:

Contact_Organization_Primary:

Contact_Organization: USDA Forest Service, Southern Research Station, Center for Forested Wetlands Research

Contact_Person: Andy Harrison

Contact_Position: Hydrology Technician

Contact_Address:

Address_Type: mailing and physical

Address: 3734 Hwy 402

City: Cordesville

State_or_Province: SC

Postal_Code: 29434

Country: USA

Contact_Voice_Telephone: 843-336-5603

Contact_Electronic_Mail_Address: charles.a.harrison@usda.gov

Contact Instructions: This contact information was current as of original publication date. For current information see Contact Us page on: https://doi.org/10.2737/RDS.

Metadata_Standard_Name: FGDC Biological Data Profile of the Content Standard for Digital Geospatial Metadata

Metadata_Standard_Version: FGDC-STD-001.1-1999