Previous Investigations

Results from the long-term multidiscipline cooperative study between the City of Cedar Rapids Water Department and the USGS have been published in data and interpretative reports. Water-quality data collected as part of this study were previously summarized at 5-year intervals in data reports by Schulmeyer and others (1995), Schnoebelen and Schulmeyer (1996), Boyd and others (1999), Littin and Schnoebelen (2010), Littin (2012), and Meppelink and others (2019).

Data collected during this study were used to understand the hydrology and water-quality characteristics of the alluvial aquifer and the interaction of aquifer water with the Cedar River. Schulmeyer (1995) analyzed the effect of the Cedar River on the quality of groundwater near the well fields. Schnoebelen and Schulmeyer (1996) documented hydrogeologic data collected and compiled from October 1992 to March 1996. Schulmeyer and Schnoebelen (1998) described the hydrogeology near the well fields, documented a groundwater-flow model constructed to simulate regional groundwater flow under steady-state conditions, identified sources of water to the well fields, and assessed temporal and spatial variations of selected water-quality constituents and properties. Boyd (1998) characterized groundwater flow near the well fields using selected environmental isotopes and tracers. Boyd (2000) evaluated the occurrence and distribution of concentrations of selected pesticides in the alluvial aquifer and Cedar River following springtime application of these pesticides to upstream cropland areas.

Kalkhoff (2018) documented spatial and temporal differences in nitrogen and phosphorus transport from the Cedar River Basin during 2000 to 2015 that can assist in documenting progress of efforts by the City of Cedar Rapids in the Middle Cedar Partnership Project, the Iowa Department of Natural Resources Watershed Management Authority in the upper and middle Cedar River, and the University of Iowa Watershed Approach in the middle Cedar River Basin to reduce downstream flooding and improve water quality. Garrett (2021) developed and documented a model based on real-time turbidity measurements that provided information to evaluate progress of nutrient reduction efforts in the Cedar River Basin. Most recently, Kalkhoff (2021) summarized the effect of pumping on spatial and temporal hydrologic and water-quality variability of the Cedar River alluvial aquifer in Linn County, Iowa, from 1990 to 2019.

During water years 2018 through 2022, geophysical data (Deszcz-Pan and others, 2018; Johnson and others, 2020) were collected to refine the lithology and extent of the aquifer (Valder and others, 2018), and an updated groundwater-flow model to simulate effects of drought on water availability was developed (Haj and others, 2021).

Purpose and Scope

This report presents the results of water-quality data-collection activities in water years 2018 through 2022 (October 2017 through September 2022) for a multidiscipline study of the Cedar River alluvial aquifer completed by the USGS, in cooperation with the City of Cedar Rapids Utilities Water Division. Selected water-quality constituents were monitored continuously in the Cedar River and Morgan Creek, and water levels were monitored continuously in selected wells during the 2018–22 period (USGS, 2023a). Only water-quality data from periodic samples are summarized in this report; all data are publicly available in the USGS Water Data for the Nation database (USGS, 2023a). Data presented in this report include results of water-quality analyses and physical characteristics of surface water flowing into Cedar Rapids from the Cedar River and Morgan Creek. Water quality within the Cedar River alluvial aquifer was determined by periodically sampling two monitoring wells, seven production wells in the Cedar Rapids well fields, and raw water entering the two drinking water treatment plants. Samples were analyzed for selected physical characteristics at the time of collection, constituents that included nitrate as nitrogen, selected pesticides of interest for drinking water supply, and major ions to document the basic water chemistry in the surface and groundwater in the study area.

Description of the Study Area

Cedar Rapids is within Linn County in east-central Iowa. Water for the City of Cedar Rapids is supplied from three well fields (Seminole, East, and West) along the Cedar River (fig. 1). The City of Cedar Rapids had a population of about 137,700 in 2020 (U.S. Census Bureau, 2023). The Cedar River flows from the northwest to the southeast in the study area (fig. 1) and drains 6,342 square miles upstream from the Cedar River at Blairs Ferry Road at Palo, Iowa, streamgage (USGS station number 05464420; hereafter referred to as “CRPalo streamgage”; fig. 1). Morgan Creek flows from the southwest to northeast in the study area and drains 16.7 square miles in Linn and Benton Counties upstream from the Morgan Creek near Covington, Iowa, streamgage (USGS station number 05464475; hereafter referred to as “MorgCr streamgage”). Morgan Creek flows into the Cedar River upstream from the Seminole well field (fig. 1). Upstream land use in the Cedar River and Morgan Creek Basins is greater than 90 percent agriculture, which is dominated by corn and soybean production (Iowa Department of Natural Resources, 2006). Livestock raised in Cedar River and Morgan Creek Basins upstream from the study area include cattle and hogs. Annual precipitation in the Cedar Rapids, Iowa, area was as follows: 26.09 in. (2017), 40.65 in. (2018), about 37.5 in. (2019), 31.55 in. (2020), 23.69 in. (2021), and 28.36 in. in 2022 (National Oceanic and Atmospheric Administration, 2025).

Extreme daily mean flows recorded at the CRPalo streamgage during this reporting period were a maximum of 61,700 cubic feet per second (ft³/s) on March 18, 2019, and minimum of 525 ft³/s on December 26, 2017 (U.S. Geological Survey, 2023a). These extreme flows compare to the long-term (1903–2022) maximum daily mean flow of 138,000 ft³/s on June 13, 2008, and minimum of 140 ft³/s on November 18, 1989, at the downstream streamgage at Cedar Rapids (USGS station number 05464500) (U.S. Geological Survey, 2023a).

Hydrogeologic units in and near the Cedar Rapids well fields include unconsolidated surficial deposits of loess, glacial till, and Cedar River alluvium (alluvial aquifer), underlain by Silurian and Devonian carbonate bedrock. The Cedar River floodplain ranges between about 1,000 and 3,300 ft wide in the study area and is bounded by steep bluffs that rise nearly 200 ft above the river valley, exposing bedrock in some places. The upland topography is characterized by rolling hills of low relief typically formed in loess and glacial till. The alluvial aquifer ranges from 5 to 95 ft thick near the well fields, with the thickest parts nearest to the present-day location of the Cedar River and the thinnest alluvium adjacent to the valley walls. The alluvial aquifer is characterized by a sequence of coarse sand and gravel at the base, grading upward to fine sand, silt, and clay near the surface. Bedrock in the study area consists primarily of jointed and fractured Paleozoic limestone and dolomite, with interbedded chert and shale (Schulmeyer and Schnoebelen, 1998). The buried bedrock surface has its own complicated erosional topography with multiple superimposed incised channel networks reflecting the area’s glacial history. This Silurian-Devonian sequence has a maximum thickness of about 700 ft in the study area, and although no production wells have been completed in this aquifer, it is used locally for private and industrial water supply. The unconsolidated surficial deposits in the Cedar River Valley, underlying Devonian and Silurian carbonate bedrock, and deeper hydrogeologic units are described in detail by Hansen (1970), Wahl and Bunker (1986), and Schulmeyer and Schnoebelen (1998).

The Cedar River is in direct hydraulic connection with the alluvial aquifer (Turco and Buchmiller, 2004), and the alluvial aquifer is recharged by infiltration from the river, as well as by precipitation and seepage from underlying and adjacent hydrogeologic units. In areas affected by production well pumping, groundwater flows from the Cedar River toward the well fields, whereas in other areas, groundwater generally flows toward the river. Hansen (1970) calculated an approximate transmissivity of the alluvial aquifer to be about 20,000 square feet per day (ft²/d), whereas Schulmeyer (1995) determined that transmissivity varies between 1,500 and 19,000 ft²/d, depending on the physical properties of the alluvium. In May 2006, a contractor to the City of Cedar Rapids performed an aquifer test using Seminole 10 (an abandoned well located on the edge of the riverbank) that yielded a transmissivity value of approximately 15,000 ft²/d (R. Hesemann, Cedar Rapids Water Department, oral commun., March 2007).

Well Construction and Nomenclature

Wells sampled during the study included 2-inch outer-diameter monitoring wells. The monitoring wells were installed using hollow-stem auger drilling techniques and completed with polyvinyl-chloride flush-joint casing. Bentonite grout was installed around the casing 6 to 8 ft below land surface, and the wells were capped with a cement pad at the surface. Well depths ranged from 18 to 97 ft. Well-construction information for all wells is listed in table 1.

Monitoring wells are named according to a convention that includes the year the well was installed (for example, 1993), the agency identifier (USGS), the local project identifier (CRM, for Cedar Rapids Monitoring), and a unique incremental number (beginning with number 1). For example, well 1993USGS CRM–3 is the third monitoring well installed by the USGS for this project. For convenience in this report, all sites have been given a short name (table 1).

Sample Preparation and Laboratory Analysis

Water samples for analysis of nutrients and major ions were filtered through a 0.45-micrometer (µm) pore size Aquaprep polycarbonate disk or Pall capsule filter in the field. Water samples for pesticide analysis in 2018 and 2019 were filtered through a 47-millimeter (mm) diameter, 0.7-µm pore size glass-fiber filter in a Teflon filter holder into a 1-liter (L) baked glass bottle. Starting in 2020, pesticide samples were filtered using a 25-mm diameter, 0.7-µm pore size borosilicate glass-fiber syringe tip filter into a baked glass 20 milliliter (mL) vial. After collection, water samples were kept chilled until shipped overnight to the USGS National Water-Quality Laboratory in Denver, Colorado, for analysis.

Samples were analyzed to determine nutrient concentrations using colorimetric methods (Patton and Kryskalla, 2011) for dissolved species and alkaline persulfate digestion (Patton and Kryskalla, 2003) for unfiltered samples. Analytical methods used for major ions are described by Fishman (1993). Inductively coupled plasma–atomic emission spectrometry was used to determine boron concentrations (Struzeski and others, 1996). The high-temperature combustion method was used to determine the total organic carbon concentration in samples (ASTM International, 2019).

Nutrients, dissolved organic carbon, physical characteristics, and major ions from water-quality samples, the Chemical Abstract Service Registry Number, laboratory reporting limits (LRL), and reporting units are listed in table 2. Pesticides from water-quality samples, followed by the Chemical Abstract Service Registry Number, and LRLs are listed in table 3. The LRL is used to specify the lowest quantifiable value for constituents listed in tables 2 and 3. The LRL is defined more rigorously by statistics than the older minimum reporting level that it replaces (Oblinger Childress and others, 1999).

During this period of record, the analytical method for pesticides was changed from a C–18 solid-phase extraction and gas chromatography/mass spectrometry method (Sandstrom and others, 2001) to direct aqueous-injection liquid chromatography–tandem mass spectrometry (LC–MS/MS) in December 2019 (Sandstrom and others, 2015). The analysis was changed to analyze for a broader range of constituents at a lower LRL (table 3). The LC–MS/MS analytical method includes more than 200 fungicides, herbicides, insecticides, and associated transformation products (TPs) with results at similar or lower concentrations than previously available methods. The pesticides represent a broad range of chemical classes and were selected based on criteria such as current-use intensity, probability of occurrence in streams and groundwater, and toxicity to humans or aquatic organisms (Sandstrom and others, 2015)

Quality Assurance and Quality Control

To properly interpret water-quality data and to verify these data are reliable and accurate, QA procedures are followed and QC samples are collected in addition to the environmental samples. In general, QA includes using correct procedures and protocols, proper documentation (log books and field sheets) and approved analytical methods. The QC samples typically are used in the estimation of the magnitude of bias and variability of the environmental samples. Bias is systematic error that can “skew” results in either a positive or negative direction. The most common source of positive bias in water-quality studies is contamination of samples from airborne gases and particulates or inadequately cleaned sampling equipment between uses and locations. Variability is the degree of random error of independent measurements of the sample quantity. Variability may be the result of errors in laboratory analytical procedures or in collection of samples in the field. The QA/QC procedures are followed to ensure the data collected meet standards of reliability and accuracy.

The QA/QC procedures for the study followed USGS protocols (U.S. Geological Survey, variously dated) and other USGS guidelines (Mueller and others, 1997) and included documenting any deviations made in the field. Approximately 5 percent of the total samples collected for the study were analyzed for QC including equipment blanks, field blanks, and replicates. Generally, field blanks are used to estimate sample bias, whereas replicates are used to estimate sample variability.

A field blank is a water sample that is intended to be free of the analytes of interest. Two types of commercially available blank waters were used for equipment and field blanks. Organic blank water (OBW) and inorganic blank water (IBW) are certified by the manufacturer (Ricca) to be free of either organic compounds (OBW) or inorganic compounds (IBW). In the case of equipment blanks, blank water was passed through all sampling equipment in a “clean environment,” such as the laboratory, to examine the cleanliness of the equipment before sampling. A field blank is a specific type of blank sample collected in the field and is used to demonstrate that (1) equipment has been adequately cleaned to remove contamination introduced by samples obtained at the previous site; (2) sample collection and processing have not resulted in contamination; and (3) sample handling, transport, and laboratory analysis have not introduced contamination (Mueller and others, 1997). Field blank samples of the OBW and IBW were collected by passing water through all pumps, filter holders, and filters to verify the cleanliness of sampling equipment and technique. Field blank sample concentrations of inorganic constituents typically were at or below the LRL. For this dataset, calcium, chloride, dissolved organic carbon, ammonia, nitrite as nitrogen, nitrite, and phosphorus were all detected in field blank samples. Of the 11 detections in field blank samples, eight of them were close to the reporting levels and three were unexplainable. There were no detections of organic constituents in field blank samples.

Replicates are two or more samples collected or processed so that the samples are considered essentially identical in composition. All replicate samples for groundwater and surface-water point samples were collected as sequential samples (that is, they were collected one after the other, utilizing the same techniques and filters, as necessary). All composited surface-water samples were collected as split replicates (that is, they were collected in and processed from the same container but were processed as replicates at the laboratory). For the purposes of this report, the terms “environmental sample” and “replicate sample” are used to identify a particular sample in a replicate pair.

One objective of collecting replicate pairs was to estimate the precision of constituent concentrations determined by sample processing and analysis. Analytical results of organic constituents generally are more variable than those of inorganic constituents. Replicate pairs were compared by using relative percent differences (RPD). The RPD between sample pairs was calculated using the following equation:

A large relative percent difference can indicate greater variability between samples. Variability for all constituents in the replicate samples generally was within 10 percent of the environmental samples. As shown in table 4, the maximum RPD for nutrients ranged from 0.00 to 7.14 percent, the maximum RPD for organic carbon and major ions ranged from 0.46 to 49.0 percent, and the maximum RPD for pesticides ranged from 0.46 to 104 percent. It should be noted that when comparing low concentrations between replicate pairs, the RPD can appear relatively large because slight differences (common at the lowest detection levels) can result in higher RPDs. For example, an environmental value of 0.029 milligram per liter (mg/L) for bromide with a replicate value of 0.024 mg/L has an RPD of 9.4 percent, although the absolute difference between the pair was 0.005 mg/L. The median RPD for nutrients ranged from 0.00 to 1.14, the median RPD for organic carbon and major ions ranged from 0.00 to 14.0, and the median RPD for pesticides ranged from 0.00 to 18.3.

Surrogates were added to all environmental and quality-control samples for pesticide analysis before sample preparation in the laboratory. A surrogate has physical and chemical properties similar to those of the analytes of interest but is not normally present in environmental samples. Surrogates provide QC by monitoring matrix effects and gross processing errors (Wershaw and others, 1987) and help control for bias, either positive or negative. Surrogate recoveries of organic chemicals are expressed in percent and typically range from 80 to 120 percent. Three surrogates—carbendazim-d4, 2,4-D-d3, and alachlor-d13—had a combined 14 recoveries of 0. The minimum median recovery for all sites and surrogates was 48.8, and the maximum was 110.5. Surrogate recoveries that consistently are less than 70 percent may indicate that many targeted compounds may be present in greater concentrations than reported. Minimum, maximum, and median surrogate percent recoveries are listed in table 5.

Physical Characteristics, Nutrients, and Major Ions

Physical characteristics were measured at each sampling site when a water-quality sample was collected. Physical characteristics reported include temperature, pH, dissolved oxygen, specific conductance, turbidity, alkalinity, bicarbonate, and carbonate. Summary statistics for the physical characteristics of sample water from individual sites are listed in table 6.

Nutrient data were compiled for ammonia as nitrogen, nitrite as nitrogen, nitrite plus nitrate as nitrogen, orthophosphate as phosphorus, ammonia plus organic nitrogen as nitrogen, and phosphorus. Dissolved organic carbon data are summarized with the nutrient data. Major ion data were compiled for total dissolved solids, iron, manganese, fluoride, calcium, magnesium, silica, sodium, chloride, sulfate, boron, potassium, and bromide. Major ion data are required for characterization of water chemistry and geochemical modeling. Nutrient and major ion concentration summary statistics for surface water and groundwater are listed in table 7.

Pesticides

Pesticides are used to control unwanted vegetation, insects, and other pests in agricultural and urban areas. Typically, large amounts (thousands of pounds per year) of common herbicides are applied during the growing season in the Cedar River Basin to corn and soybean crops (Schnoebelen and others, 2003). In 2019, the most recent pesticide-use data available indicated glyphosate, acetochlor, atrazine, and metolachlor were the most commonly used herbicides in Iowa (U.S. Geological Survey, 2024). Insecticides are detected less often in water, most likely because they are used in smaller amounts than herbicides, have shorter persistence, and are selectively applied during periods of reduced runoff (Schnoebelen and others, 2003). Pesticide TPs are formed when a parent pesticide compound breaks down or degrades. TPs often have been detected at higher concentrations than their parent compounds (Kolpin and others, 2000, 2004; Schnoebelen and others, 2003). There were 123 TPs analyzed using the gas chromatography/mass spectrometry and LC–MS/MS analytical methods during the 2017 to 2022 time period. The complete list of pesticides and TPs analyzed are included in table 4. Summary statistics for pesticides detected in samples are summarized in table 8.