Methods:
Data from multiple water quality sensors were collected at the deep hole of Falling Creek Reservoir in 2018-2021 using water temperature thermistors, dissolved oxygen sensors, pressure sensors, and a YSI EXO2 sonde. Profiles were collected every ten minutes from July 2018 to December 2021. We deployed the thermistors and 5 and 9 m dissolved oxygen sensors in July 2018 and the EXO2 sonde at 1 m depth was deployed in August 2018. The EXO2 sonde was moved from 1.0 m to 1.6 m depth on 2019-05-20 at 12:40:00 (1.6 m is the depth of one of the outtake valves for drinking water treatment). As the EXO2 sonde has been moored near the outtake valve at 1.6 m (assuming full pond), the exact depth of the EXO2 sonde varied slightly over time due to small water level changes in the reservoir. We deployed a pressure transducer at 9 m in May 2020 to help track water level changes.
At the time of deployment in summer 2018, the thermistor string was factory-calibrated and verified against a NIST-traceable thermistor to meet measurement precision. The in situ dissolved oxygen sensors were calibrated using a one point calibration according to the standards and practices outlined in the user manual. The YSI EXO2 Sonde sensors, except for temperature, were user-calibrated as outlined in the user manual and KOR EXO software.
The reservoir has an oxygenation system that has been manipulated for whole ecosystem experiments throughout the deployment of the sensors which results in various oxygen conditions because of this different levels of QAQC are done to the sensors but especially the dissolved oxygen sensors at 5m and 9m. See Additional Info for the oxygenations system schedule.
Maintenance/Issues
All maintenance, calibration times and sensor issues are recorded in the FCR_CAT_MaintenanceLog_2018_2021.txt. The text below explains the issues and how they were dealt with in the data.
Maintenance data gaps
The sensor string was regularly brought up to the surface for routine maintenance, dates and times of which were recorded in the FCR_CAT_MaintenanceLog_2018_2021.txt file. To account for this, we used an automated QAQC script (FCR_catwalk_QAQC_function_2021.R) which indexed the dates and times recorded in the maintenance log and set the data to NA and the Flag_All and other Flag columns to 1. Any data points outside of the immediate time frame that the sensor string was up that do not reflect ambient conditions due to sensor re-equilibration were also set to NA and Flag_All to 1 manually and the corresponding Flag column.
Missing Data Gaps:
Missing Data Gaps EXO:
2018-08-20 14:30-2018-08-27 17:50; 2018-09-25 10:40-2018-10-01 11:00; 2020-12-02 12:50-2020-12-27 11:50-removed for maintenance; 2021-04-26 11:00-13:00-Removed for calibration; 2021-12-13 10:25:00- 2021-12-14 12:25:00; 2021-06-15 11:40-2021-06-18 10:20-Connection issue so no data was streaming; 2021-12-13 10:25:00- 2021-12-14 12:25:00 -Removed for cleaning
Dissolved Oxygen sondes at 5 and 9 m
Maintenance performed on the sensors-
When the oxygenation system is first activated, we have observed a rapid fouling of the 9m Do probe, which we typically flag with a 5 or 6.
2021-
At 9m there were some high oxygen readings on 2021-06-07-2021-06-08 about 3 hours after the temperature string had been pulled up. There was no oxygen at 9m, as the oxygenation system had not been turned on yet, so these readings were changed to NA and flagged.
The oxygenation was first turned on 11 June at 9:40EST but we were having issues with the oxygen regulator so we were unsure how well the system was working until the regulator was fixed and deployed on 14 September at 13:30 EST. During this time we have flagged the data with a flag of 5 because of the fluctuations from the system.
The temperature string was tangled after some maintenance on June 18th, 2021 at 10:00 and was fixed on June 21, 2021 at 10:00. It was tangled again on July 21st, 2021 13:20 and fixed on July 26th, 2021 at 12:00. Consequently, there was a data gap for both DO sensors with a data flag of 1 in the Flag_All column and individual flag columns of the sensors affected. The 9m RDO-PRO_X sensor was calibrated on August 18th, 2021 with a two point calibration using 100 percent air saturation and a 0 percent saturation with a sodium sulfite solution. On August 24, 2021, the anti-fouling caps on the RDO-PRO-X sensors were removed as they were causing fouling especially in anoxic hypolimnion. The RDO-PRO-X sensor cap at 5m was replaced on August 31st, 2021 and checked with a 100 percent air saturation and 0 percent saturation.
2020-On August 7, 2020, the RDO-PRO-X Sensor caps at 5 m and 9 m were replaced. Both dissolved oxygen sensors were calibrated on August 17, 2020 using a one point 100 percent air saturation calibration.
QAQC methods: After any maintenance on the DO sensors, Do measurements for the next two hours are changed to NAs to give ample time for the sensors to equilibrate. There were two periods in the summer of 2019 and four periods in the summer of 2020 when hypolimnetic DO sondes gave low readings that did not align with simultaneous measurements from a high-resolution CTD profiler (likely due to fouling). We adjusted these values using linear or square root functions to best match the CTD and present those data in a separate column (original values are retained and data are flagged for this column with Flag = 6). The equations for each adjustment are given below, as well as in the R code provided with this data product.
5 meter oxygen sonde:
This equation will give you the values in the RDO_mgL_5_adjusted and RDOsat_percent_5_adjusted:
2019-08-11 00:00:00 to 2019-08-12 12:40:00
- DO (mg/L) = DO observed plus the square root of the difference between the current time and 2019-08-11 00:00:00 in minutes, divided by 30
- DO (percent saturation) = DO observed plus the square root of the difference between the current time and 2019-08-11 00:00:00 in minutes, divided by 30/11.3*100, with 11.3/100 being the potential saturation of oxygen in water at 10 degrees C
9 meter oxygen sonde
This equation will give you the values in the RDO_mgL_9_adjusted and RDOsat_percent_9_adjusted:
2019-08-05 12:45:00 to 2019-08-12 12:40:00
- DO (mg/L) = DO observed plus the square root of the difference between the current time and 2019-08-05 12:45:00 in minutes, divided by 70
- DO (percent saturation) = DO observed plus the square root of the difference between the current time and 2019-08-05 12:45:00 in minutes, divided by 70/11.3*100, with 11.3/100 being the potential saturation of oxygen in water at 10 degrees C
2019-09-02 13:00:00 to 2019-09-27 12:30:00
- DO (mg/L) = DO observed plus the difference between the current time and 2019-09-02 13:00:00 in minutes, divided by 30000
- DO (percent saturation) = DO observed plus the difference between the current time and 2019-09-02 13:00:00 in minutes, divided by 30000/11.3*100, with 11.3/100 being the potential saturation of oxygen in water at 10 degrees C
2020-08-11 07:00:00 to 2020-08-17 12:50:00
- DO (mg/L) = DO observed plus the difference between the current time and 2020-08-11 07:00:00 in minutes, divided by 6500
- DO (percent saturation) = DO observed plus the difference between the current time and 2020-08-11 07:00:00 in minutes, divided by 6500/11.3*100, with 11.3/100 being the potential saturation of oxygen in water at 10 degrees C
2020-08-19 20:00:00 to 2020-08-24 10:40:00
- DO (mg/L) = DO observed plus the difference between the current time and 2020-08-19 20:00:00 in minutes, divided by 6500
- DO (percent saturation) = DO observed plus the difference between the current time and 2020-08-19 20:00:00 in minutes, divided by 6500/11.3*100, with 11.3/100 being the potential saturation of oxygen in water at 10 degrees C
2020-08-26 12:00:00 to 2020-09-02 10:50:00
- DO (mg/L) = DO observed plus the difference between the current time and 2020-08-26 12:00:00 in minutes, divided by 10000
- DO (percent saturation) = DO observed plus the difference between the current time and 2020-08-26 12:00:00 in minutes, divided by 10000/11.3*100, with 11.3/100 being the potential saturation of oxygen in water at 10 degrees C
2020-09-05 06:00:00 to 2020-09-09 17:50:00
- DO (mg/L) = DO observed plus the difference between the current time and 2020-09-05 06:00:00 in minutes, divided by 3000
- DO (percent saturation) = DO observed plus the difference between the current time and 2020-09-05 06:00:00 in minutes, divided by 3000/11.3*100, with 11.3/100 being the potential saturation of oxygen in water at 10 degrees C
Water temperature (thermistors)
2021:
Thermistors at 1m(SN 3724) and 4m(SN 3747) were reading higher than normal (See below late October-December 2020) and were removed on January 7, 2021 14:30 and replaced on February 26, 2021 12:00 with SN 3966 and SN 4013. No data was recorded at those depths and was flagged in the Flag_All column and individual flag columns of the sensors affected.
The temperature string was tangled after some maintenance on June 18th, 2021 at 10:00 and was fixed on June 21, 2021 at 10:00. It was tangled again on July 21st, 2021 13:20 and fixed on July 26th, 2021 at 12:00. Consequently, there was a data gap for all the temperature sensors with a data flag of 1 in the Flag_All column and individual flag columns of the sensors affected.
2020:
The temperature string was tangled after some maintenance on October 9th, 2020 and was fixed on October 11, 2020. Consequently, there was a data gap for the temperature sensors at 5, 6, 7, 8 and 9 m from October 9-11, 2020, with a data flag of 1 in the Flag_All column and individual flag columns of the sensors affected.
On October 19, 2020, the NexSens T-Node FR Thermistor (Serial No. 3718) deployed at 0.1 m stopped working and read NA until October 20, when we took the thermistor from 2 m and put it at 0.1 m. There was a data gap at 2m from October 20-22, 2020. We put the NexSens T-Node Thermistor (Serial No. 3903) from Beaverdam Reservoir, another one of our sites, at 0.1 m in FCR and returned the 2 m thermistor back to its original depth at 2 m on October 22, 2020. Consequently, there was a data gap for 0.1 m temperature from October 19-20, 2020 and a data gap for 2 m temperature from October 20-22, 2020, all flagged with data code 1 .
In late October-December 2020, two of the Nexus T-Node FR Thermistors (at 1 m and 4 m) appeared to be reading water temperatures higher than the sensors deployed above them. On 7 Jan 2021, we put the 2 suspect thermistors and 4 other thermistors adjacently located (0.1, 2, 3, and 9 m) in a water bath in the field for an hour. The water bath showed that thermistors at 1 m and 4m were exhibiting higher readings than the others in the bath. The 1 m thermistor was reading about 0.2496℃ higher, and the 4 m thermistor was reading about 0.3862℃. To correct for this discrepancy, we created an offset for both the 1 m and 4 m thermistors from November 1, 2020 to December 31, 2020 (the time period in which the thermistors were consistently exhibiting higher temperature readings). First, we took the average temperature of the thermistors directly above and below the suspect ones for each 10-min interval observation during November 1 to December 31, 2020. Second, we calculated the difference between the suspect thermistor’s reading and its neighbor average, which we averaged over the entire time period to calculate a mean offset. Third, this mean offset was subtracted from the observed thermistor reading on each observation. For example, for the 1 m thermistor we took the average of the 0.1 m thermistor and the 2m thermistor for the neighbor average at each time interval, then we found the difference between the 1 m thermistor and the neighbor average to calculate the offset value for the entire November 1-December 31 period. The offset value for the 1 m thermistor was 0.22617oC and for the 4m thermistor was 0.18122oC. The offset was subtracted from the 1 m thermistor values from October 29, 2020 13:00EST to the end of 2020 and the offset for the 4m thermistor was subtracted from the raw values on October 31, 2020 05:00EST to the end of 2020. These times were chosen because the broken thermistors were reading consistently higher than the one above them. The observations that received an offset were flagged with data code 8.
Pressure Transducer at 9 m
The temperature string was tangled after some maintenance on June 18th, 2021 at 10:00 and was fixed on June 21, 2021 at 10:00. It was tangled again on July 21st, 2021 13:20 and fixed on July 26th, 2021 at 12:00. Consequently, there was a data gap for data from the pressure transducer with a data flag of 1 in the Flag_All column and individual flag columns of the sensors affected.
EXO sonde
2021 Maintenance:
On March 22, 2021, the dissolved oxygen sensor and depth sensor were calibrated. The dissolved oxygen sensor was calibrated with a one point 100 percent air saturation. On April 16, 2021, the fluorescent dissolved organic matter(fDOM) sensor was calibrated with a one point calibration in DI water or a 0 point calibration. The chlorophyll sensor was checked with DI water but was not calibrated. Due to a bad calibration of the fDOM sensor it was recalibrated on April 26, 2021 and all data were set to NA during that time. On June 15, 2021 a one point 100 percent air saturation calibration was done on the dissolved oxygen sensor. On August 16, 2021 a two point calibration was done on the dissolved oxygen sensor with 100 percent air saturation and a 0 percent calibration. On December 13, 2021 the EXO was brought in the lab for a deep clean, o ring maintenance, and zero calibration check of chlorophyll and fDOm and redeployed the next day.
2020 Maintenance:
On August 7, 2020, the YSI EXO (1.6 m) Optical Dissolved Oxygen Sensor Cap (Serial No. 20E101629) was replaced, and the dissolved oxygen sensor was calibrated on August 17, 2020 using a one point 100 percent air saturation calibration. The EXO 1.6 m dissolved oxygen sensor was recalibrated on 15 September 2020, 19 October 2020, and 27 December 2020 using a one point 100 percent air saturation calibration.
The YSI EXO sonde was removed from the reservoir on December 2, 2020 for manufacturer cleaning, maintenance, and calibration. It was re-deployed in the reservoir on December 27, 2020. Below is the status of each sensor: pass or fail, and the calibration, or maintenance information. All sensors were calibrated by the manufacturer on December 21, 2020 unless otherwise stated.
YSI EXO2 Sonde Serial No. 18D103367, Pass
Replaced power-communication circuit board due to issues at final testing, depth sensor passed
Wiped Conductivity/Temperature Sensor Serial No. 18C102992, Pass
Temperature reads 20.78 °C in a 20.70 °C bath with a cell constant of 0.47
Conductivity reads zero sp cond. in air.
Central Wiper for Conductivity/Temperature Sensor Serial No. 18D101236, Pass
Replaced head & motor O-rings
Dissolved Oxygen Sensor Serial No. 18D101168, Pass
Probe ODO Gain: 1.24
Total Algae Sensor Serial No. 18D102222, Pass
Chlorophyll-a: Probe reads 14.73 RFU in 16.4 standard-calibrated to standard. Probe reads 58.65 µg/L in 66 standard-calibrated to standard.
Phycocyanin: Probe reads 13.36 RFU in 16 standard-calibrated. Probe reads 13.37 µg/L in 16 standard-calibrated to standard
fDOM Sensor Serial No. 18C100625, Pass
Probe reads 275.69 uncalibrated QSU in 300 QSU standard- calibrated to 300QSU. Probe reads 92.18 uncalibrated RFU in 100 RFU standard- calibrated to 100 RFU
EXO Chlorophyll-a and Phycocyanin QAQC
The chlorophyll-a and phycocyanin sensor data underwent a QAQC process to remove extreme outliers (see ‘chla and phyco qaqc’ section of catwalk_EDI_QAQC_all_variables.R script). Outliers were defined as absolute values greater than four standard deviations from previous and subsequent measurements for each individual observation. These outliers were set to NA and identified in the Flag_Chla or Flag_Phyco column as a value of 2. In October 1st, 2018- March 3rd, 2019 outliers were identified as the absolute values greater than four standard deviations from the average measurement. These outliers were set to NA and flagged as 4 in the Flag_Chla or Flag_Phyo. After March 3rd, 2019 outlier values identified this what were left in the dataset and flagged as 5 for questionable value.
EXO fDOM QAQC
fDOM sensor data underwent a QAQC process to remove extreme outliers (see ‘fdom qaqc’ section of catwalk_EDI_QAQC_all_variables.R script). Outliers were defined as absolute values greater than two standard deviations from previous and subsequent measurements for each individual observation. These outliers were set to NA and identified in the Flag_fDOM column as a value of 2. Any value greater than 4 standard deviations above the mean was flagged with a 5 and left in the dataset.
EXO Conductivity, Specific Conductivity and Total Dissolved Solids(TDS)
Conductivity, Specific Conductivity and TDS sensor data underwent a QAQC process to remove extreme outliers (see ‘fdom cond, spcond and TDS’ section of catwalk_EDI_QAQC_all_variables.R script). Outliers were defined as absolute values greater than two standard deviations from previous and subsequent measurements for each individual observation. These outliers were set to NA and identified in the Flag_fDOM column as a value of 2. Any value greater than 4 standard deviations above the mean was flagged with a 5 and left in the dataset. During the summer of 2020 there were some high conductivity events. Any value above 42 microsemanis for conductivity and specific conductivity were flagged and over 30 mg/L for TDS were flagged as 5 but not removed from the data set.
Instruments, manual, and serial numbers used in 2018, 2019, 2020 and 2021:
Multiple sensors and instruments were added over time; below we note serial numbers of sensors with accompanying depths and provide links to the accompanying instrument manuals.
NexSens T-Node FR Thermistors (July 2018 - current)
https://www.fondriest.com/pdf/nexsens_t-node_fr_manual.pdf
Serial No. 3718; Depth 0.1m (July 2018- October 19, 2020)
Serial No. 3741; Depth 0.1m (October 20-22, 2020)
Serial No. 3903: Depth 0.1m (October 22, 2020- current)
Serial No. 3724; Depth 1m (July 2018-January 7, 2021)
Serial No. 3966; Depth 1m (February 26, 2021-current)
Serial No. 3741; Depth 2m (July 2018-October 20, 2020; October 22, 2020-current)
Serial No. 3746; Depth 3m
Serial No. 3747; Depth 4m (July 2018-January 7, 2021)
Serial No. 4013; Depth 4m (February 26, 2021-current)
Serial No. 3751; Depth 5m
Serial No. 3786; Depth 6m
Serial No. 3823; Depth 7m
Serial No. 3851; Depth 8m
Serial No. 3857; Depth 9m
InSitu RDO Pro-X Dissolved Oxygen Sensor (July 2018 - December 2021)
https://in-situ.com/wp-content/uploads/2014/11/RDO_PRO-X_Manual.pdf
Serial No. 584491; Depth 5m
Calibration from July 5, 2018 to December 19, 2018; slope = 1.0361, offset = 0.00
Calibration from August 17, 2020 to September 15, 2020; slope = 1.1074, offset = 0.00
Calibration from October 19, 2020 to current; slope = 1.1447, offset = 0.00
Serial No. 584488; Depth 9m
Calibration from July 5, 2018 to August 20, 2018; slope = 1.0743, offset = 0.00
Calibration from August 20, 2018 to December 19, 2018; slope = 1.0000, offset = 0.00
Calibration from August 17, 2020 to September 15, 2020; slope = 1.1656, offset = 0.00
Calibration from September 15, 2020 to October 19, 2020; slope = 1.0916, offset= 0.00
Calibration from October 19, 2020 to August 17, 2021; slope=1.1221; offset=0.00
Calibration from August 17, 2021 to current; slope is 1.1534; offset is 0.0054.
Campbell Scientific Stainless-Steel Pressure Transducer CS451 (May 2020-current)
https://s.campbellsci.com/documents/us/manuals/cs451-cs456.pdf
Serial No. 29010799
YSI EXO2 Sonde Serial No. 18D103367 (August 2018 - December 2021)
https://www.ysi.com/File%20Library/Documents/Manuals/EXO-User-Manual-Web.pdf
Wiped Conductivity/Temperature Sensor Serial No. 18C102992
Central Wiper for Conductivity/Temperature Sensor Serial No. 18D101236
Total Algae Sensor Serial No. 18D102222
fDOM Sensor Serial No. 18C100625
Calibration on April 16, 2021; pre: 1.38 RFU, 4.68 QSU; post: 0.00 RFU, -0.01 QSU - that was a bad calibration
Calibration on April 26, 2021; pre: 0.34 RFU, 1.13 QSU ; post: 0.00 RFU, 0.00 QSU
Pressure Transducer Serial No. 17K100864
Dissolved Oxygen Sensor Serial No. 18D101168
YSI EXO Optical Dissolved Oxygen Sensor Cap Serial No. 18C105373 (August 2018- August 7, 2020)
YSI EXO Optical Dissolved Oxygen Sensor Cap Serial No. 20E101629 (August 7, 2020-December 2021)
Calibration from August 17, 2020 to September 15, 2020; DO Gain=1.13
Calibration from September 15, 2020 to October 19, 2020; DO Gain=1.11
Calibration from October 19, 2020 to December 2, 2020; DO Gain=1.12
Calibration fromDecember 27, 2020 to December 31, 2020; DO Gain=1.10
Calibration from March 22, 2021 to June 15, 2021; DO Gain=1.10
Calibration from June 15, 2021 to August 16, 2021; DO Gain=1.13
Calibration from August 16, 2021 to current; DO Gain= 1.10
