Bear Spring served as the site of our experimental field study west of the town of Eyota in southeastern Minnesota (4,869,610N, 557,713E, UTM Zone 15). Bear Spring consists of a perennial spring (Bear Perennial Spring, ID 55A0000406 in the Minnesota Karst Features Database (Minnesota Department of Natural Resources, 2022)) that includes multiple boils within a spring house and another point just south of the spring house as well as an overflow spring (Bear Overflow Spring, ID 55A0000572 in the Minnesota Karst Features Database (Minnesota Department of Natural Resources, 2022), 4,869,551N, 557,689E, UTM Zone 15) ~60 m south-southwest of the perennial spring that becomes active when water levels are high.

We planned to conduct three artificial recharge experiments that involved emptying a pool full of water at different rates into the overflow spring. We pumped water from the perennial spring to fill up the pool for each experiment, and salt was added to the water in the pool for the first and third artificial recharge experiments. The water temperature in the pool was impacted by the surrounding surficial conditions and the time that the water sat in the pool before each experiment began. We also poured fluorescent dye into the overflow spring with the injection stream of water from the first artificial recharge experiment at 18:04:11 UTC on June 11, 2016, and we confirmed a direct preferential flow path between the overflow spring and the perennial spring with the first dye appearance at the perennial spring at 18:19:44 UTC.

Rain fell before we were able to start the third experiment on the morning of June 12, 2016. We did not monitor for rain, but this natural recharge event increased spring discharge and caused the overflow spring to start flowing sometime between 13:13-13:40 UTC on June 12. We eventually poured the pool, but the third experiment resulted in the pool water flowing along the overflow run at the surface instead of through the subsurface conduits because of the precipitation-induced hydrologic changes to the system.

Hydrologic sensors and pressure transducers were submerged below the water surface at the spring or spring run. We measured water level, temperature, and electrical conductivity just upstream of the weir using a Van Essen CTD-Diver and atmospheric pressure compensation using barometric pressure measurements collected onsite. At the same site in the spring run, we also measured the same parameters using a Campbell Scientific CR10 data logger with a Campbell Scientific 247-L conductivity/temperature probe and a vented Druck PDCR 830 pressure transducer. We also measured electrical conductivity and temperature in the spring house and at the far upstream perennial emanation using two additional Campbell Scientific CR10 data loggers with Campbell Scientific 247-L conductivity/temperature probes. Resolution and accuracy for the CTD-Diver conductivity measurements were 0.1% and 1% of the reading, respectively, and 0.01C and 0.1C for temperature, respectively. Resolution and accuracy for the Campbell conductivity measurements were 0.001 mS/cm and 5% (within the range of 0.44 to 7.0 mS/cm), respectively, and 0.01C and 0.4C for temperature (-24C to 48C), respectively. NaCl standards were used to convert electrical conductivity to the excess dissolved NaCl concentration that resulted from our injection. We converted water level to discharge using a 120 v-notch weir that we attached to the end of the culvert on the upstream side of the road. We also used a salt trace discharge measurement (at ~16:05 UTC on June 12) to calibrate the discharge measurements using the weir.

The raw hydrologic data are included. There are also two additional files that include processed data from the Van Essen CTD-Diver and the Campbell Scientific CR10 data logger by the weir, and the processing steps are described below. It is important to note that this was a short-term field experiment, so sensors, materials, etc., were all adjusting to new environmental conditions on warm summer days. In addition, the water level and discharge data capture the spring response to the installation of the v-notch weir.

Files
File BearSpring-AtmosphericPressure.csv includes Date/time (UTC), Atmospheric pressure (cm), and Atmospheric temperature (C) data that was recorded in the spring house.
 
File BearSpring-WeirCTD-raw.csv includes Date/time (UTC), Pressure-water and air (cm), Conductivity (mS/cm), and Temperature (C) data that was recorded by the Van Essen CTD-Diver by the weir.
 
File BearSpring-WeirCTD-adjusted.csv includes Date/time (UTC), Level above v-notch bottom (ft), Discharge (L/s), and NaCl (mg/kg H2O) data that was calculated from the data from the Van Essen CTD-Diver by the weir.
-The Level above v-notch bottom (ft) column was produced by subtracting the atmospheric pressure (in cm, from the BearSpring-AtmosphericPressure.csv file) from the Pressure-water and air (in cm, from the BearSpring-WeirCTD-raw.csv file) from the CTD-Diver, subtracting an offset of 3.34 cm from the entire time series (so that the average level from 16:01:00 to 16:10:55 UTC on June 12, 2016 yielded a discharge value (276 L/s) that was in agreement with the salt trace discharge measurement taken during that time), and then converting the level to ft.
-The Discharge (L/s) column was calculated using the equation for a 120 v-notch weir:
Q = 4.33*H^2.5					 (1)
where Q is the discharge in ft^3/s and H is the height above the bottom of the v-notch in feet, and then applying the conversion from cubic feet to liters.
-The NaCl (mg/kg H2O) column was determined for times around the first two pool pour experiments by subtracting the background conductivity just before the pours and then applying a calibration determined from conductivity measurements in known NaCl standards (multiplying conductivity in mS/cm by 730.6 to convert to NaCl in mg/kg H2O). Nan values in this column correspond to outliers in the raw conductivity data or calculations that resulted in a negative value.

File BearSpring-WeirCampbell-raw.csv includes Date/time (UTC), Voltage (V), Conductivity (mS/cm), and Temperature (C) data that was recorded by the Campbell Scientific CR10 data logger by the weir.
 
File BearSpring-WeirCampbell-adjusted.csv includes Date/time (UTC), Level above v-notch bottom (ft), Discharge (L/s), Discharge-cleaned up (L/s), and NaCl (mg/kg H2O) data that was calculated from the data from the Campbell Scientific CR10 data logger by the weir.
-The Level above v-notch bottom (ft) column was produced by applying the following calibration
L = 1.1787*V - 0.417864				(2)
where L is the level in ft and V is the voltage in V from the pressure transducer. With this calibration, the average level from 16:01 to 16:10 UTC on June 12, 2016 yielded the salt trace discharge value (276 L/s) measured during this time.
-The Discharge (L/s) column was calculated using equation (1) above for a 120 v-notch weir, and then applying the conversion from cubic feet to liters.
-The Discharge-cleaned up (L/s) column includes some of the data from the previous column, but offsets were applied to the level data before the discharge calculation to intervals from 21:35 on June 10 to 14:03 on June 11, and from 14:04 to 14:50, 14:51 to 16:34, and 18:05:47 to 18:13:21 on June 11 (all in UTC). There are clearly some impacts on the data while changing between 1 min and 1 s output intervals, but we do not fully understand some of the offsets. It is possible that the sensor or the pole where it was mounted was bumped/adjusted, or it may be malfunctions of the pressure transducer. The offsets from 21:04 to 22:43 on June 11 and 23:55 on June 11 to 1:08 on June 12 (all in UTC) are due to pump operation at the perennial spring to fill up the pool before and after the second pool pour (in preparation for the third pool pour). Thus, there is a physical mechanism that accounts for these two offsets, and these offsets are preserved in this column. Finally, some intervals were deleted where the data appeared to be inaccurate.
-The NaCl (mg/kg H2O) column was determined for times around the first two pool pour experiments by subtracting the background conductivity just before the pours and then applying a calibration determined from conductivity measurements in known NaCl standards (multiplying conductivity in mS/cm by 524.3 to convert to NaCl in mg/kg H2O). The calibration for the third pool pour experiment was slightly different due to the larger conductivity and NaCl concentrations (multiplying conductivity in mS/cm by 582.8 to convert to NaCl in mg/kg H2O), and it also included a linear, shifting conductivity background since conductivity was increasing in response to the natural recharge event as it was returning to the pre-recharge condition.

File BearSpring-SpringHouse-raw.csv includes Date/time (UTC), Conductivity (mS/cm), and Temperature (C) data that was recorded by the Campbell Scientific CR10 data logger in the spring house.
 
File BearSpring-UpstreamEmanation-raw.csv includes Date/time (UTC), Conductivity (mS/cm), and Temperature (C) data that was recorded by the Campbell Scientific CR10 data logger at the upstream emanation.

References
Susan Bilek. (2016). Delineating preferential flow paths from recharge for water planning and management [Data set]. International Federation of Digital Seismograph Networks. https://doi.org/10.7914/SN/XK_2016.

Minnesota Department of Natural Resources. (2022). Minnesota Karst Features Database. Minnesota Department of Natural Resources, Groundwater Atlas Program, https://www.dnr.state.mn.us/waters/groundwater_section/mapping/springs.html, accessed July 18, 2022.