One Ranney Well Can Make a Difference: The Impacts of a Radial Collector Well on Groundwater Level and Quality in the Cedar River Alluvial Aquifer

The City of Cedar Rapids, Iowa, depends on groundwater from the Cedar River alluvial aquifer for residential and industrial use. In 2020, the city completed an additional radial collector well, or Ranney well, and was concerned that pumping from the well at high rates may lower water level elevations in the aquifer, reduce yields from nearby production wells, and change the quality of produced water. During an operational test of the well's pumps, the U.S. Geological Survey and the city collected water level and water quality data to evaluate the effects of increased pumping rates on the aquifer and nearby production wells. Results indicated that a high rate of pumping from the new well caused sustained declines in water levels near the well and other nearby production wells, and, if maintained, the aquifer water level in the vicinity would continue to decline to levels observed during the 2012 drought. Aquifer specific conductance and temperature were altered and matched trends and values of the river, and river‐to‐well travel time was shortened from 7‐17 days to about 3 days. Results may also provide insights to other municipal water resource managers when considering wellfield design, production expectations, and long‐term management strategies for radial collector well production during drought, low streamflow, and times when high concentrations of nitrate‐N or organic pesticide compounds in the river may limit production options.


Introduction
The City of Cedar Rapids is the second largest city in Iowa with a 2019 estimated metropolitan population of 133,562, a 5.5% increase since 2010 (U.S. Census Bureau 2018Bureau , 2021. The city is located along the Cedar River in Linn County, Iowa, and depends on groundwater from the Cedar River alluvial aquifer (aquifer) for municipal and industrial use. The Cedar Rapids Utilities Department (city) pumps and treats between 30 and 50 million gallons per day (MGD; Dustin Elin, Cedar Rapids Utilities Process Control Specialist, written commun., 2020) to meet the municipal and industrial needs of the City of Cedar Rapids and surrounding area. In 2020, the city completed an additional radial collector well (RCW; Ranney well), RCW5, in the City of Cedar Rapids Seminole wellfield (Figure 1). The RCW5 is located about 200 ft east of the Cedar River channel and consists of a 16-ft, reinforced-concrete vertical shaft that terminates on a concrete caisson. Seven horizontal perforated steel pipes (laterals) connect near the bottom of the shaft (about 55 ft depth) and extend in a radial pattern horizontally up to 200 ft into the aquifer. The RCW5 was designed to produce up to 15 MGD from the aquifer (Bruce Jacobs, Cedar Rapids Utilities Engineering Manager, written commun., 2020). The city desired to optimize operation of the well to minimize interference with other wells and to ensure consistent water quality. The U.S. Geological Survey partnered with the city in an opportunistic study to improve understanding of the aquifer providing groundwater to RCW5 and to measure groundwater level and water quality during an operational test of the RCW5 pumps (test).
The test was conducted from 10:00 AM on September 1, 2021, to 9:00 AM on September 14, 2021. During that time, three pumps in RCW5 were cycled on and off to maintain an average pumping rate of 10.5 MGD. The USGS and the city collected continuous water level and water quality data from nearby observation wells, city production wells, and the Cedar River before, during, and after the test. In this paper, water level and water quality data (water temperature and specific conductance, SC) collected during the test are presented and interpreted, and the effects of pumping from RCW5 on water level and water quality in nearby production and monitoring wells are evaluated. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Scope and Purpose
Water level and water quality data collected from August 30, 2020, to February 28, 2021, in RCW5, nearby vertical wells, RCWs, and the Cedar River were used to determine the effects of pumping from RCW5 during an operational test on water level and water quality in the aquifer, and neighboring wells in the Seminole wellfield. The results of this study will help the city determine optimal pumping rates for RCWs and provide a better understanding of the immediate and long-term effects of pumping from RCW5 on water quantity and quality in the Seminole wellfield. The city can also use these results to develop new and more adaptive management strategies for periods of drought when low streamflow and increased demand on RCW production cause low water levels in the aquifer and reduced stream recharge rates. Results may also provide insights to other municipal water resource managers when considering wellfield design, production expectations, and long-term management strategies for radial collector well production during drought, low streamflow, and times when high concentrations of nitrate-N or organic pesticide compounds in the river may limit production options.

Description of the Study Area
The RCW5 is located near the bank of the Cedar River and is one of several RCWs among many vertical wells in the Seminole wellfield ( Figure 1). The aquifer near the Seminole wellfield is generally composed of 42-72 feet (ft) of sand and gravel and eolian sand deposits (Schulmeyer and Schnoebelen 1998;Iowa Department of Natural Resources 1998a). Although the aquifer is considered hydraulically connected to the underlying Silurian, carbonate bedrock (Iowa Department of Natural Resources 1998b; Turco and Buchmiller 2004;Iowa Geological and Water Survey 2010;Gannon et al. 2011;Haj et al. 2021), the connection may be limited due to the presence of shale at the bedrock contact (Boyd 1998;Johnson et al. 2020;Kalkhoff 2021). During low flows in the Cedar River, vertical and horizontal groundwater flows in the aquifer near areas unaffected by municipal pumping are toward the river channel (Squillace 1996). During high flows, flow directions near the channel are reversed and water from the Cedar River infiltrates the aquifer. Flow direction may also be reversed near production wells (Schulmeyer and Schnoebelen 1998;Haj et al. 2021). Production wells in the Seminole wellfield were screened in sand and gravel overlying the bedrock surface (Schulmeyer and Schnoebelen 1998;Turco and Buchmiller 2004;Haj et al. 2021). The mean pumping rates from vertical wells range from 0.97 to 1.42 MGD (Turco and Buchmiller 2004), and pumping from a RCW ranges from 3.0 to 11.0 MGD Haj et al. 2021).
The RCW5 is located near the Seminole (SEM) #10 vertical production well (SEM10), where aquifer properties were evaluated by previous studies. Schulmeyer (1995) reported an aquifer transmissivity at SEM10 of 10,836 ft 2 /d and a hydraulic conductivity of 158.0 ft/d. Additionally, Schnoebelen (2008) reported an aquifer transmissivity of 15,000 ft 2 /d, determined from an aquifer test on SEM10 conducted by the city. Boyd (1998) used deuterium and sulfate tracers to estimate that 49% of the water pumped from SEM10 originated from the aquifer, 5% of the water was from the underlying bedrock aquifer, and 46% of the water was from the Cedar River ( Figure 1). Travel time from the Cedar River to SEM10 ranged from 7 to 10 days, based on measurements of changing SC (Schulmeyer, 1995).

Methods
Data collected by the USGS used in this study included water level (stage) and water quality data in the Cedar River (from August 30 to February 28, 2021), groundwater level and water quality data in SEM10 (August 30 to February 28, 2021), and groundwater level data in monitoring wells GPW-2, CRM-15, CRM-16, CRM-17, and CRM-18 (August 30 to September 28, 2021) (Figure 1 and Table 1). Groundwater levels were collected using Hobo U20L Water Level Loggers ® (pressure transducers), following methods described in Cunningham and Schalk (2011); continuous data (one reading recorded every 15 min) were verified with discrete water level measurements made by field personnel. River level was measured at two USGS streamgages, Cedar River at Blairs Ferry Road at Palo, Iowa ( Figure 1 and Table 1; SW-1) and Cedar River at Edgewood Rd. (Figure 1 and Table 1; SW-3), using non-submersible pressure transducers as the pressure sensor for bubble-gage systems (Sauer and Turnipseed 2010; Table 1). One site, 083N08W13CBDA USGS CRM-SW-2, was temporarily monitored in the Cedar River ( Figure 1; SW-2) near RCW5 to measure water temperature and SC (Table 1). All water level data collected in wells and the Cedar River were converted from local datum to North American Vertical Datum of 1988 (NAVD88). Continuous water quality data (SC and water temperature) were collected using methods described in Wagner et al. (2006), and SC was used as a surrogate for dissolved solids content (Hem 1985). An In-Situ Aqua TROLL 100 Data Logger ® was used to measure SC and water temperature in SEM10 and deployed below the pressure transducer near the top of the screened interval to measure conditions representative of the aquifer and not the casing. The top of the screened interval for SEM10 is at a depth of 58.4 ft (Turco and Buchmiller 2004). All data collected are available in the USGS National Water Information System database at https://water data.usgs.gov (U.S. Geological Survey 2021).
Additional data were provided by the city and included RCW1-6 15-min time series data for water level elevation (NAVD88) and flow rates (MGD) from August 30 to September 18, 2020, and Seminole wellfield vertical well pumping ON/OFF records (Dustin Elin, Cedar Rapids Utilities Process Control Specialist, written commun., 2020). The city also provided RCW5 60-min times series data for SC (μS/cm), flow rate (MGD), water level elevation (ft), and water temperature (degrees Fahrenheit) from August 30, 2020, to February 28, 2021 (Amy Knudsen, Cedar Rapids Utilities Water Quality Specialist, written commun, 2021). Additional data provided by the city of Cedar Rapids are available (see Data S1).

Results
For presentation of test results and discussion purposes, the test was subdivided into four periods based on changes in the RCW5 pumping rate. Period 1 (P1) began at the start of the test and had a pumping rate of approximately 10 MGD ( Figure 2). Period 2 (P2) began on September 2, 2020, when the pumping rate was increased to approximately 12 MGD for most of the period; pumping rates during this period were the highest sustained pumping rates during the test. Period 3 (P3) began on September 5 when the pumping rate was decreased to, and maintained at, 10 MGD. Period 4 (P4) began on September 10 when the pumping rate was increased to 11 MGD; although, during most of P4 the pumping rate was maintained at about 10.5 MGD until the end of the test on September 14. After the test, the pumping rate in RCW5 was decreased and maintained at approximately 5 MGD ( Figure 2). Average rates of water level change were calculated over several, 7-h intervals to evaluate water level recovery or decline within selected periods ( Figure 2). Interval start and end times, summarized in Table 2, were selected to capture water level responses to constant pumping rates, and to minimize the influence of sudden increases and decreases in pumping rates. Water level data collected from the Cedar River, production wells (SEM10, RCW1, RCW2, RCW4, and RCW5), and USGS observation wells (GPW-2, CRM-15, CRM-16, CRM-17, and CRM-18) were compared to pumping records for vertical wells and RCWs during the test. River stage during the test was nearly constant at about 717.8 ft at SW-3 (U.S. Geological Survey 2021). Pumping occurred from seven vertical production wells in the Seminole wellfield (SEM6, SEM10, SEM17, SEM20-23) during all or part of the test (see Data S1). Water level data from SEM10, GPW-2, RCW1 and RCW5 showed no observed changes associated with pump on/off changes in SEM6, SEM17, and SEM20-23. Water level changes were observed in SEM10 because of pumping from that well. Additionally, USGS observation wells CRM-15, CRM-16, CRM-17, and CRM-18 located upstream, near the active vertical wells showed water levels that gradually increased from about 716.6 to 718.2 ft during the test and recorded no fluctuations associated with pump on/off changes in vertical production wells or pumping from RCW5 during the test (Figure 1; U.S. Geological Survey 2021). This gradual increase was associated with an increase in Cedar River discharge that was attributed to a rainfall event in the basin from September 6 to September 11, 2020, recorded at Cedar Rapids Municipal Airport weather station located about 8 miles south of the study area (USW00014990, not shown) (National Centers for Environmental Information 2021). From this comparison it was concluded that pump on/off changes in vertical wells had no direct, discernable effect on water levels recorded in SEM10, GPW-2, RCW1 and RCW5 during the test, and, in turn, that pumping from RCW5 during the test did not affect water levels upstream, near the vertical production wells.
Pumping from three nearby RCWs (RCW1, RCW2, and RCW4) occurred during the test (Figure 1). Pumping from RCW2 and RCW4 occurred at constant rates with few, brief changes that had no discernable, associated response in SEM10, GPW-2, RCW1 and RCW5 water levels (see Data S1). Although the water levels decreased in RCW2 and RCW4, water level changes observed near RCW5 were not associated with any changes in pumping from RCW2 and RCW4. It was concluded from this comparison that water level changes observed in SEM10, GPW-2, RCW1 and RCW5 were responses to pumping from RCW5 and RCW1 during the test. Therefore, the analysis of the test's effect on water levels in the aquifer was limited to water levels and pumping rates recorded in SEM10, GPW-2, RCW1 and RCW5.

The Effects of Radial Collector Well 5 Pumping on Water Levels
Prior to the start of the test, no pumping occurred from RCW5 and the water level in the well was 718.20 ft, 0.50 ft higher than river stage recorded downstream at Edgewood Road, and likely represents approximate river stage at RCW5 (site number 05464480; SW-3 Figures 1 and 2). Pumping was occurring from SEM10 with a water level in that well of 700.20 ft and from RCW1 at a rate of about 3.6 MGD with a water level in that well of 701.00 ft. A water level was recorded in observation well GPW-2 of 713.4 ft, 4.8 ft below river stage at Edgewood Road, and represented a lowering of the water levels near RCW1 due to pumping.
The test began on September 1, 2020, at 10:00 AM with an initial pumping rate from RCW5 of about 5.0 MGD and reached about 10.0 MGD within the first 30 min of the test; this rate was maintained throughout P1 (Table 2). Pumping from RCW1 was maintained at about 3.6 MGD and the water level in the well fluctuated from about 701 to 698 ft during P1; abrupt changes in the water level appear associated with changes in the rate of pumping from RCW1 ( Figure 2). Water level measurements in SEM10 recorded prior to the test were about 700 ft with pumping occurring from the well (Figure 2). Just after the start of the test, pumping from SEM10 was stopped (at 11:52 AM) and the water level recovered rapidly, in less than 15 min, to about 714 ft, and then after decreased for the remainder of the period, reflecting the continued drawdown of the water level near RCW5 due to pumping ( Figure 2). The water level decline at SEM10 due to RCW5 pumping was steady, about 0.067 ft/h (Table 2). During this time, RCW1 maintained a constant pumping rate, and a very small water level decline, 0.001 ft/h, was recorded in GPW-2 and indicated that pumping from RCW5 during this period had little impact on water levels near RCW1.
Period 2 was defined by an increase in the pumping rate from RCW5 to about 12 MGD and steady decline in water levels near RCW1 and RCW5. At the beginning of this period, water level in SEM10 continued to decline at 0.050 ft/h, then appeared to recover when pumping in RCW5 was decreased to about 11 MGD, after which time the water level declined at 0.067 ft/h when RCW5 pumping was increased to about 12 MGD (Table 2 and Figure 2). Two rapid recoveries were recorded during this period in RCW5 and SEM10 and reflected brief time intervals when pumping was stopped and switched to alternate pumps in RCW5. These pump switch spikes also occur in periods 3, 4, and after the test (Figure 2). During P2, pumping from RCW1 was increased to about 4.6 MGD, and a steady decline in the water level was recorded in GPW-2 of about 0.019 ft/h ( Table 2). Although abrupt water level changes in RCW1 appear linked to abrupt changes in pumping rate from RCW1, the steady decline in water level observed in RCW1 and GPW-2 during this period were attributed to combined pumping from RCW5 and RCW1, the only wells with active pumping during this period. Period 3 was defined by a decrease in pumping from RCW5, to about 10 MGD. This pumping rate was sustained for most of P3. Observation well SEM10 and RCW5 initially recorded recovery, followed by a lower rate of decline in water level that lasted about 24 h and was due to the decrease in pumping from RCW5. This recovery event occurred in the aquifer near RCW5 and was not recorded in GPW-2 near RCW1 (Figure 2). After about 24 h, water level in SEM10 continued to decrease at a rate of 0.039 ft/h when RCW1 was pumping at 4.6 MGD, and then at 0.031 ft/h when RCW1 pumping decreased to 3.5 MGD; water levels recorded in GPW-2 at these times showed a decline of 0.014 ft/h and a recovery at a rate of 0.001 ft/h. The latter recovery was temporary and likely because the pumping rate decreased in RCW1. The decrease in pumping from RCW1 also affected groundwater flow near RCW5, slowing the rate of water level decline in SEM10 (Table 2). This slowing decline indicated that a change in the pumping rate from RCW1 influenced water levels near RCW5, albeit temporary. Sustained pumping from RCW5 and RCW1 during this period collectively caused a continued decline in water level near both RCWs.
Period 4 began with a slight increase in RCW5 pumping to 11 MGD that was sustained for about the first 21 h of the period, then decreased to about 10.5 MGD for the remainder of the period. Initially, pumping from RCW1 was maintained at about 3.51 MGD and water level declined in SEM10 at 0.037 ft/h and in GPW-2 at 0.004 ft/h. At about halfway into the period, pumping from RCW1 was increased to 4.55 MGD (Figure 2), after which time water level decline in SEM10 was relatively unchanged (from 0.037 to 0.036 ft/h) while the rate of water level decline in GPW-2 increased, from 0.004 to 0.024 ft/h ( Table 2). The change in pumping from RCW1 influenced water level decline observed in nearby GPW-2 but had less effect on decline in SEM10 than observed in period 3, likely due to water level recovery near SEM10 associated with a decrease in RCW5 pumping that occurred within the previous 24 h. Water level recovery near SEM10 was also observed in periods 2 and 3 after pumping from RCW5 decreased (Figure 2). The end of period 4 coincided with the end of the test.
After the end of the test, pumping rate from RCW5 was reduced to about 5.0 MGD and water levels in RCW5 and SEM10 recovered. Although the water level in these wells declined briefly when the pumping from RCW5 was increased to about 7.0 MGD, they both showed recovery at the new pumping rate. Pumping from RCW1 continued after the close of the test, and water level in RCW1 and GPW-2 also showed recovery, after a lag of about 12 h, in response to the reduced RCW5 pumping rates.

The Effects of Radial Collector Well 5 Pumping on Water Quality
Water quality data collected from SEM10 and RCW5 were used to determine the effects of RCW5 pumping on the aquifer. Because of the nature of the operational test of RCW5 and the need for continuous municipal water supply, limitations existed in data collection. The SEM10 well is one of the city's municipal production wells, and as such, pumping from the well occurred prior to, and for the first 97 min of the test, drawing down water level in the well to approximately 700.2 ft (see Data S1; Figure 2). During that time, SC measurements fluctuated and were about 480 μS/ cm, higher than values measured in the river (SW-2) and lower than values measured in the aquifer (500 to 600 μS/ cm; Kalkhoff 2021; Figure 3A). Water temperature in SEM10 also fluctuated during this time and was about 73 °F, cooler than the temperature of the river (SW-2) and warmer than the aquifer (CRM-15) ( Figure 3B). The decreased water level and the presence of both stream recharge and aquifer water components present in the well indicate that during pumping, a steep hydraulic gradient from the Cedar River to SEM10 was established that induced recharge from the river into the aquifer. Water quality data from RCW5 were not available from September 1 to September 2, 2020 (the first 14 h of the test), after which time the immediate effects of pumping on water quality in RCW5 were measured. These limitations were unavoidable and considered during analysis.
When pumping from RCW5 began on September 1, 2020, water quality in the aquifer was immediately affected. The SC in SEM10 decreased from about 480 to 440 μS/cm during the first 3 h of the test ( Figure 3A). The SC continued to decrease during the following 4 days to 408 μS/cm on September 5, then gradually increased to 430 μS/cm at the end of the test on September 14. The SC of water pumped from RCW5 was greater than in SEM10 but followed a similar but dampened trend during the test. SC in RCW5 decreased from 534 μS/cm on September 2 to 506 μS/cm on September 5 and remained relatively constant with a gradual increase to 522 μS/cm at the end of the test. The SC trend in SEM10 generally followed that of the Cedar River (SW-2) during the test; SC in the Cedar River decreased from 436 μS/cm on September 1 to between 360 and 370 μS/cm on September 6, after which the value increased to about 480 μS/cm on September 13. The increase in the Cedar River SC coincided with an increase in discharge attributed to a rainfall event in the basin from September 6 to September 11 at the Cedar Rapids Municipal Airport weather station (USW00014990, not shown) (National Centers for Environmental Information 2021).
The water temperature response in the aquifer to the test showed a strong connection to the Cedar River. Water temperature in monitoring well SEM10 and in RCW5 initially increased when pumping started ( Figure 3B), due to streambank infiltration that occurred prior to the test (August 25 to August 29, 2020), when pumping occurred from SEM10 and the temperature of the Cedar River (SW-1) averaged approximately 81.3 °F (U.S. Geological Survey 2021). During that time prior to the test, water from the Cedar River was drawn into the aquifer and toward SEM10. This warmer recharge was then drawn toward RCW5 when RCW5 pumping began, passing through SEM10. In SEM10, water temperature increased to 81.5 °F on September 2 before falling to about 75 °F on September 4, after which time it remained relatively constant until September 11. During that time the temperature in SEM10 was like the temperature of the Cedar River from about September 1 to September 7. (Figure 3B). After September 11, water temperature in SEM10 decreased to 67.5 °F at the end of the test.
The water temperature changes in SEM10 were like the changes observed in the Cedar River (SW-1 and SW-2), with diurnal fluctuations dampened and a lag time during the test of about 4 days. Water temperature in the Cedar River (SW-2) fluctuated diurnally from late August until September 8 at which point water temperature gradually decreased until September 10 when it again increased ( Figure 3B). The diurnal fluctuation was not apparent from September 8 to September 13. The decrease in stream temperature and lack of diurnal fluctuation was likely due to basin runoff that occurred from September 6 to September 11. The travel time of stream recharge from the Cedar River to SEM10 of 3 days was calculated as the difference of the peak mean daily water temperature of the Cedar River (SW-1) on September 8, 2020, of 82.2 °F and the peak mean daily water temperature in SEM10 on September 12, 2020, of 81.3 °F (U.S. Geological Survey 2021).
The water temperature in RCW5 initially increased from 56.4 °F (when first measured but was likely lower) to about 65.0 °F on September 5 and then remained constant during through P3 and indicated an increase in stream recharge relative to aquifer water in the well. During P4, RCW5 water temperature began to decrease about 1 °F by the end of the test, likely caused by the general decrease in river water temperature ( Figure 3B). After the test, pumping from RCW5 was decreased and water temperature in SEM10 and RCW5 remained like the water temperature of the Cedar River and indicated RCW5 maintained a connection with and continued to inducing warmer recharge from the Cedar River deep into the aquifer.
Although SEM10 SC and water temperature data during the test lacked the variability observed in the Cedar River, the general trends in the river data are observed in the SEM10 SC and water temperature data with a time lag. The SC of water in the aquifer near the Seminole wellfield is typically between 500 and 600 μS/cm (Kalkhoff 2021; Figure 3A). Although SC values measured in RCW5 during the test were in the typical range, the values fluctuated like the river and with a delay (lag). The subtleness of the SC changes in RCW5 were due to the SC values in the Cedar River being close in value to those measured in the aquifer; this was not the case with water temperature values. The temperature of water in the aquifer (measured in CRM-15) was relatively constant, ranging from 56.3 to 56.6 °F during the test, while the temperature in the Cedar River was generally higher, ranging from 75.5 °F on September 2 near the beginning of the test to 56.8 °F on September 10 ( Figure 3B). Water temperature and SC results indicate a strong connection between RCW5 and the Cedar River.
Previous studies have documented the connection between SEM10 and the Cedar River. Boyd (1998) used deuterium and sulfate tracers as evidence of a strong connection between the SEM10 production well and the Cedar River. Groundwater flow models (Schulmeyer 1995;  Schulmeyer and Schnoebelen 1998;Turco and Buchmiller 2004;Haj et al. 2021) were used and estimated that most water pumped from SEM10 originated from storage in the aquifer (49%) and the Cedar River (46%). During this study, pumping rates from RCW5 were up to 10 times greater than previously observed in nearby vertical production well SEM10, and SC and water temperature data from SEM10 and RCW5 indicated that much of the water pumped from RCW5 was likely induced recharge from the Cedar River.
Potential Long-Term Implications of Radial Collector Well 5 Production Schulmeyer (1995) indicated that water levels in the SEM10 vertical production well ranged from 12 to 20 ft below river stage, resulting in a steep gradient from the Cedar River toward the well. Following the test, RCW5 began pumping to provide municipal supply and caused decreased water levels at SEM10 where groundwater level continued to decline throughout the fall and winter months ( Figure 4A). When the SEM10 pump turned on and off quickly (15 min), it caused large responses (a change of 16-22 ft) in the water level in the well (U.S. Geological Survey 2021). On January 21, 2021, water levels were measured at about 31 ft below land surface when the SEM10 pump was off. On February 3, 2021, the SEM10 pump turned on and caused water level to rapidly decline to a level below the pressure sensor ( Figure 4A); pumping continued during this time and the water level was assumed to have continued to decline. On February 23, 2021, a discrete water level measurement of 56.7 ft below land surface was made while the SEM10 pump was on, and the pressure sensor was above the water level ( Figure 4A). These data indicated that pumping from RCW5 greatly lowered the water level in the aquifer near the well, greatly steepened the hydraulic gradient from the river to RCW5, increased the rate that the Cedar River recharges the aquifer, and impacted water quality.
Pumping from RCW5 after the test, into the fall and winter months, at a rate of about of 5.0 MGD continued to induced recharge from the river into the aquifer and altered the previously observed water temperature profile and the SC measured in SEM10 ( Figure 4B and 4C). The SC in SEM10 was like the SC in the Cedar River and lower than the aquifer SC measured in areas not impacted by municipal pumping, 700 to over 800 μS/cm (Squillace 1996) and in CRM-15 (Figures 3 and 4). The SC in SEM10 ranged from 411 to 609 μS/cm, substantially less than three discrete values measured (711 to 751 μS/cm) in CRM-15, which is located near the upland slope (Figure 1). The SC trend in SEM10 from late summer into winter closely followed the SC trend in the Cedar River. The SC was variable in SEM10, but at the end of the test it was 430 μS/cm and had increased to 572 μS/cm when monitoring ended on January 20, 2021 ( Figure 4B). Conductance in the Cedar River increased from 480 μS/cm on September 15 to 585 μS/cm on December 10, 2020, when the sensors were removed for winter.
In a pattern like that of SC, water temperatures in the aquifer in the vicinity of RCW5 reflected those in the Cedar River rather than those from the aquifer in areas unaffected by municipal pumping. Water temperature from monitoring well CRM-15 gradually increased from 55.0 °F on August 1 through the summer and fall to 57 °F at the end of September before decreasing through the fall and winter to 47.8 °F ( Figure 4C). In contrast, water temperature in SEM10 was 81.3 °F on September 2 during the test before decreasing through the remainder of the summer, fall, and winter months to 32.4 °F on January 20, 2021. Water temperature in SEM10 closely followed those recorded in the Cedar River. These water temperatures were warmer than expected (values measured in CRM-15) in summer and fall months and were colder than expected in late fall and winter months. Pumping from RCW5 resulted in cold (near freezing) Cedar River water to flow into the aquifer in January 2021 based on recorded observations in SEM10.
Collected data indicated that greater pumping rate shortened travel time from the Cedar River to pumping wells. Comparison of water temperature data from the Cedar River (SW-1) and SEM10 collected from September 1 to December 10, 2021, indicated an estimated travel time of 3 days, which contrasts Schulemeyer's (1995) documented travel time of 7 to 17 days between the river and SEM10 (Figure 4C). Specific conductance data were not sufficient to estimate travel time during the test. Additional water level and water quality data and refined groundwater flow simulations are needed to better estimate travel time from the Cedar River to RCW5.

Discussion and Conclusions
The need for a sustainable source of water for a growing population and industry required that the City of Cedar Rapids, Iowa, increase production from the aquifer, an abundant, high-quality source for the city that typically requires minimal treatment before distribution. Pumping from RCW5 in the Seminole wellfield at high rates will result in decreased water levels in the aquifer, decreased production from nearby wells, and an increase in induced recharge from the Cedar River. Water from the Cedar River will mix with aquifer water resulting in changes in the quality of water typically pumped from the aquifer.
Water level results from data collected during the RCW5 test indicated that a high rate of pumping from RCW5 caused sustained declines in water levels near RCW5 and RCW1, and, when pumping from RCW5 was greatly reduced after the test, the water level recovered near both RCWs. The observations also indicated that changes in pumping from RCW5 affected the water level and rate of water level decline near RCW1, and, similarly, changes in pumping from RCW1 affected the water level and rate of water level decline near RCW5. Additionally, it can be inferred by comparing pumping rates and water levels during this test that a sustainable rate of pumping from RCW5 exists between 7.0 and 10.0 MGD-while the Cedar River is at low flow and the pumping rate from RCW1 is like rates observed during the test (Figure 2). Moreover, the data indicate that if a high rate of pumping from RCW5 and RCW1 continued, further declines in the water level would have occurred.
The rates of water level decline measured during the test in SEM10, from 0.031 to 0.067 ft/h, were greater than those observed during the drought of 2012 (August 1, 2012, to November 12, 2012) when the water level in RCW1 declined from 685.97 to 673.77 ft (0.005 ft/h) and caused a reduction in pumping from RCW1, triggering the emergency construction of a diversion ditch near RCW1 to augment recharge from the river Haj et al. 2021). If water level declines observed in 2012 had persisted, the city would have needed to shut down production or greatly reduce pumping from one or several RCWs (Bruce Jacobs, Cedar Rapids Utilities Engineering Manager, personal commun., 2020).
In addition to declining water levels, high pumping rates from RCW5 caused changes in water temperature and SC in the aquifer. Water quality changes were due in a large part to water level decline in the aquifer that created a steep water level gradient from the Cedar River to RCW5. The SC of water pumped from RCW5 during the test was less and the water temperature was substantially greater than in the aquifer in areas not affected by pumping.
Continued pumping of 5.0 MGD from RCW5 after the pump test through the following fall and winter substantially altered the normal water temperature profile and the dissolved solids content as measured by SC of the shallow groundwater by continuing to induce recharge from the river. As observed during the test, SC and water temperature trends in SEM10 from late summer into winter closely followed the trends in the Cedar River rather than those in areas of the aquifer unaffected by municipal pumping.
Collected data and results from this study can be used to design additional, tailored stress tests of the aquifer to better understand and quantify the interactions of the RCWs and the Cedar River. Data and results will also be used in additional calibration of an existing groundwater flow model for simulating the interactions of the RCWs, the Cedar River, and the aquifer to improve forecasting the effects of pumping from RCWs on the quantity and quality of groundwater during drought or low-flow conditions .
The high production rates from the RCWs can induce more recharge from the Cedar River than smaller vertical production wells. The continued pumping from RCW5 after the test resulted in a gradual decrease in water temperature to near freezing in SEM10 (32.4 °F), and, when pumping from SEM10 began, water levels rapidly declined to a level below the sensor. Pumping continued during that time and the water level is presumed to have continued to decline until a discrete measurement made on February 23, 2021 (when the SEM10 pump was running), showed a water level of 669.04 ft, which is 56.69 ft below land-surface, approaching the elevation of the RCW5 laterals. Water quality in water withdrawn from the RCWs was comparable to that of the Cedar River, which at times, has had concentrations of nitrate-N that exceed drinking water standards (Iowa Department of Natural Resources 2006; Kalkhoff 2021) and measurable amounts of organic pesticide compounds (Boyd 2000;Kalkhoff 2021). Increased production from a small number of RCWs may limit the city's ability to blend water from other parts of the aquifer that have lower nitrate-N concentrations.

Disclaimer
Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Supporting Information
Additional supporting information may be found online in the Supporting Information section at the end of the article.
Data S1 Supporting Information