Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters East of Hot Springs National Park, Arkansas, 2006–09

Home , Geochemistry

Prepared in cooperation with the Arkansas State Highway and Transportation Department and the Federal Highway Administration

Scientific Investigations Report 2009-5263 Revised February 2011

U.S. Department of the Interior U.S. Geological Survey Cover photograph. Display Springs in Hot Springs National Park, Arkansas. Photograph by Ralf Montanus, U.S. Geological Survey. U.S. Department of the Interior KEN SALAZAR, Secretary U.S. Geological Survey Marcia K. McNutt, Director

U.S. Geological Survey, Reston, Virginia: 2009 Revised 2011

For product and ordering information: World Wide Web: http://www.usgs.gov/pubprod Telephone: 1-888-ASK-USGS For more information on the USGS--the Federal source for science about the Earth, its natural and living resources, natural hazards, and the environment: World Wide Web: http://www.usgs.gov Telephone: 1-888-ASK-USGS

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

Although this report is in the public domain, permission must be secured from the individual copyright owners to reproduce any copyrighted materials contained within this report.

Suggested citation: Kresse, T.M., and Hay, P.D., 2009, Geochemistry, comparative analysis, and physical and chemical characteristics of the thermal waters east of Hot Springs National Park, Arkansas, 2006–09: U.S. Geological Survey Scientific Investiga- tions Report 2009-5263, 48 p. Revised February 2011 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters East of Hot Springs National Park, Arkansas, 2006–09

By Timothy M. Kresse and Phillip D. Hays

ing of water from production wells appears to be of limited Abstract possibility based on review of relevant studies. Characteristics of hydraulic conductivity, storage, and A study was conducted by the U.S Geological Survey in fracture porosity were interpreted from flow rates observed in cooperation with the Arkansas State Highway and Transporta- individual wells completed in the Bigfork Chert and Stanley tion Department to characterize the source and hydrogeologic Shale; from hydrographs produced from continuous measure- conditions responsible for thermal water in a domestic well ments of water levels in wells completed in the Arkansas 5.5 miles east of Hot Springs National Park, Hot Springs, Novaculite, the Bigfork Chert, and Stanley Shale; and from Arkansas, and to determine the degree of hydraulic connectiv- a potentiometric-surface map constructed using water levels ity between the thermal water in the well and the hot springs in wells throughout the study area. Data gathered from these in Hot Springs National Park. The water temperature in the three separate exercises showed that fracture porosity is much well, which was completed in the Stanley Shale, measured greater in the Bigfork Chert relative to that in the Stanley 33.9 degrees Celsius, March 1, 2006, and dropped to 21.7 Shale, shallow groundwater flows from elevated recharge degrees Celsius after 2 hours of pumping—still more than 4 areas with exposures of Bigfork Chert along and into streams degrees above typical local groundwater temperature. A sec- within the valleys formed on exposures of the Stanley Shale, ond domestic well located 3 miles from the hot springs in Hot and there was no evidence of interbasin transfer of groundwa- Springs National Park was discovered to have a thermal water ter within the shallow flow system. component during a reconnaissance of the area. This second Fifteen shallow wells and two cold-water springs were well was completed in the Bigfork Chert and field measure- sampled from the various exposed formations in the study ment of well water revealed a maximum temperature of 26.6 area to characterize the water quality and geochemistry for degrees Celsius. Mean temperature for shallow groundwater in the shallow groundwater system and for comparison to the the area is approximately 17 degrees Celsius. The occurrence geochemistry of the hot springs in Hot Springs National Park. of thermal water in these wells raised questions and concerns For the quartz formations (novaculite, chert, and sandstone with regard to the timing for the appearance of the thermal formations), total dissolved solids concentrations were very water, which appeared to coincide with construction (includ- low with a median concentration of 23 milligrams per liter, ing blasting activities) of the Highway 270 bypass-Highway whereas the median concentration for groundwater from the 70 interchange. These concerns were heightened by the shale formations was 184 milligrams per liter. Ten hot springs planned extension of the Highway 270 bypass to the north—a in Hot Springs National Park were sampled for the study. corridor that takes the highway across a section of the eroded Several chemical constituents for the hot springs, including anticlinal complex responsible for recharge to the hot springs pH, total dissolved solids, major cations and anions, and trace of Hot Springs National Park. metals, show similarity with the shale formations in exhibiting Concerns regarding the possible effects of blasting asso- elevated concentrations, though mean and median concen- ciated with highway construction near the first thermal well trations for most constituents were lower in the hot springs necessitated a technical review on the effects of blasting on compared to water from the shale formations. The chemistry shallow groundwater systems. Results from available studies of the hot springs in Hot Springs National Park is likely a suggested that propagation of new fractures near blasting sites result of rock/water interaction in the shale formations in the is of limited extent. Vibrations from blasting can result in rock deeper sections of the hot springs flow path; the initially, low collapse for uncased wells completed in highly fractured rock. ionic strength waters, originating as shallow recharge through However, the propagation of newly formed large fractures that quartz formations, move through the upper section of the flow potentially could damage well structures or result in pirat- path into deeper sections of the flow path and are modified 2 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs by passage through shale formations present at depth. Mixing hydrogeologic conditions responsible for of the occurrence of curves for lithium, manganese, and strontium concentrations, thermal water east of Hot Springs National Park (HSNP), Hot which were greatest in the shale formations, indicate that the Springs, Arkansas, and to determine the degree of hydraulic hot springs represent an approximate 40-percent contribution connectivity between the thermal water in the well and the of water from the shale formations and a 60-percent contribu- hot springs in HSNP. The well (Bratton site) is approximately tion of water from the quartz formations. 5.5 miles east of the hot springs in HSNP. In addition to the Characterization of strontium geochemistry and analysis distance, the well is separated from HSNP by two anticline of strontium isotopic signatures were conducted by compar- complexes, at least two major thrust faults, and one surface ing geochemical analyses from the hot springs in Hot Springs watershed. On March 1, 2006, the well owner contacted HSNP National Park, shallow groundwater samples in the study area personnel regarding the appearance of warm water in a well from the quartz and shale formations, and samples from two on his property. The well had supplied water to the residence western hot springs 30 to 50 miles west of Hot Springs in for 15 years until a municipal system began supplying water in Montgomery County. Mixing model analysis indicated that 2002 and was used intermittently thereafter, without previ- the strontium geochemistry of the hot springs results from an ously yielding any noticeably warm water. A visit to the site approximately 35-percent contribution from the shales and a by HSNP personnel on March 1, 2006, documented a tempera- 65-percent contribution from the quartz formations, similar to ture of 33.9°C for water flowing from a hose outside of the that found from trace-metal analysis curves. The geochemistry house (Steve Rudd, Hot Springs National Park, oral com- results of the newly discovered thermal water sites at the two mun.), and the owner reported this phenomena had not been domestic wells were not similar to the geochemistry results experienced during past use of the well. On March 13, 2006, for the hot springs in Hot Springs National Park with respect USGS personnel visited the site and pumped the well while to strontium chemistry, but rather resemble that of the ground- monitoring water temperature. The beginning temperature was water of the local outcropping formations. The very different measured at 30.5°C and never dropped below 21.7°C after 2 isotopic and geochemical signatures observed for the thermal hours of pumping. Mean temperature for shallow (less than water from the two domestic wells as compared with the Hot 400 ft) groundwater in the area is approximately 17°C, and the Springs National Park springs do not provide evidence of any occurrence of thermal water in the well raised questions and direct hydraulic connection with the hot springs in Hot Springs concerns with regard to the timing for the appearance of the National Park. thermal water. For the purposes of this study, “thermal water” The occurrence of thermal water throughout the Ouachita is defined as groundwater having a temperature substantially Mountains tends to be found in similar geologic settings; exceeding that of normal shallow groundwater adjusted for along the nose of plunging anticlines and closely aligned seasonal variability; the maximum temperature of the hot with mapped thrust faults. The shallow flow systems have springs in HSNP—based on silica geothermometry—was been identified as being confined within the regional surface 66.6°C (Bell and Hays, 2007). watershed boundaries. The deep flow systems for thermal Of interest to HSNP personnel was that the discovery of water are likely a result of the local hydrologic and geologic the thermal water in the well appeared to coincide with the framework and represent an analog to the geologic framework construction and associated blasting of the State Highway model for propagation of groundwater to the hot springs in 270 East bypass, which terminated less than 1 mile from the Hot Springs National Park, rather than being systems in direct residence at the intersection with State Highway 70 West communication with the Hot Springs National Park thermal and was completed in the spring of 2006. HSNP personnel system. Thermal-water sites across the Ouachita Mountains raised concerns over the continuation of the bypass north appear to represent discrete systems that are associated with a of State Highway 70 West, which would be routed east of specific set of hydrologic and geologic conditions, which oc- HSNP through the previously defined recharge area for the cur at numerous locations across the region, and are viewed as hot springs (Bedinger and others, 1979) to eventually intersect analogs to the hot spring flow system in Hot Springs National with State Highway 7 South (fig.1). Park rather than connected components of the same hydraulic The discovery of thermal water outside of HSNP implied system. Concerns related to pirating of water from the hot that the geologic formations constituting the geothermal springs in Hot Springs National Park because of blasting near system in HSNP— the highly fractured Hot Springs Sandstone the thermal well sites were not supported by the data gathered Member of the Stanley Shale, the Big Fork Chert, and the for this study. Arkansas Novaculite—are capable of supporting geothermal activity east of HSNP, and that the two spatially separated systems might be hydraulically connected. The existence of the thermal system east of HSNP and observed hydrologic Introduction behavior highlighted the potential vulnerability of the thermal system resource in HSNP to changes resulting from human A study by the U.S Geological Survey (USGS) in coop- activities. Land-use changes in the area of HSNP include eration with the Arkansas State Highway and Transportation continuing urban and suburban development and expansion of Department was conducted to characterize the source and infrastructure; this expansion includes building and extension Introduction 3 171 59 !

River

e TION in al S River study area study area Hot Springs Hot Springs ta 92°53' i uach O 22 ! ARKANSAS 84 ! ! spring 83 ! ! Spring Spring EXPLAN A Caddo Gap Ouachita Mountains Mountains 50 40 298 23 39 ! ! ! change 2 ! oposed highway Cold wate r Cold water spring with water-quality data Shallow well Shallow well with water-quality data Shallow well with continuous water-level data Shallow well with water quality data and continuous water-level data Thermal water with water-quality data Thermal water with water-quality data and continous water-level data Thermal water with water-quality data, continuous water-level data, and geophysical data

Little

Spring Location map Missouri Zigzag Mountains Well or spring— Number corresponds to appendixes P r Construction limits Inte r Precipitation station 13 ! ! ! ! ! ! ! ! ! ! MILE 92°55' 1 13 ! 66 44 KILOMETER 68 71 67 1 63 ! ! 58 3 ! ! ! ! ! 65 ! 0.5 ! ! 64 ! 0.5 62 41 28 45 ! ! ! 29 ! 0 0 30 5 ! U 70 £ 9 4 !

34 Bratton ! 92°57' 15 ! ! 24 12 70 ! ! ! 31 69 25 ! ! 26 ! 8 32 ! ! 7 55 ! ! 5 72 27 53 ! ! ! 6 ! 2 78 48 ! 73 ! 42 ! ! ! 74 ! ! ! 77 ! 38 ! 270 ! £ 75 11 54 10 76 21 ! 49 ! 82 ! ! ! ! 47 ! ! ! ! 20 51 52 46 14 ! 81 92°59' ! 37 Greer ! 36 35 ! ! 19 ! 33 18 7 1 ! ! U ! 16 17 ! City of 43 ! ! Hot Springs 80 ! 93°01' 61 7 U 79 ! ! City of N A L

Hot Springs

O n i

t H O T n 60

n P A R K i

u a

S P R I N G t o

N A T I n

u M

M a

l h

t r

u S ransverse Mercator projection Hot springs 93°03' ! ! ! ! ! ! ! ! Location of study area. 94 93 92 91 90 89 88 87 86 85 Base modified from U.S. Geological Survey digital data, 1:25,000 Universal T Zone 15 34°36' 34°34' 34°32' 34°30' Figure 1. 4 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs of major roadways. Such activities are of particular interest 30 and 50 miles west of the city of Hot Springs. The geo- to the National Park Service and other entities interested in chemistry of these springs was characterized and compared to protecting the hot springs. groundwater in the study area. Because there was some initial Personnel with the HSNP initially were concerned with concern related to potential effects of blasting associated with potential new fracture connections to the hot springs as a highway construction in the study area, a literature review was result of blasting. If no fracture connections were identified, made of the effects of blasting on shallow groundwater. the area still proved of interest as a useful analog and natural model to assist in understanding the effect of human activities in the area of recharge for the hot springs. For example, Description of Study Area if the hydrologic system at the residence with thermal water The study area is located in west-central Arkansas, east of had been altered by blasting activities, especially as related and including the HSNP in the city of Hot Springs, Arkansas to development of new flow paths, this discovery would aid (fig. 1). The study area is approximately 70 mi2 and generally in understanding potential effects of similar activities on the is bounded to the north by State Highways 5 and 7 and to the shallow and deep flow components of the HSNP hot springs south by State Highway 70. The western extent of the study thermal system. area generally coincides with and includes the hot springs The HSNP, the Arkansas State Highway and Transporta- in HSNP located in the city of Hot Springs, and the eastern tion Department (AHTD), the USGS, and stakeholders includ- extent of the study is approximately 8 miles east of HSNP. ing other State and Federal agencies identified the need for a The study area is in the eastern part of the Ouachita study to characterize the newly discovered thermal water. The Mountains and is characterized by a series of steep ridges and study by the USGS, in cooperation with the AHTD, began in valleys. The land surface ranges from an altitude of greater January 2007, and new highway construction was suspended than 1,050 ft above National Geodetic Vertical Datum of 1929 until the study was completed. The main components of the (NGVD29) on the ridge tops to less than 450 ft NGVD29 in study are to: (1) perform a groundwater temperature recon- the valleys with an average slope of 0.35 ft/ft. The southern naissance, documenting the existence and distribution of part of the study area is in the Ouachita River Basin, and the thermal waters outside of HSNP; (2) conduct borehole geo- northern part is in the Saline River Basin. Except for the hot physical surveys of thermal wells; (3) conduct discrete-zone springs in HSNP, most well and spring sites inventoried and geochemical sampling of the thermal well and other additional sampled for this study were located in the sparsely populated, sites; (4) assess the degree of hydraulic connection between densely forested rural area east of the city of Hot Springs. the geothermal system of the hot springs in HSNP and the thermal well through continuous water-level monitoring at sites near both locations; (5) assess the degree of hydraulic Overview of Hydrogeologic Setting connection between the HSNP waters and the thermal well through geochemical characterization of groundwater from A basic knowledge of the geology and hydrology of the suspected recharge areas for both systems; and (6) construct study area is important in understanding the occurrence and a potentiometric-surface map defining local flow directions distribution of the various sources of thermal water inside and and boundaries in the surficial aquifer underlying potential outside the area of the hot springs in HSNP. The ridge and val- recharge areas. All data for the study were collected during ley terrain that defines the topography of the area in and near 2006 through 2009. Hot Springs is part of the Zigzag Mountains, which extend northeast and southwest from Hot Springs over a total length of approximately 22 miles and a width of up to 7 miles north- Purpose and Scope west of Hot Springs (Arndt and Stroud, 1953). The Zigzag Mountains comprise rocks ranging in age from Ordovician to The purpose of this report is to describe the geochemistry, Mississippian (table 1). These strata are intensely folded, the physical, and chemical characteristics of thermal waters and folds are closely compressed, and all folds are overturned to to perform a comparative analysis and evaluate the degree of the south. As a result, the dips off of fold axes are to the north hydraulic connectivity, if any, of thermal waters outside of (Purdue, 1910). Fold axes in the study area strike generally to HSNP to the hot springs in HSNP. This report presents field the northeast, and major faults are essentially parallel to the and water-quality data and an interpretation based on the data trends of the axes of folds and thus are longitudinal faults (fig. for the occurrence, source, and distribution of thermal water 2). Because the system of fractures lies along east-west trend- outside of HSNP. ing lines, wells of equal water-yielding capacity typically lie Included in the report is a description and comparison of east or west of each other, and yields of wells north or south the geochemistry of shallow groundwater from cold springs of each other generally are unrelated (Halberg and others, and wells, thermal water from springs and wells outside of 1968). Because the topographic basins generally are defined HSNP, and water from the hot springs in HSNP. Although the by the folded strata in the study area, resulting in synclinal report focuses on groundwater in the study area, two thermal and anticlinal basins, and shallow groundwater yields and springs were identified outside of the study area located about flow directions reflect the structural geology, orientation, and Introduction 5

Table 1. Generalized section of sedimentary rocks in the vicinity of Nature and Origin of the Hot Springs in Hot the hot springs. Springs National Park [o, outcrops in the vicinity of Hot Springs National Park; modified after Bedinger and Because this report characterizes the occurrence of ther- others, 1979] mal water outside of HSNP and assesses any hydraulic rela- Maximum tions with the hot springs in HSNP, better understanding of the thickness in hot springs flow path, particularly including the source of heat, Hot Springs area is critical for defining potential hydraulic connections and the System Formation (feet) potential for a similar hydrogeologic origin where there is no physical hydraulic connection. The origin, chemistry, and ther- Stanley Shale (o) 8,500 apeutic value of hot springs in HSNP have been of interest to Hot Springs Sandstone (o) 150 researchers from the early 1800s to the present. References to the hot springs, general chemistry of the springs, and the sur-

Mississippian Arkansas Novaculite (o) 650 rounding geology were made in early reports by Owen (1860), Devonian Glasgow (1860), Branner (1892), Haywood (1902) and Weed (1902). Purdue (1910) was one of the first researchers to Silurian Missouri Mountain Shale, (o) systematically address the origin of the hot springs. Based on Blaylock Sandstone, and 195 the intervening structure of the compressed folds, their lateral Polk Creek Shale, undif- overlapping, and the large degree of fracturing in the Ouachita ferentiated Mountains, Purdue (1910) concluded that the recharge area Bigfork Chert (o) 700 must be in the near vicinity of the springs, and, based on the Ordovician topography, stratigraphy, and structure, that the recharge area Womble Shale 1,500 must be located on the overturned, anticline valley between Sugarloaf and North Mountains (fig. 1) with the Bigfork Chert as the dominant recharging formation. The Bigfork Chert was hydraulic connection of the major east-west trending faults, identified as the most likely formation through which recharge early researchers speculated that shallow groundwater basins occurs based on the considerable thickness and outcrop extent, within the Zigzag Mountains likely had little interbasin flow of the intensely fractured nature of the rock, and the thin lay- groundwater (Halberg and others, 1968). ers that define the unit. Purdue (1910) attributed the location The primary porosity of most rocks in the study area is of the springs to the southwestern plunge of the Hot Springs negligible, and the occurrence and distribution of secondary anticline (fig. 2) and to resulting fracturing and faulting in the porosity, such as joints, fractures, and separation along bed- process of folding, although no major faults were identified in ding planes, controls hydrologic behavior and water avail- his report. ability; hence, the amount of water available from the rocks is The most comprehensive study of the origin of the hot a function of the amount of fracture density and distribution springs was conducted by Bedinger and others (1979). The (Halberg and others, 1968). A review of domestic wells in the radioactivity and chemical composition of the hot springs was study area revealed that shallow wells are to be found com- compared to that of cold-water springs; tritium and carbon-14 pleted in all of the formations, but that yields are considerably isotopes were analyzed to age date the hot springs; and a lower in shale units than in novaculite and chert rocks. The mathematical model was used to test the various conceptual brittle nature and extreme fracturing of the chert and novacu- models of the hot springs flow system. Bedinger and others lite rocks in the area lead to enhanced secondary porosity, and (1979) demonstrated a similarity between chemistry of the realization of this fact led early researchers to identify the Big- hot springs and cold-water springs and wells in the vicinity of fork Chert as the most important aquifer throughout the area in the hot springs. The main difference was in the elevated silica terms of yield (Purdue, 1910; Purdue and Miser, 1923; Albin, concentrations of the hot springs, which were used to estimate 1965; Halberg and others, 1968; Bedinger and others, 1979). the temperature at depth based on the increased solubility of The broad exposure and distribution of the Bigfork Chert silica with increasing water temperature (Fournier, 1981). in the area of the hot springs combined with its hydrologic Bedinger and others (1979) estimated that the maximum tem- importance resulted in most researchers identifying the forma- perature at depth was no more than a few degrees higher than tion, potentially in combination with the Arkansas Novaculite, the temperature of the emerging springs. as the primary recharge formation for the hot springs flow Bell and Hays (2007) calculated the maximum tempera- system (Purdue, 1910; Bedinger and others, 1979). ture at depth of 66.6°C; the mean temperature of the emerging hot springs is approximately 62°C. Hot springs temperatures vary from spring to spring and vary with time for individual springs because of changing cold-water input and to seasonal surface temperature variation. The maximum temperature attained by different thermal flow systems in the area likely 6 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs spring TION omble Shale Cold wate r Shallow well Thermal water Thrust fault— Sawteeth on upper plate Well or Spring— Number corresponds to appendixes Stanley Shale Hot Springs Sandstone Arkansas Novaculite Missouri Mountain Shale and Polk Creek Shale Big Fork Chert W Blakely Sandstone Mazarn Formation Anticline— Shows trace of crest plane and direction of plunge Syncline— Shows trace of trough plane and direction of plunge Formation EXPLAN A 50 ( (

( ( (

! ! ! Ms SO Ow Om ( Obf ( Oby MDa Mshs

( ( (

( ( ( ( (

Ms ( 92°53' Ow 22 ! Ow 2 9 8 U 40 50

Obf 23

39 !

! ! (

! ( SO

Mshs ( ( (

Ms ( SO ( ( Geology modified from Arkansas Geological Survey, 2009 92°55'

( 13 MDa !

66 ( 44 68 ( 67 58 ! ! 3 ! ! ( ! ! ! ( ! ! (71

63 ! ! ( 41 62 ( 28 64

! Obf 45 ( ! ! ! 7 0 5 29 30 £

U ! (

( ( ( 9 4 ! Mshs ! 15 MILE 34

92°57' 1

( ! 24 ! 69

70 (

12 MDa ! ( ( ! 31 ( !

KILOMETER

! ! 1 ! ( 26 8 ( 25 32 ! Bratton ! 0.5

7 Mshs

55 ( ( ! 0.5 ! 5 27 72 53 ! ! !

! ( 2 0 0 ! 48 ! ! 73 78 ! Ms ! ! ! ( 74 75

38 42

( ! ! ( !

! 54 77 Ms ( 11 10

( 76 ! ! 82 49 ! !

( ! 47 ! ( ! ! 14 51 ( ! 20 81 ! 52 46 92°59' ! 37 Greer ! Obf 35 36

! ! (

( ( 18 ( 19 33 1 ! ! 7 ( ( ( ( !

U ( 16 17 43 ! ! !

MDa Om

( ( 80

! 93°01' ( 61 7

79 ! U Oby ( Hot Springs anticline ! (

( (

(

( (

( ( ( ! (

60 ( ( MDa Ow ( Ms

Hot ( ( ( ransverse Mercator projection springs

Geology of study area. ( ( ( ! ! ! 93°03' ( ( ! !

! ! ( ! ! ! ! ! ( ( ( 94 93 93 92 92 91 91 90 90 89 89 88 88 87 87 86 86 85 85 Base modified from U.S. Geological Survey digital data, 1:25,000 Universal T Zone 15 34°36' 34°34' 34°32' Figure 2. Methods 7 varies based on maximum depths of flow and local variation in Water-Quality Sampling and Analysis geothermal gradient; temperature in these springs also can be expected to vary because of cold-water input, as is the case for A general reconnaissance of wells and springs in the the springs in HSNP. Tritium and carbon-14 isotope analyses study area was conducted during numerous trips between indicated a mixture of a small volume of young water less than October 2006 and March 2007. The reconnaissance included 20 years in age and a preponderance of water approximately spring and well location and georeferencing, collection of 4,400 years old. The mathematical model, together with the water-level data (where accessible), and field measurement of geochemical data, flow measurements, and regional geologic specific conductance, temperature, and pH of the water. Water structure, supported a concept of local, meteoric origin by samples were collected at 27 sites, including 15 wells, 10 hot recharge through the Bigfork Chert and Arkansas Novaculite. springs in HSNP, and 2 cold-water springs in September 2007 The water was theorized to be heated at depths of 5,000 to and analyzed for major cations and anions, trace metals, and 8,000 ft through the normal geothermal gradient (thought to be isotopes at selected sites, including carbon-14 and tritium between 0.006°C/ft to 0.009°C/ft; Bedinger and others,1979), abundance, and strontium, carbon, oxygen, and hydrogen and fault conduits provided the avenue for rapid transport to stable isotopes. An additional 28 well and spring sites were the surface. This model remains the most widely accepted by sampled in April 2008 for strontium isotope analysis only. the scientific community for the origin of the hot springs. For Samples for major cations, anions, and trace metals were reference, past authors have posited upwelling of juvenile analyzed by the USGS National Water Quality Laboratory water (Arndt and Stroud, 1953), the existence of an underlying (NWQL) in Denver, Colorado, following procedures described magmatic body (Weed, 1902), or reaction of minerals, such as in Fishman (1993) and Garbarino (1999). Complete analyses sulfides (Bergfelder, 1976) to explain the heat. are found in appendix 1. Concentrations preceded by a “<” Bell and Hays (2007) conducted a comprehensive symbol in appendix 1 denote that the concentration was less geochemical and water-quality study and quantified the hot than the laboratory reporting level; one-half of the laboratory springs’ local, cold-water component. Yeatts (2006) investi- reporting level concentration was used for statistical purposes. gated the recharge area for the shallow, cold-water component For instance, the laboratory reporting level for iron is less than of the hot springs, and performed groundwater tracer tests to 6.0 µg/L, and 3.0 µg/L was the value used to calculate mean calculate recharge areas. Yeatts (2006) showed a traveltime concentrations for iron. Estimated concentrations for some of 1 to 3 weeks from various dye-release points to the hot constituents are noted by an “E” in front of the concentration. springs. Continuous temperature monitoring data (Bell and Concentrations are marked as estimated for cases in which Hays, 2007; Yeatts, 2006) showed that recharge from rainfall the concentration is between the long-term method detection generally caused downward spikes in water temperature for limit and the laboratory reporting level (Childress and oth- many of the springs, providing direct evidence that cooler, ers, 1999). Because there is a 95 percent confidence level that shallow groundwater was mixing with the thermal water of only 1 percent of concentrations above the long-term method the springs. Yeatts (2006) calculated the recharge area for the detection limit are false positives, estimated concentrations cold-water component of flow to be approximately 0.1 to 0.2 were used as listed for statistical purposes in this report. Two mi2. Bell and Hays (2007) collected geochemical data on the thermal springs, Little Missouri and Caddo Gap Springs, listed hot springs and showed statistically significant differences in Bryan (1922) and located in western Arkansas, were addi- for silica, total dissolved solids (TDS), strontium, sulfate, tionally sampled in June 2008 for comparison to the well and and barium for analyses of hot springs during base-flow and spring data from the study area; samples were submitted to stormflow events. Using silica in a binary mixing model, it the Arkansas Department of Environmental Quality (ADEQ) was demonstrated that contribution of cold-water recharge to Water Quality Laboratory for analysis of major ions and hot springs during base-flow conditions varied from 0 to 16 samples for strontium analysis were submitted to the USGS percent, and the contributions during stormflow conditions NWQL. Field measurements including water temperature, pH, varied from 21 to 31 percent. dissolved oxygen, and specific conductance were collected at reconnaissance and water-quality sampling sites following procedures described in Wilde (2008). Methods Analysis of samples for determination of isotopic composition was performed at various laboratories. Oxygen and hydrogen stable isotope (oxygen-18/oxygen-16, 18O/16O, and The following sections describe methods used for collec- deuterium/protium, 2H/H) analyses were performed at the tion and analysis of water-quality samples; collection of field USGS Reston Stable Isotope Laboratory in Reston, Virginia, data including water level, pH, temperature, dissolved oxygen, by isotope-ratio mass spectrometry (Revesz and Coplen, and specific conductance; collection of borehole geophysical 2008a, 2008b). Tritium (3H) and strontium isotope (stron- data from selected wells; collection of continuous water-level tium-87/strontium-86, 87Sr/86Sr) analyses were performed by and temperature data at selected sites in the study area; and the USGS Water Resources Radiogenic Isotope Laboratory in collection of precipitation data. All data were collected by Menlo Park, California; tritium analysis was performed by use USGS personnel from March 2006 through June 2009. of electrolytic enrichment and liquid scintillation (Thatcher 8 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs and others, 1977; Ostlund and Werner, 1962), and strontium Four replicate samples and one blank sample were submit- analysis was performed by mass spectrometry (Bullen and ted with the September 2007 sampling event. A stainless- others, 1996). Carbon stable isotope (carbon-13/carbon-12, steel, positive-displacement pump with Teflon rotors (gears) 13C/12C) composition and carbon-14 analyses were performed was used to purge and sample wells (including constructed at the Environmental Isotope Laboratory on the campus of monitoring wells) that lacked dedicated pumps. An equipment University of Waterloo, Waterloo, Ontario, Canada, by accel- blank was collected using inorganic-free water passed through erator mass spectrometry (Drimmie and others, 1990; Stuiver the pump prior to field use for the September 2007 sampling and Polach, 1977). Stable isotopes are measured as the ratio of event. All results from the QA/QC sample data indicated the two most abundant isotopes of a given element. For oxy- that cleaning procedures were adequate in preventing cross gen, the ratio is 18O/16O. The relative abundance of the oxygen contamination of samples and that the laboratory results isotopes can be expressed as a deviation from a standard quan- were reproducible. Results for QA/QC samples are available tity using delta (δ) notation and symbolized as δ18O; values are at the USGS National Water Information System webpage expressed in per mil (‰): (http://waterdata.usgs.gov/nwis) and can be obtained through the USGS Arkansas Water Science Center in Little Rock, δ 18 Ox=( Rx − Rstd ) / Rstd × 1000‰ (1) Arkansas. where δ18Ox is the δ value of a sample x, Collection and Analysis of Field Data Rx is the 18O/16O ratio of that sample, and Continuous monitoring of water-level and temperature Rstd is the 18O/16O ratio of the standard. was conducted at 1-hour intervals for five wells (well 1, 16, 24, 48, and 52) located in the study area. Two of the wells A positive value of δ18O represents water with more 18O were equipped with real-time monitoring capability; data relative to 16O than the standard water, Vienna Standard Mean were transmitted by satellite on an hourly basis. The other Ocean Water (VSMOW). A zero value of δ18O represents three wells were equipped with data loggers and the data water that has the same ratio of 18O to 16O as VSMOW. A were downloaded monthly. Data were subsequently checked negative value of δ18O represents water with less 18O relative for accuracy, corrected where necessary, and stored in the to 16O than VSMOW. Similar notation is used to express the USGS National Water Information System. Water levels were concentrations of the stable isotopes of hydrogen (deuterium/ measured manually on a monthly basis and compared to trans- protium, δ2H) also referenced to VSMOW, carbon (13C/12C) ducer levels to check for cable stretch and instrument drift; referenced to the Pee Dee Belemnite (PDB), and sulfur iso- any discrepancies were entered into the calculation for the topes (34S/32S) referenced to the Canyon Diablo Troilite. corrected data to account for the variation. Precipitation data Strontium is a divalent, reactive metal that behaves simi- were collected by using a tipping-bucket rain gage at a USGS larly to calcium in chemical interactions and substitutes for gaging station (07358280) in Hot Springs, Arkansas. The data calcium in many minerals. Like Ca2+, Sr2+ is a trace component were collected at 15-minute intervals, and daily totals were of most groundwaters. Strontium has four naturally occur- calculated from these data for comparison to water-level data ring isotopes: 84Sr, 86Sr, 87Sr, and 88Sr. Strontium isotopes are at selected well sites in the study area. reported as absolute 87Sr/86Sr ratios, typically to a four- to six- A potentiometric-surface map was constructed using decimal notation (Kendall and McDonnell, 1998). Variability water levels from 53 shallow (less than 400 ft) wells and 24 in strontium isotope ratios in rocks is caused by radiogenic cold-water springs located in the study area. Water levels were formation of 87Sr at the expense of parent rubidium. Strontium measured in the wells using an electric tape, which was cali- isotope ratios show great variability through much of geologic brated to one-hundredth of a foot. Water levels were subtracted time, particularly for the Phanerzoic, due to relative contribu- from surface altitudes at each site, which were extracted from tions of strontium to the ocean from terrestrial weathering and topographic maps with a default accuracy of plus or minus hydrothermal influx from the mid-ocean ridges. Rock-water 10 ft for each well altitude. Potentiometric-contour lines were interaction imparts this strontium isotope variability as a manually contoured within a geographical information system signature to aquifers. Strontium isotopic ratios (87Sr/86Sr) have using 50-ft contour intervals. Abundant control points in the proven to be a useful tracer for groundwater movement (Clark southwestern part of the study area reflected strong topo- and Fritz, 1997; Kendall and McDonnell, 1998). Aquifer- graphic control, and topographic contours were followed for framework minerals often have distinctive strontium content estimated depth to water to extend contour intervals at loca- and specific strontium isotopic signatures that can be picked tions with a lower density of control points. up by water, enabling identification of recharge sources, trans- Borehole geophysical logging was conducted on one port pathways, and mixing. existing well (well 24) and one newly installed monitoring Protocols for instrument calibration (Wilde, 2008) and well (well 34) in the study area to characterize the geology, equipment cleaning (Wilde and others, 1998) were followed fracture zones, relative fracture yields, fluid temperature, to maintain proper quality assurance and quality control (QA/ and other elements of site hydrogeology. All logging was QC) of water-quality data associated with all sampling events. Effects of Blasting on Shallow Groundwater 9 conducted by staff of the USGS Arkansas Water Science researchers to conduct a separate study using project-specific, Center. Geophysical logs included natural gamma radiation, newly constructed wells with monitoring conducted before spontaneous potential, fluid resistivity, long and short normal and after blasting in the vicinity of each well. Four wells were resistivity, lateral resistivity, specific conductance, tempera- drilled for that project, enabling collection of project-specific ture, single-point resistance, caliper data, induction data, water-quality data, aquifer data, well specific capacity, and acoustic televiewer data (Keys, 1997), and electromagnetic other information gathered before and after advancement of flowmeter data (Century Geophysical Corporation, 2006). blasting and excavation associated with mining activities. Three monitoring wells were constructed from June Ground vibrations and water levels were monitored continu- 27-29, 2007, at two sites for the purpose of isolating zones ously. As mining and pit development approached within of thermal water contribution noted in domestic wells at each approximately 300 ft of the wells, a drop in static water level site. A 4-inch monitoring well (Bratton 2) completed with occurred in three of the four wells, followed by a substantial polyvinyl chloride (PVC) casing and screen was drilled within improvement in well performance as measured by specific 20 ft of the domestic well at the Bratton site to a depth of capacity. At the fourth site, there were no recorded changes 165 ft; a 20-ft screen was installed across the thermal zone. in the water levels. The timing of the changes recorded in the Groundwater with elevated temperature (maximum tem- three wells indicated the changes were not the direct result of perature of 26.6°C) additionally was discovered in a second blasting. The period of change, the length of time involved, domestic well, the Greer site, on June 26, 2007, during the and the location of active mining at the time of the occurrence original reconnaissance of existing wells and springs (fig. 1). indicated that the lowering of the static water levels resulted Two monitoring wells (Greer2, Greer3), one 4-inch and one from increased storage space in the aquifers as the result of 2-inch, were installed at this site at depths of 120 ft and 155 fractures becoming more open because of downslope exca- ft. Conventional rotary drilling methods were employed, and vation and resulting lateral stress relief associated with the the wells were developed until clear water appeared and field mining activities. Subsequent filling of excavations resulted measurements (pH, temperature, dissolved oxygen, and spe- in lowering yields in the wells back to previous production cific conductance) stabilized. values. Fracturing of rock commonly is used to improve permeability for enhanced recovery of hydrocarbons and at groundwater-remediation sites for improving recovery efficiency of Effects of Blasting on Shallow contaminated water and hydraulic control of groundwater-flow Groundwater directions. Lane and others (1996) used various geophysical techniques to characterize a fractured bedrock aquifer Concerns regarding the possible effects of blasting asso- and a blast-fractured contaminant recovery trench. Borehole ciated with highway construction near the newly discovered geophysical methods applied included video, acoustic tele- thermal well (Bratton site) necessitated a literature review viewer, heat-pulse flowmeter under nonpumping and low-rate on the effects of blasting on shallow groundwater systems. pumping conditions, natural gamma, electromagnetic induc- Although the effects of blasting on surface structures, includ- tion, fluid temperature and conductivity, caliper, deviation, ing buildings, have been well documented in the literature and borehole radar. An azimuthal square-array, direct-current (Duvall and Fogelson, 1962; Siskind and others, 1980), few resistivity surface geophysical technique also was used. Aqui- studies have been performed that document the effects of fer testing was performed before and after blasting. Charac- blasting on the hydrology and water quality of shallow aquifer terization was conducted before and after blasting of a trench systems, including physical effects on production wells com- used for recovery of a petroleum plume in low-permeability pleted in these aquifers. fractured bedrock. Information from six observation wells near One of the first investigations into the effects of blasting the trench, borehole and surface geophysical methods, and on groundwater supplies was in the coal production region of aquifer tests indicated that the blasting created an intensely the Appalachian Range physiographic province (Robertson fractured zone about 10-ft wide and 85-ft deep, along the and others, 1980). One part of the investigation focused on 165-ft length of the recovery trench. An increase in poros- researching 36 case histories of complaints from homeowners ity and transmissivity near the trench resulted from apparent reporting problems with their domestic wells. The com- hydraulic cleaning that occurred when water was ejected out plaints ranged from turbidity issues to perceived reduction in of the observation wells within 30 ft of the trench during the yield. Of the 36 case histories of complaints, 24 of these sites blast (water was ejected 40 ft to 50 ft into the air during blast- were investigated with no clear evidence that any of the well ing). Estimated aquifer transmissivity after the blast in wells problems were related to blasting; however, evaluating exist- within 30 ft of the trench was an order of magnitude higher ing wells with little or no knowledge of construction details, than before the blast; however, no new blast-induced fractures preblasting water-quality, preblasting specific yields, and other or increases in fracture apertures in the observation wells were information makes a preconstruction/postconstruction compar- indentified on the video or televiewer logs and none of the ison and comprehensive evaluation of blasting effects incon- calcite-filled fractures appeared to be opened by the blast. As clusive. As such, results from the investigations prompted the such, fractures developed during the blasting did not extend 10 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs to observation wells, which were within 30 ft of the blast- structures (Siskind and others, 1980), and this range in fractured trench. vibrations appears to be safe for protection of groundwater A key observation of Lane and others (1996) was that the wells ranging from 300-5,100 ft from blasting sites based on relatively high-energy blasting approach—designed specifi- a review of pertinent studies. Vibrations from blasting can cally to fracture the aquifer and increase hydraulic connectiv- result in rock collapse for uncased wells completed in highly ity—did not result in observable fracture propagation any fractured rock. However, the propagation of newly formed large distance from the blasting locus. Begor and others (1989) large fractures that potentially could damage well structures or similarly reported on an artificially produced fracture zone for result in pirating of water from production wells appears to be hydraulic control of a contaminant plume. The blast-induced of limited possibility based on review of relevant studies. fracture zone was 325-ft long and 7-ft wide and was created to provide enhanced fracture connectivity for intercepting a contaminated groundwater plume. Extensive fracturing was extended several feet radially from each shot hole with Hydrogeology of the Shallow Aquifer hairline cracks extending as far as 10 to 15 ft from blasting System locations. Much of the damage created away from blasts is associ- The hydrogeology of the shallow groundwater-flow sys- ated with the travel of wave fronts at the ground surface, and tem in the study area was defined by constructing a potentio- studies have shown much lower vibration at depths in observa- metric-surface map to evaluate flow directions and reviewing tion wells; these observations are consistent with large-scale hydrologic data from continuously monitored wells to evaluate seismic events in that most of the damage caused during the hydrologic response of wells completed in different forma- earthquakes is a result of surface waves. Ground vibrations tions to regional rain events. A potentiometric-surface map normally are measured with seismographs and are reported in was developed from water levels measured in 53 wells and 25 feet per second. The point during a measuring event where the springs within the study area. Well depths were obtained for velocity is at a maximum is termed the peak particle velocity. 35 of the wells cataloged during the reconnaissance. Of these Robertson and others (1980) noted that vibrations near the 35 wells with known well depths, 29 (83 percent) were less bottom of wells were considerably attenuated at depths of 140 than or equal to 200 ft in depth and all of the 35 wells were to 160 ft. Daniel B. Stephens and Associates (2002) studied less than 400 ft in depth. Because wells produce mainly from effects on yield and water quality in domestic wells at five fractures in the study area, and fractures tend to be increas- coal-mining sites in Pennsylvania (at least two wells at each ingly smaller in aperture and frequency at greater depths, site). This study, similar to Robertson and others (1980), also aquifer yields below depths greater than approximately 400 ft observed very little change in water quality and other factors; typically are insufficient to supply user needs. Potentiometric the minor changes in water quality noted for the study were contours of the shallow groundwater system (fig. 3) present related to household well use and sensor drift. There were a subdued reflection of topography, which indicates a high no changes in well yield over the period of measurement for degree of topographic control on shallow groundwater flow the study; however, distances of the study wells from blast- in the area. Hydraulic behavior of this shallow aquifer also is ing sites ranged from 1,293 to 5,140 ft and averaged 2,607 strongly controlled by local geology. Steep and topographi- ft. Daniel B. Stephens and Associates (2002) also placed cally rugged ridges in the area are defined and supported by transducers to measure peak particle velocities near the sur- the resistant Arkansas Novaculite and Bigfork Chert with face (0.42 ft) and at depth (from 1.1 to 20 ft) at the sites and gentle sloping valleys formed on the Stanley Shale. Recharge showed that peak particle velocities decreased with depth. The areas for infiltrating rainwater are located in the topographi- average decrease in velocity was 0.0015 inches per second per cally higher areas of exposures of Bigfork Chert and Arkansas foot below the ground surface. Golder and Associates (2004) Novaculite, and the valleys serve as groundwater-discharge reviewed numerous studies for a report on potential effects areas. Discharge occurs naturally as base flow in streams from extension of a quarry in Canada and stated that fracture draining the study area and through numerous springs. Springs development from blasting was usually limited to the equiva- can be found throughout the region in topographically higher lent distance of about 20 times the blast borehole diameter areas and in the valleys. Most springs were located in the (typically 0.25 ft), except for minor microfracturing beyond Bigfork Chert and Arkansas Novaculite formations, although this distance. Golder and Associates (2004) postulated that a springs were present in all formations exposed in the study ground-vibration limit of 2.0 inches per second is adequate to area (fig. 2). protect wells from any damage. Potentiometric-surface gradients are steeper in the val- Although additional studies are needed to fully charac- leys and less steep near the top of the ridges (fig. 3). The terize the effects of blasting on the hydrogeology of shallow high storage and high permeability of the Bigfork Chert and aquifer systems, results from available studies indicate that Arkansas Novaculite allow rapid conveyance of recharg- propagation of new fractures near blasting sites is of limited ing water from rain events through these rocks and distribute extent. Safe levels of ground vibration from blasting range this water throughout the numerous fractures; thus maintain- from a peak particle velocity of 0.5 to 2.0 in/sec for residential ing lower gradients. The lower fracture density and resulting

Hydrogeology of the Shallow Aquifer System 11

500 500 600 600 600 550 ection MILE

650 ! ( 700 1 TION flow di r KILOMETER spring 1 0.5 92°53' 0.5 EXPLAN A oundwate r ! ( Cold wate r Shallow well Thermal water Potentiometric contour— Shows altitude at which water level would have stood in tightly cased wells. Contour interval 50 feet. Dashed where approximately located. Datum is NGVD of 1929 Surface drainage divide G r 0 0 ! ( ! ( ! ! !

! ( 500

! (

650 650 600 600

550 550 92°55'

550 ! ( ! ( ! ( ! ( ! ( ! ( ! ( ! ( ! ( ! ( ! (

! (

700 ! ( 700

! (

700 650 650 ! (

! (

600 600

550 550

500 ! ( ! (

92°57'

Bratton ! ( 700 ! ( ! (

! ( ! (

! (

650

650 ! ( 700 ! (

600 600 ! ( ! ( ! ( ! ( ! ( ! ( ! (

! (

650 650

! (

700 700 550

500

! (

! ( 600

! ( 600 ! ( 650 650 ! ( ! ( ! ( ! ( ! ( ! ( ! ( ! ( ! ( Greer Greer ! ( ! ( ! ( ! ( ! ( 92°59' ! ( ! ( ! (

! (

500

500 ! ( ! (

550 550

650 650 600 ! ( 600 ! (

! ( 700 ! ( ! ( 93°01' ! ( ! ( N A L

t H O T n

n P A R K i ! (

o S P R I N G

N A T I n

u M

M a

r t

o g

N u S ransverse Mercator projection 93°03' Hot springs ! ( ! ( ! ( ! ( ! ( ! ( ! ( ! ( ! ( Potentiometric surface of shallow groundwater in study area, 2006–07. Base modified from U.S. Geological Survey digital data, 1:25,000 Universal T Zone 15 34°36' 34°34' 34°32' 34°30' Figure 3. 12 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs lower permeability, porosity, and storage in the shale forma- Comparisons of the net response of these two wells tions, including the Stanley Shale, result in a highly uneven over a time period of more than 1.5 years revealed seasonal distribution of recharged water, a rapid filling of fractures, responses in water level and provided information on the stor- and a correspondingly greater increase in water levels and age characteristics for each formation (fig. 5). From December flow gradients with rain events. The mean depth to water 2007 through May 2008, water levels in the Stanley Shale well in 25 wells completed in the Stanley Shale was 16 ft below responded strongly for individual rain events, but dropped to a land surface, as compared to mean depth to water of 75 ft in relative average base level that was approximately 31 ft below 11 wells completed in the Bigfork Chert (fig. 2). The greater land surface from September 2007 through December 2007. depth to water in wells completed in the Bigfork Chert also fits This base level increased slightly by 1 ft to approximately 30 regional models, which show that deep unsaturated zones exist ft below land surface during the wet season from February in recharge areas between the water table and the land surface; 2008 through April 2008; however, the rapid post rain event whereas, the water table is found close to or at the land surface water-level decreases continued. Water levels dropped to the in discharge areas (Fetter, 1988). Nine of the 25 water levels original average base level of approximately 31 ft below land measured in wells completed in the Stanley Shale were less surface from May 2008 through August 2008 and began to rise than 10 ft below land surface at the time of measurement. Four in September 2008 to the slightly higher base level of 30 ft. flowing artesian wells were identified during the early field The hydrograph from the Greer well, completed in the Bigfork reconnaissance, and seven of the wells measured during the Chert, showed a very different response from the hydrograph reconnaissance had water levels less than 5 ft below the land from the Bratton well, completed in the Stanley Shale, for the surface at the time of measurement. same period of measurement. Water levels slowly rose more Aquifer properties, including porosity and storage, within than 7.8 ft in the Bigfork Chert well from December 2007 the various formations were investigated by the continu- through April 2008 with only subtle inflections as a result of ous monitoring and evaluation of water-level responses to individual rain events during this period. In late May 2008, regional precipitation events for selected wells in the study water levels in the Bigfork Chert well began to decrease, and area. Hydrographs for two wells were examined, including a this trend continued into December 2008 when water levels well completed in the Stanley Shale and one completed in the began to increase. Winter and early spring are wet seasons for Bigfork Chert, representing water-level changes within these Arkansas and much of the annual aquifer recharge typically two lithologies in the study area. The relation between con- occurs during this period; whereas, the summer to early fall tinuously recorded water levels for a 3-month period (August seasons are times of lower rainfall, marked by falling water through October 2008) in a well completed in the Stanley levels for aquifers throughout Arkansas. Shale (Bratton site) and a well completed in the Bigfork Chert The differing water-level responses of these two wells to (Greer site) are shown in figure 4. Five major rain events (one individual rain events and to seasonal variation in rainfall are or more days) greater than 1 inch for each day occurred during a result of the differing storage and permeability characteris- the 3-month period. The first rain event occurred from August tics of the shale and chert formations. The brittle nature and 10-12, 2008, and resulted in an approximate 2-ft increase in intense fracturing of the Bigfork Chert resulted in substan- the water level in the Stanley Shale well (Bratton site) but an tially greater fracture porosity as compared to the fracturing increase of only 0.2 ft in the Bigfork Chert well (Greer site). in the Stanley Shale. The greater storage and permeability of The next 1-inch rain event on August 21, 2008, resulted in a the fractured chert allowed greater storage of recharging rain water-level increase of 1.2 ft in the Stanley Shale well but an water with lower water-level increases for individual rain increase of only 0.1 ft in the Bigfork Chert well. Responses for events. The lower permeability and secondary porosity in the four of the five major rain events during this period resulted Stanley Shale resulted in rapid filling of the fractures, followed in increases in the Bigfork Chert well of less than 0.3 ft, by rapid water-level declines driven by the steep gradients though increases for these four events in the Stanley Shale induced during filling of the fractures. These differences in ranged from 1.2 to 2.0 ft. The only substantial increase in the permeability correspond directly to differences in observed Bigfork Chert well was for the third rain event (two consecu- yields for wells in these formations. The well completed in tive days of 2.0 inches and 1.8 inches of rain, respectively), the Stanley Shale (Bratton site), which penetrated 140 ft of which resulted in a water-level increase of 1.4-ft increase in saturated thickness, can be pumped dry within 30 minutes of the Bigfork Chert well compared to an increase of 3.7 ft in the pumping at less than 3 gal/min; however, an observation well Stanley Shale well. Similar responses were noted for all rain drilled in the Bigfork Chert (Greer site), which penetrated only events in both wells with water levels rising substantially less 20 ft of saturated thickness, maintained a rate of 20 gal/min in the Bigfork Chert well than in the Stanley Shale well for the with negligible drawdown over a period of 30 minutes. entire period of measurement (September 2007 through May Three other sites were instrumented with long-term, 2009). Following the steep increases in water level after major continuous monitoring for water level and temperature. Exist- rain events, water levels were observed to drop sharply in the ing wells at these sites were completed in the Stanley Shale Stanley Shale well; whereas, decreases occurring over a longer (Thornton site), the Arkansas Novaculite (Thomas site), and period of time were typical of the Bigfork Chert well. the Bigfork Chert (Panther Valley site) (fig. 5). Hydrographs for wells at these three sites corroborated the observations Hydrogeology of the Shallow Aquifer System 13

26 2.5 Stanley Shale (Bratton) Depth to water Daily precipitation

27 2.0

28 1.5

29 1.0

30 0.5

31 0

95 2.5

Bigfork Chert (Greer)

Depth to water PRECIPITATION, IN INCHES Daily precipitation

DEPTH TO WATER, IN FEET BELOW LAND SURFACE 96 2.0

97 1.5

98 1.0

99 0.5

100 0 Aug. 1 Sept. 1 Oct. 1 Nov. 1 2008

Figure 4. Short-term hydrographs of Bratton and Greer wells and daily precipitation at Hot Springs (U.S. Geological Survey gaging station 07358280). 14 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

26 3.5

Stanley Shale (Bratton) Depth to water Daily precipitation 3.0

28 2.5

2.0

30 1.5

1.0

32 0.5

0 Oct. 1 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1 May 1 June 1 July 1 Aug. 1 Sept. 1 Oct. 1 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1 2007 2008 2009

92 3.5 PRECIPITATION, IN INCHES

Bigfork Chert (Greer) Depth to water 3.0 DEPTH TO WATER, IN FEET BELOW LAND SURFACE 94 Daily precipitation

2.5

96 2.0

1.5 98

1.0

100

0.5

102 0 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1 May 1 June 1 July 1 Aug. 1 Sept. 1 Oct. 1 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1 2007 2008 2009

5 3.5

Stanley Shale (Thornton) Depth to water Daily precipitation 3.0

7 2.5

2.0

1.5

1.0 11

0.5

13 0 Feb. 1 Mar. 1 Apr. 1 May 1 June 1 July 1 Aug. 1 Sept. 1 Oct. 1 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1

2008 2009

30 3.5 PRECIPITATION, IN INCHES

Akansas Novaculite (Thomas) Depth to water DEPTH TO WATER, IN FEET BELOW LAND SURFACE Daily precipitation 3.0 35

2.5

40 2.0

1.5 45

1.0

0.5

55 0 Feb. 1 Mar. 1 Apr. 1 May 1 June 1 July 1 Aug. 1 Sept. 1 Oct. 1 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1 2008 2009

0 3.5

Bigfork Chert (Panther Valley) 1 Depth to water 3.0 Daily precipitation

2 2.5

3 2.0

4 Missing data

1.5 Missing 5 data PRECIPITATION, IN INCHES

1.0 6 DEPTH TO WATER, IN FEET BELOW LAND SURFACE

0.5 7

8 0 Feb. 1 Mar. 1 Apr. 1 May 1 June 1 July 1 Aug. 1 Sept. 1 Oct. 1 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1 2008 2009

Figure 5. Seasonal hydrographs of Bratton, Greer, Thornton, Thomas, and Panther Valley wells and precipitation at Hot Springs (U.S. Geological Survey gaging station 07358280).—Continued in regard to porosity and storage in the Stanley Shale and potentiometric map was developed, shallow groundwater flow Bigfork Chert from the two wells discussed in the previous is strongly influenced by surface topography with no discern- paragraph. Water-level increases of 5.5 ft and 16.5 ft are noted able influence by faulting or other geological structural fea- from approximately January 2008 through April 2008 for the tures. The Bigfork Chert generally is located along the hinge wells completely in the Bigfork Chert (Panther Valley site) line of the topographically higher anticlinal ridges, and the and Arkansas Novaculite (Thomas site), respectively, with a higher groundwater elevations from the potentiometric-surface steady decline from approximately May 2008 through August map indicate that exposures of Bigfork Chert serve as major 2008, followed by a steady increase beginning in January recharge areas for the regional shallow groundwater system 2009, which corresponds to the rainy season in Arkansas. in the study area. Flow lines developed on the potentiometric- However, the well completed in the Stanley Shale (Thornton surface map indicated that flow moves from these elevated site) had large water-level increases in response to individual recharge areas and into streams within the synclinal valleys rain events (up to 3 ft for consecutive day rain events), but formed on exposures of the Stanley Shale. Therefore, ground- consistently fell back to a relative base level, which generally water flow follows topography, and there is no evidence of fluctuated between 9 to 11 ft below the land surface for the interbasin transfer of groundwater within the shallow flow period of measurement, similar to the well completed in the system. Stanley Shale at the Bratton site. In summary, characteristics of storage and secondary fracture porosity were interpreted from yields observed in individual wells completed in the Bigfork Chert and Stanley Geochemistry and Comparative Shale; from hydrographs produced from continuous measure- Analysis of the Shallow Aquifer and ments of water levels in wells in the Arkansas Novaculite, the Bigfork Chert, and Stanley Shale; and from a potentiometric- Thermal Water Systems surface map constructed using water levels in wells throughout the study area. The data gathered from these three separate A groundwater-sampling program was conducted to exercises corroborate findings from earlier reports (Purdue, determine water quality and geochemistry of shallow ground- 1910; Purdue and Miser, 1923; Albin, 1965; Halberg and oth- water from the various shallow formations in the study area ers, 1968) that fracture porosity is much greater in the Bigfork for comparison to the geochemistry of the hot springs in Chert relative to that in the Stanley Shale. At the scale that the HSNP. The goal of this study component was collection of Geochemistry and Comparative Analysis of the Shallow Aquifer and Thermal Water Systems 17 water-quality data that would enable further characterization Sandstone); all but one site exhibited specific-conductance of shallow flow paths, identification of recharge source litholo- values of less than 50 µS/cm with a median of 30 µS/cm. gies for shallow groundwater, and establishment of potential Specific conductance in wells completed in the shale forma- connections between shallow flow paths and deep flow paths tions (Stanley Shale, Womble Shale, and the undifferentiated of thermal waters within and outside of HSNP. Important Missouri Mountain Shale and Polk Creek Shale) ranged from information was gleaned from a general reconnaissance of 97 µS/cm to 490 µS/cm with a median value of 290 µS/cm. the study area prior to the selection of sites for sampling and The pH values observed for wells completed in shale forma- geochemical analysis. Because of the large amount of data tions ranged from 5.8 to 7.7 with a median value of 7.3; pH presented in this section of the report, discussion of the data values for wells completed in quartz formations ranged from is divided into the following subsections for sake of clar- 3.6 to 5.9 with a median of 4.5. ity: reconnaissance, general geochemistry, trace metals, and The relation of pH and conductance of groundwater isotope geochemistry. measured in 25 shallow wells during the reconnaissance was analyzed to assess the effects of geology on the field chemistry of groundwater in the study area (fig. 6). Values of pH increase Reconnaissance with dissolution of mineral species with buffering capacity, particularly carbonate minerals, with a positive correlation in A reconnaissance of existing wells and springs in the pH and conductance as a result. As such, the higher pH and study area was conducted during several trips between Octo- conductance values were observed in groundwater samples ber 2006 and March 2007. Data gathered during the reconnais- from wells completed in the shale formations as a result of sance phase are found in Appendix 2 with statistical results of the greater abundance of carbonate and other acid-buffering the data summarized below. These data enabled development minerals. Subsequent water-quality analyses revealed that and enhancement of a conceptual model for groundwater flow all water samples from the shale formations in the study area in the area and design of an efficient and effective geochemi- were a calcium-bicarbonate water type (see “General Geo- cal sampling event. chemistry” section). Rainwater in Arkansas typically is within Specific-conductance values collected from the well a pH range of 4.7 to 4.8 standard units (National Atmospheric and spring reconnaissance showed low values for springs Deposition Program, 2009). Eight of the 10 wells completed in issuing from and wells completed in the quartz rock forma- the quartz formations had pH values below that of rainwater. tions (Bigfork Chert, Arkansas Novaculite, and Hot Springs

8.00

STNL STNL 7.50 STNL STNL STNL STNL STNL 7.00 STNL STNL WMBL STNL

6.50 STNL STNL STNL

6.00 ARKS STNL BGFK 5.50 pH, IN STANDARD UNITS

5.00 ARKS Rainwater

HSPG BGFK 4.50 Groundwater samples from shallow aquifer BGFK STNL—Stanley Shale BGFK BGFK HSPG—Hot Springs Sandstone 4.00 ARKS—Arkansas Novaculite BGFK—Bigfork Chert HSPG WMBL—Womble Shale BGFK 3.50 0 100 200 300 400 500 600 SPECIFIC CONDUCTANCE, IN MICROSIEMENS PER CENTIMETER AT 25 DEGREES CELSIUS

Figure 6. Specific conductance and pH values for groundwater samples from wells completed in shallow aquifers in the study area. 18 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

In the carbonate-poor soils and rock units of the quartz The complete set of analyses for all samples is found in appen- formations, carbonic acid, formed by dissolution of soil-gas dix 1. The laboratory analyses substantiate the geochemical carbon dioxide, resulted in decreased pH relative to rainwater. contrast between groundwater from the quartz formations and Equilibrium constants show carbon dioxide generally to be groundwater from the shale formations initially indicated by far more abundant than carbonic acid in solution (Appelo and the field reconnaissance data. For the quartz formations, total Postma, 1999), and gases, which exsolved and caused bubbles dissolved solids (TDS) concentrations, as expected from the to form in sampling tubes during sampling of water from the low specific conductance values, were very low with a median quartz rocks in the study area, were interpreted as carbon concentration of 23 mg/L, whereas median TDS concentration dioxide gas. for groundwater from the shale formations was 153 mg/L. To The apparent lack of carbonates in the quartz formations, illustrate the uniquely low ionic strength of the quartz ground- evidenced by groundwater chemistry, is supported by early water, the median specific conductance of 30 μS/cm for these mineralogical studies (Weed, 1902). Chemical tests performed waters compares very closely with distilled, membrane filtra- by Weed (1902) on novaculites in the study area showed that tion, or reverse-osmosis bottled water available commercially, they consist of nearly pure silica, and that the silica is pure which typically ranges from 10 to 25 μS/cm (North Dakota quartz and not amorphous silica. Numerous rhomboidal cavi- State University, 1992; Emerson Process Management, 2004). ties were observed in novaculite rock, which corresponded in Calcium and bicarbonate concentrations for samples from form and position to calcite inclusions found in the rock. The quartz formations also were low, exhibiting median values cavities were speculated to have formed by the dissolution and of 0.4 mg/L and 5 mg/L, respectively. Silica, with a median removal of the calcite (Weed, 1902). Purdue and Miser (1923) concentration of 10.7 mg/L, generally constituted the greatest describe the upper division of the Arkansas Novaculite as proportion by weight of the TDS. being composed of massive, highly calcareous novaculite, and A Piper diagram constructed using the major ion chemis- the Bigfork Chert as containing small quantities of calcite and try, in milliequivalents per liter, for all water samples revealed pyrite. However, the poor buffering capacity implied by field substantial scattering for the percentage distribution of cations measurements of water from the quartz formations indicates and anions for groundwater samples from the quartz forma- that these formations were noncalcareous in the study area tions (fig. 7); such a distribution is likely the result of the low or that weathering at the surfaces of the fractured rock and ionic strength for the water where small concentration changes flushing by recharge water over time has removed the calcite, can considerably affect percent ion contributions. Because eliminating any buffering capacity for infiltrating slightly of the low buffering capacity and low TDS concentrations in acidic rainfall. groundwater from the quartz formations, even small changes resulting from various processes including chemical reactions, evapotranspiration, cation exchange, and the quantity, distribu- General Geochemistry tion, and solubility of mineral assemblages, can greatly affect the resulting ion percentages. Water type for groundwater from Based on results from the initial field reconnaissance, 15 the quartz formations also varied widely with sodium and shallow wells and 2 cold-water springs were sampled for a calcium as the major percentage of cations in milliequivalents more comprehensive suite of constituents to evaluate geo- per liter and sulfate, chloride and bicarbonate as the major chemistry for the shallow groundwater system. Four of the percentage of anions in milliequivalents per liter. wells were completed in the Stanley Shale, two wells in the The shale formations, the hot springs in HSNP, and the Hot Springs Sandstone, one well in the Arkansas Novaculite, western hot springs in Montgomery County all exhibited seven wells in the Bigfork Chert, and one well in the Womble strongly calcium-bicarbonate water type (fig. 7). The chem- Shale. One of the two cold-water springs sampled for the istry of the hot springs of HSNP was remarkably consistent; investigation issued from the Stanley Shale (ArScenic Spring) the 10 springs appear to be represented by a single point on and the other from the Bigfork Chert (Sleepy Valley Spring). figure 7. Most analytes for the hot springs show similarity Additionally, 10 hot springs in HSNP were sampled for the with the shale formations in exhibiting elevated concentra- investigation; all of which emerged from the Hot Springs tions; important exceptions to this similarity in water chem- Sandstone. Sampling of all sites took place from September istry being low sodium and chloride concentrations (table 9 through September 13, 2008. Review of an earlier report by 2), which compared more closely to the quartz formations. Bryan (1922) revealed three springs in Montgomery County Because the Bigfork Chert and Arkansas Novaculite have that were listed as thermal springs. Two of these thermal been identified as the reservoir rock for recharge to the hot springs (Bryan, 1922), Caddo Gap and Little Missouri Spring, springs, the chemical composition for the hot springs would hereinafter referred to as the western hot springs, were located be expected to be closer to the chemistry of samples from the in the field and sampled in June 2008. quartz formations. One potential explanation for the higher A statistical summary of selected analytes from the sam- TDS, pH, calcium, and bicarbonate concentrations observed pling of 15 shallow wells and 2 cold-water springs in the study for the hot springs relative to those of the quartz formations is area (divided into quartz and shale formations), 10 hot springs dissolution of unweathered calcium carbonate in the Big Fork in HSNP, and the 2 western hot springs is shown in table 2. Chert and Arkansas Novaculite at depth. However, trace metal Geochemistry and Comparative Analysis of the Shallow Aquifer and Thermal Water Systems 19 ------1.5 6.7 4.8 4.2 3.2 22 35 83 Value 149 233 133 Spring (1 sample) Caddo Gap ------4.0 6.2 5.3 1.6 1.6 11 17 73 98 62 37.1 Little Value Western hot springs Western Spring Missouri (1 sample) --- 7.2 7.5 1.2 3.9 1.8 4.7 39.8 45 <6 213 138 179 298 160 106 Median 1 1 --- 7.2 7.5 3.5 3.9 1.8 4.7 9.5 45 40.0 123 179 298 158 104 162 Mean --- 7.5 7.8 4.0 1.9 5.0 46 25 41.3 13.3 Hot springs 143 184 316 167 108 229 (10 samples) mum Maxi - --- 6.3 7.3 1.6 3.7 1.8 4.2 39.6 42 35 99 <6 <0.2 173 284 151 mum Mini - 0.6 6.6 2.4 3.8 5.9 11.4 17.0 22.7 36 35 113 127 153 228 164 222 Median 0.7 6.6 6.8 5.6 9.1 39 59 10.7 19.8 20.4 186 179 260 187 226 377 Mean 1.0 7.4 11.2 78 14.8 31.3 35.4 29.8 18.3 (6 samples) 611 146 273 477 249 324 mum Maxi - 1,115 Shale formations µ g/L, micrograms per liter; <, less than; --, not measured; E, estimated] 0.5 5.8 1.7 2.2 6.2 1.2 2.9 12 10.7 13 12 83 42 91 141 154 mum Mini - 3.2 4.4 0.4 1.5 1.7 3.5 9 5 0.6 4.0 10.7 27.4 38 23 30 12 Median 1 1 1 8 3.6 4.6 1.9 2.0 1.9 4.6 1.4 8.7 12 25 34 10.2 47.6 18 491 Mean 6.8 8.3 6.1 7.6 9.2 3.0 60 40.5 13.5 13.5 27 43 63 20 (11 samples) 214 mum Maxi - 2,510 Quartz formations 1 0.5 3.6 0.7 1.0 7.2 1.0 2.5 11 17 0.01 0.05 E4 <0.6 <0.2 <0.4 µ S/cm, microsiemens per centimeter at 25 degrees Celsius; mum Mini - - Selected chemical characteristics for samples collected from shallow groundwater in study area. tance, µS/cm gen, mg/L solids, mg/L mg/L µg/L Mean calculated by substituting one-half of the laboratory reporting level for censored values. 1 Characteristics Dissolved oxy - pH dissolved Total Specific conduc Calcium, mg/L Sodium, mg/L Bicarbonate, Chloride, mg/L Silica, mg/L Sulfate, mg/L Aluminum, µg/L Barium, µg/L Lithium, µg/L Iron, µg/L Manganese, Strontium, µg/L Table 2. Table [mg/L, milligrams per liter; 20 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

100 100 90 EXPLANATION 90 80 Samples from 80 Shale formations CALCIUM PLUS MAGNESIUM 70 Quartz formations 70 Western hot springs 60 60 Hot springs 50 50

SULFATE PLUS40 CHLORIDE 40 30 30 20 20 10 10 0 0

0 0 100 100

10 10 90 90

SODIUM PLUS POTASSIUM 20 20 80 80

30 30 70 70

40 40 60 60 SULFATE 50 50 50 50 CARBONATE PLUS BICARBONATE MAGNESIUM 60 60 40 40

70 70 30 30

80 80 20 20

90 90 10 10

100 100 0 0 100 90 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 80 90 100

CALCIUM CHLORIDE PERCENTAGE OF TOTAL MILLIEQUIVALENTS PER LITER

Figure 7. Piper diagram of major cations and anions from shallow groundwater and thermal water samples. Geochemistry and Comparative Analysis of the Shallow Aquifer and Thermal Water Systems 21 concentrations of the hot springs, including barium, lithium, minerals and that calcium and bicarbonate account for the manganese, and strontium, generally were substantially higher majority of dissolved constituents contributing to the TDS than those of the quartz groundwater and more similar to the concentration for the thermal water and the shale formations. shale formation groundwater. Hot springs water chemistry is The wide range of TDS, calcium, and bicarbonate concentra- likely a result of rock/water interaction in the shale formations in the shale formations is explained by varying degree tions in the deeper sections of the groundwater flow path to of rock-water interaction along shallow, cold-water flow the hot springs with the initial, low ionic strength character paths and increased dissolution of carbonate as slightly acidic inherited from shallow recharge through quartz formations recharge waters are buffered. being modified by passage through the shales at depth. This Because median constituent concentrations for the hot observation for the hot spring chemistry is in agreement with springs generally compare more closely to the median concen- the conceptual model of flow for the hot springs, whereas, the trations in shales, rather than the maximum shale concentra- water must traverse shale sequences to exit through the Hot tions (fig. 8) as one might expect for waters evolving along Springs Sandstone (table 1), likely along thrust-fault conduits a very long flow path, this provides further argument against that bring water up from depth (Bedinger and others, 1979). recharge through surficial shales as being an important part Substantial contributions to recharge near the surface are not of the recharge to the hot springs in HSNP. To illustrate, if expected from the shale formations because of relatively low hot springs water was recharging through surficial shales and permeability, low fracture porosity, and limited storage associ- traveling a long shale-dominated flow path, the increased ated with the shale formations and the lack of outcrop at eleva- opportunity for greater rock-water interaction would lead to tions sufficiently high to drive flow to the hot springs. TDS concentrations that approach the maximum constituent The two western hot springs issue from the exposures concentrations observed for shales; this is not the case for the of Arkansas Novaculite, and results from the field reconnais- hot springs waters. This observation better supports a shale sance in the study area indicate that these springs also would geochemical influence arising from consistent, well con- have water chemistry similar to that of water samples from the strained rock-water interaction along a shorter flow path, such quartz rock formations in the study area (low pH, TDS, cal- as is thought to occur as the waters of the hot springs move out cium, and bicarbonate); however, analyses revealed a chem- of the quartz formations then move through faults traversing istry closer to that of water from shale formations rather than the shales at depth. from the quartz formations (table 2, figs.7 and 8). The high Sulfate concentrations (fig. 8) were greatest in the shale values of pH and specific conductance and high concentrations formations relative to all other site types. Lower sulfate con- of calcium and bicarbonate match more closely with those of centrations in the hot springs relative to the shale formations water from the shale formations rather than from the Arkansas may be the result of sulfate-reducing conditions and precipita- Novaculite and other quartz formations. Because the springs tion of iron-sulfide minerals at depth, which is supported by have a thermal component, a plausible explanation is that the the extremely low and nondetectable concentrations of iron water is a mixture of shallow groundwater from the Arkansas in the hot springs (table 2 and fig. 8). Sulfate reduction also Novaculite and deep flow path thermal water that has traveled was reported as a possibility by Bell and Hays (2007), whose along fracture zones at depth, interacting with shale forma- sulfur isotope analyses provided results indicative of this tions to modify the water chemistry in similar fashion to that microbially mediated process. for the hot springs in HSNP. Silica concentrations (fig. 8) were highest for the hot A series of box plots were constructed for selected springs relative to the other site types and can be explained by analytes to provide a visual representation of the distribu- the increased solubility of quartz and other silica minerals with tion of concentrations for each site type and the relation of increased temperatures at depth (Fournier and Rowe, 1966). water chemistry between the four site types: the hot springs For the shallow groundwater samples, the variability in silica in HSNP, the quartz formations, the shale formations, and is a function of the silica phase present in the rock. Because the two western hot springs (fig. 8). The box distribution of silica in the quartz rocks is present dominantly in the form concentrations for samples collected from the hot springs in of pure quartz (Weed, 1902), the mineral form of silica that HSNP show the very consistent water chemistry for the 10 has the lowest solubility in typical groundwater (Hem, 1989), hot springs and little variability in samples collected from silica concentrations were lowest in cold-water groundwater the quartz formations as compared to water from the shale samples from the quartz rocks. The presence of amorphous formations, which show considerable range for most of these and cryptocrystalline silica, common silica forms in typical analytes (fig. 8). Plots for TDS, calcium, and bicarbonate (fig. shales, likely explains the greater cold-water silica concentra- 8) show the similar, relatively high concentrations for the ther- tions in the shale formations. mal water (hot springs in HSNP and western hot springs) and Aluminum concentrations (fig. 8) were highest in the the shale formations with much lower concentrations for water quartz formations relative to the other site types, and are from the quartz formations. The overall similarity in appear- explained by the increased solubility of aluminum at lower pH ance for TDS, calcium, and bicarbonate is explained by the values rather than rock type (fig. 9). Aluminum solubility and fact that calcium and bicarbonate concentrations are derived concentrations increase substantially at pH values approaching stoichiometrically from the dissolution of calcium carbonate 4.0 (Hem, 1989). The median pH of groundwater from 11 sites 22 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

300 100 Total dissolved solids Calcium 250 80

200 60

150 n = 10 n = 2 n = 5 40 n = 10 100

20 n = 2 50 n = 6

0 0 n = 11 n = 11

300 40 Bicarbonate Sulfate

250 30 200

20 150 n = 10 100 10 n = 2 n = 5 n = 5 50 n = 10 n = 2 0 0 n = 11

CONCENTRATION, IN MILLIGRAMS PER LITER n = 11

Hot Quartz Shale Western springs formations formations hot springs

45 EXPLANATION Silica 40 90th percentile

35 75th percentile n = 10 Median (50th percentile) 30 25th percentile 25 10th percentile n = 11 20 Data value greater than or less than 90th or 10th percentile n = 9 Number of samples 15 n = 2

10 n = 6 5 Hot Quartz Shale Western springs formations formations hot springs

Figure 8. Distribution of concentrations of selected constituents from shallow groundwater and thermal water samples.

Figure 8. Distribution of concentrations of selected constituents from shallow groundwater and thermal water samples. Geochemistry and Comparative Analysis of the Shallow Aquifer and Thermal Water Systems 23

3,000 250 Iron n = 11 Aluminum n = 11

2,500 200

2,000 150

1,500 100 1,000

50 500

0 0 n = 6 n = 10 n = 6 n = 10

1,200 160 Manganese Barium n = 10 140 1,000 120 800 100

600 80

400 60 40 200 20 0 n = 6 n = 6 0 n = 9 CONCENTRATION, IN MICROGRAMS PER LITER n = 10 n = 11

Hot Quartz Shale springs formations formations

20 350 Lithium Strontium n = 6 n = 6 300 15 250

200 10 n = 11 150

5 100 n = 10 n = 10 n = 2 50 0 0 n = 11

Hot Quartz Shale Hot Quartz Shale Western springs formations formations springs formations formations hot springs

Figure 8. Distribution of concentrations of selected constituents from shallow groundwater and thermal water samples.—Continued 24 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

250 sources for, or are likely genetically related to, trace metal occurrence in groundwater throughout the region. Shale formations Wagner and Steele (1985) reported on metal concentra- 200 Quartz formations Hot springs tions in groundwater of the Ouachita Mountains as related to mineral deposits based on 91 samples, predominantly from 150 springs in the area, and their analyses including the following elements: lithium (Li), barium (Ba), strontium (Sr), iron (Fe), 100 manganese (Mn), zinc (Zn), copper (Cu), cobalt (Co), nickel (Ni), lead (Pb), mercury (Hg), and antimony (Sb). Primary 50 mineralization near the study area included Mn deposits (found dominantly in the Arkansas Novaculite), contain- 0 ing up to 1 percent by weight of each of the base metals Co, ALUMINUM, IN MICROGRAMS PER LITER Ni, Cu, and Zn. Four barite (barium sulfate) mining districts

3 4 5 6 7 8 were present near the study area in the lower Stanley Shale pH, IN STANDARD UNITS (Wagner and Steele, 1985). Analyses from Wagner and Steele (1985), for samples from shallow groundwater in an area, which encompassed the study area for this report, are shown Figure 9. pH and aluminum concentrations for groundwater in table 3 and compared to metal concentrations analyzed from samples from study area. samples of shallow groundwater in the study area. Most metal concentrations measured in samples in the shallow groundwater in the study area of this report were in the quartz formations at the time of sampling was 4.4, and within the ranges of metal concentrations from Wagner and 6 of the 11 samples had pH values of 4.4 or less. A detailed Steele (1985) (table 3). Median concentrations for most trace discussion of additional trace metal concentrations related to metals from Wagner and Steele (1985) were between the the major site types is presented in the next section. median concentrations listed for the quartz and shale formations; except for median Co, Ni, and Pb concentrations, which were higher in the analyses from Wagner and Steele (1985) Trace Metals than those from the study area shale and quartz formations. Concentrations for various trace metals were compared However, because the median concentrations for the shale to those documented in previous studies, to ascertain potential and quartz formations are lower than the laboratory reporting mineral sources, and to provide additional information on the level for Pb (10 µg/L) from Wagner and Steele (1985) (table chemistry of the shallow groundwater system as related to 3), the higher median concentration for the data from Wagner that of the hot springs in HSNP. Trace metals provide use- and Steele (1985) may simply be a laboratory reporting level ful adventitious tracers for understanding groundwater flow artifact. In general, because Wagner and Steele (1985) did paths. Important potential sources for trace metals are organic- not differentiate formation type in their data, it is reasonable rich black shales and igneous rocks and fluids. Both of these that median concentrations for most metals would be between sources are present in the Ouachita Mountains. The presence those differentiated by formation type. of economically viable ore emplacements, and the great Median and maximum concentrations for Mn, Ba, Li, historical interest in and development of economic metal ore and Sr were greater in groundwater from the shale formations deposits in the Ouachita Mountains, highlights the abundant than in groundwater from the quartz formations in the study presence of metals and the importance of the role of hydrology area (fig. 8, table 3). Barite deposits occur most frequently and geologic framework in metals movement and distribution in the Stanley Shale and likely account for the higher Ba at local to regional scales. concentrations in samples from shale formations. Higher Mn Early prospect mining for valuable minerals occurred median concentrations were noted in the shale formations as throughout the Ouachita Mountains, including some unsuc- compared to the quartz formations, and demonstrate that Mn cessful gold prospects, from the mid 1800s to 1980. Prospects mineralization may be more prevalent in the shale formations were worked on veins of pyrite and marcasite in 1830, 1915, in the study area. Lithium is associated with Mn ore deposits, and 1916. Traces of lead, zinc, and silver were associated with and the greater Li concentrations coincide with the greater Mn quartz veins and manganese deposits, and one quartz vein in concentrations in the shale formations. Strontium is the most the Arkansas Novaculite and lower Stanley Shale, with associ- commonly substituted element for Ba in the barite structure ated minerals of galena, sphalerite, chalcopyrite, pyrite and (Crecelius and others, 2007) and also is commonly substituted other minerals, was discovered in the 1890s and worked for for calcium in carbonate minerals. Strontium ores are associ- several years (Stone and Bush, 1984). Native copper is found ated with the Stanley Shale in Howard County (Stroud and in veins in the Arkansas Novaculite. The various iron, manga- others, 1969; Wagner and Steele, 1985) and possibly account nese, and other ores in the Ouachita Mountains are potential for the greater Sr concentrations in samples from the shale formations compared to those from quartz formations. Geochemistry and Comparative Analysis of the Shallow Aquifer and Thermal Water Systems 25 4.0 44 <0.4-41 (µg/L) 222 Strontium 91-324 1-2,200 9 35 11.5 2-27 (µg/L) <1-70 13-146 Barium 0.6 5.9 2.3 (µg/L) <1-70 Lithium 2.9-18.3 <0.6-6.8 10 0.67 0.14 Lead (µg/L) <10-37 <0.12-50 <0.12-1.7 ------<0.1 (µg/L) <0.1-2.1 Mercury 9 2.1 2.5 2-39 (µg/L) Nickel 0.05-14 0.46-7.8 5 0.8 1.1 (µg/L) <2-31 Cobalt 0.08-2.5 <0.04-3.9 3.4 1.2 3.2 1-85 (µg/L) Copper <0.4-20 <0.4-446 32 10 18 Zinc 2-144 (µg/L) 0.5-198 <0.3-114 12 21 113 (µg/L) 1-1,220 <0.2-60 42-1,115 Manganese 38 56 127 Iron (µg/L) 12-611 4-2,510 2-5,220 range range range median median median Comparison of trace metal concentrations from Wagner and Steele (1985) to concentrations from samples collected shallow groundwater in the study area. Comparison of trace metal concentrations from Wagner Shale formations Quartz formations Wagner and Steele (1985) Wagner Table 3. Table [ µ g/L, micrograms per liter; <, less than; --, not measured] 26 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

Although median Fe concentrations were greater for samples from the shale formations compared to those from 250 the quartz formations, maximum concentrations (in addition STRONTIUM Shale groundwater to three of the greatest concentrations) were noted in samples 200 from the quartz formations. The range and median concentrations for Cu were greater in the quartz formations than in the shale formations. Copper primarily is associated with Mn ores 150 found in the Arkansas Novaculite (Stone and Bush, 1984; Hot springs Wagner and Steele, 1985), and the greatest concentration of 100 Cu (446 µg/L) was in a groundwater sample from a well completed in the Arkansas Novaculite. 50 Comparisons of trace metal concentrations of samples from the hot springs in HSNP with shallow groundwater in Quartz groundwater the study area are difficult to explain because of the lack of 0 knowledge of mineralization, diagenesis, and other important 400 processes at depth in the deep flow system of the hot springs Shale groundwater in HSNP. However, inspection of water geochemistry of- 350 MANGANESE ten can provide important information in regard to possible 300 processes resulting from water/rock interactions. For example, concentrations of Fe were above the laboratory reporting level 250 of 6 µg/L in all shallow groundwater samples in the study area 200 and ranged upward to 2,510 µg/L; however, Fe concentra- Hot springs tions in all of the samples from the hot springs were less than 150

or equal to 25 µg/L and concentrations in 6 of the 10 samples MICROGRAMS PER LITER 100 from the hot springs were below a laboratory reporting level of 6 µg/L (table 2). Potentially reducing conditions in the 50 Quartz groundwater deep flow system may result in precipitation of FeS, which is 0 stable in reducing conditions, and may account for the lower Fe concentrations. Bell and Hays (2007) investigated isotope 10 Shale groundwater chemistry of the hot springs in HSNP and indicated poten- 9 LITHIUM tial sulfate (SO4) reduction based on sulfur isotope analyses, 8 which provides some evidence that both Fe- and SO -reducing 4 7 conditions exists within the hot springs flow system. 6 Reducing conditions at depth additionally may explain Hot springs elevated Ba concentrations in the hot springs. Barium concen- 5 trations were greatest in the hot springs compared to shallow 4 groundwater samples from the quartz and shale formations. 3 Dissolution of barite (and other minerals containing Ba) along 2 Quartz groundwater a long flow path and transport of dissolved Ba and SO4 into 1 Fe- and SO4- reduction zones at depth can result in FeS forma- 0 tion and would tend to differentially remove Fe and SO4 with 0 20 40 60 80 100 respect to Ba, thus increasing Ba in solution relative to other PERCENT MIXING OF QUARTZ AND SHALE FORMATION WATER metals. A review of general geochemistry (see previous subsec- tion) indicated that the chemistry of samples from the hot Figuregroundwater 10. Mixing from curvesthe shallow for strontium, aquifer in themanganese, study area and and lithium the springs in HSNP may represent a mixture of water from the forhot groundwater springs of Hot from Springs the shallow National aquifer Park. in the study area and shale and quartz formations. Where metal concentrations were the hot springs of Hot Springs National Park. substantially greater in the shale formations as compared to the quartz formations and the hot springs, relations were investigated using mixing curves. Mixing curves were developed As such, the mixing curves provides a method for describ- for Sr, Mn, and Li concentrations, which were greatest in the ing and quantifying the mixture of water from the quartz and shale formations (fig. 10). Mean concentrations for each of the shale formations that is required to produce the chemistry of major rock types were used to construct the mixing curves. For water in the hot springs in HSNP. Various processes, however, each of these metals, the hot springs represent an approximate including evidence for both Fe- and SO4-reducing conditions 40-percent contribution of water from the shale formations and at depth, occur along the flow path and can affect the chemical a 60-percent contribution of water from the quartz formations. fingerprint with respect to some of the chemical constituents. Geochemistry and Comparative Analysis of the Shallow Aquifer and Thermal Water Systems 27

Isotope Geochemistry -27 Linear regression of quartz formations δ2H = 6.96(δ18O) + 8.25 Isotopic data enabled further comparison of groundwater -28 r2 = 0.88 H) in the study area with that of the HSNP system. δ18O and δ2H 2 provide evidence of the distinctive nature of the quartz and -29 shale formation groundwaters and information on sources and recharge characteristics for quartz and shale groundwaters and for the hot springs in HSNP. Quartz formation groundwater -30 isotope data define a strong and distinct trend (fig. 11) with -31 linear regression describing a relation for these data of: Shale formations Quartz formations 2 18 2 (2) Hot springs (data δ HCQ = 6.96δ OCQ + 8.25, r = 0.88 Global meteoric line DEUTERIUM/PROTIUM, PER MIL ( δ -32 collected in 2001; (Rozanski and Bell and Hays, 2007) others, 1993) Hot springs where -33 2 2 -6.0 -5.8 -5.6 -5.4 -5.2 -5.0 -4.8 δ HCQ is the δ H of the quartz formation groundwater, 18 18 OXYGEN-18/OXYGEN-16, PER MIL ( 18O) δ OCQ is the δ O of the quartz formation groundwater, δ and Figure 11. Hydrogen and oxygen isotopic compositions for 2 Figure 11. Hydrogen and oxygen isotopic compositions for r is the regression coefficient. groundwater samples from the study area. groundwater samples from the study area. Comparison criteria for analysis of meteoric water lines hot spring water isotopic composition to current groundwater include slope—which provides information on humidity recharge isotopic compositions difficult. Hot springs isotopic and effects of evaporation in the area where precipitation is compositions are not explained by high-temperature rock- occurring—and deuterium excess, defined asd = δ2H - 8 δ18O, water interaction, evaporation (each of which would cause where d is the deuterium excess. Deuterium excess provides isotopic compositions to be displaced to heavier oxygen isoto- information on humidity at the site of original vapor forma- pic compositions to the right of the plot), or mixing of quartz tion; the global meteoric water deuterium excess is about formation or shale groundwater (giving an intermediate value). 10.8‰ (Clark and Fritz, 1997; Rozanski and others, 1993). These data indicate that the more ancient component of water Groundwater data collected for this study yield a trend that discharging from the hot springs is meteoric water, but mete- plots with a deuterium excess of 14.1‰, a value slightly more oric water that fell to earth under slightly different climatic than 3‰ greater than the global meteoric water line deuterium conditions prevalent more than 4,000 years ago (Bedinger and excess of 10.8‰ (fig. 11), likely indicating that the marine others, 1979); distribution of the data potentially indicates vapor source for area recharge has a lower humidity than the that precipitation occurred under higher humidity condi- global average of 85 percent (Clark and Fritz, 1997). The tions (giving a higher local meteoric water line slope) or that quartz formation groundwater trend exhibits a slope of 6.96, marine source vapors were generated under conditions of slightly less than the global average of 8.1 and essentially lower humidity (higher deuterium excess). Most (five of six) identical to the nearest available local meteoric water line of the hot springs that plot near the quartz formation ground- (for the Fayetteville area—δ2H = 6.95 δ18O + 10; (E. Pollock, water trendline were noted by Bell and Hays (2007) to exhibit University of Arkansas Stable Isotope Laboratory, written relatively large components of locally derived, short flow-path commun., 2009), and indicates lower humidity during precipi- cold-water recharge—further highlighting the distinctive char- tation as compared to the global average. One quartz forma- acter of the ancient, thermal component of the hot springs. tion sample is strongly displaced to the right of the quartz Characterization of Sr geochemistry and analysis of Sr formation groundwater trend, indicating strong evaporation. isotopic signatures were conducted, comparing groundwater Shale groundwater δ2H and δ18O values exhibit no samples from the hot springs in HSNP, shallow groundwater notable correlation. Shale groundwater isotope data all plot in samples in the study area from the quartz and shale forma- the isotopically heavier end of quartz formation groundwater tions, and samples from the western hot springs. Strontium trend, and most shale data plot to the right side of the trend; isotopic data were used in concert with other geochemical data as such, oxygen isotopic compositions are heavy relative to to discern geochemical signatures that can provide informa- hydrogen isotopic compositions for the samples. These obser- tion on recharge origin of the two geothermal waters (western vations indicate that shale formations received recharge during hot springs and the hot springs in HSNP) and whether these a relatively warmer time of year and were subject to greater groundwater systems share any physical hydraulic connection. evaporation than were the quartz formation groundwaters. Strontium (Sr2+ in ionic form ) is a divalent, reactive, and Hot springs isotopic compositions range from δ18O of soluble metal that behaves similarly to calcium in chemical -5.61 to -5.16‰ and δ2H of -30.2 to -28.2‰, showing no dis- interactions and substitutes for calcium in many minerals; of cernable trend. Most of the hot springs data plot near or above particular importance for the HSNP region of the Ouachita the quartz formation groundwater-regression line at the isoto- Mountains is the substitution and occurrence of Sr in carbon- pically heavy end. This position makes any attempt at relating ates and sulfates (such as barite). Like Ca2+, Sr2+ is soluble 28 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs and reactive and participates in water-rock reactions; Sr2+ The quartz mineralogy predominated rocks constitute one is a trace component of most rocks and most groundwater. of the most remarkable groundwaters in the region by exhibit- Strontium ores have been described and mined in the Ouachita ing some of the lowest TDS and most minimally buffered pH Mountains (Wagner and Steele, 1985). Strontium has four values to be found in natural, uncontaminated groundwater naturally occurring isotopes: 84Sr, 86Sr, 87Sr and 88Sr. Stron- (Hem, 1989). The shale formations exhibit a geochemistry tium isotopic compositions are reported as 87Sr/86Sr ratios. typical of regional shale groundwater—high TDS concentra- Strontium isotopic ratios (87Sr/86Sr) are a useful indicator of tions provided by the soluble and reactive mineral matrix water-rock interaction, and a valuable tracer for groundwater of the rock and well buffered pH values. The hot springs in source and movement. As groundwater interacts with rock HSNP gain their distinctive chemical signature along a com- matrix, the water brings Sr into solution and adopts the rock plex flow path through varying formation types. Low ionic Sr isotope signature. Aquifer-framework minerals often have strength rainwater recharges through the minimally soluble distinctive Sr content and specific Sr isotopic signatures that and reactive Bigfork Chert and Arkansas Novaculite, garner- can be picked up by water, enabling identification of recharge ing an initial, baseline Sr chemical and isotopic signature (as sources, transport pathways, and mixing. Unlike oxygen well as a general geochemical and trace metal signature). This isotopes and other light stable isotopes, Sr isotopes are not recharging water reacts with more soluble shales further along subject to fractionation, and Sr isotopic signatures are deter- the flow path, altering the signature. This process is not purely mined only by Sr input (Clark and Fritz, 1997). Sr geochemi- a mixing process but a dissolution/reaction-based geochemical analysis of Hot Springs recharge area rocks provided very cal evolution, and a degree of understanding of the relative specific information on thermal water sources, pathways, and contributions along the two distinctive flow path segments can relations. Fifty-two groundwater samples were characterized be derived from an assessment using simple, hyperbolic mix- for Sr geochemistry, including 10 samples from hot springs ing analyses. For isotopes, mixing of two end members does in HSNP, 16 Bigfork Chert, 10 Arkansas Novaculite, 4 Hot not result in a number of mixtures that fall on a straight line in Springs Sandstone, 9 Stanley Shale, 2 Silurian/Ordovician concentration-isotopic ratio space. Instead, the isotope ratio of undifferentiated shales, and 1 Womble Shale sample. the mixture is weighted by the concentration of Sr in each end Groundwater Sr concentration and isotopic data from member. More generally, the weighting of ratios in the mixture the hot springs in HSNP, quartz formations, and shale forma- is controlled by the denominator of the ratio, in this case 86Sr, tions indicate three end-member data groupings (fig. 12): (1) causing mixtures to plot on a hyperbola whose curvature is high isotopic ratio (ranging from about 0.712 to 0.723), low dependent upon the difference in concentration of the two end Sr concentration quartz rocks; (2) low isotopic ratio (ranging members. from about 0.708 to 0.712), low Sr concentration quartz rocks, To explore the respective contributions of the quartz and (3) mid-range isotopic ratio, high-concentration shales; and shale formations and whether more complex controls and one intermediate grouping—the hot springs in HSNP. The are important in setting hot springs geochemistry, a simple quartz formations of end-member groupings 1 and 2 exhibit hyperbolic mixing equation (Faure, 1986) was used to model a narrow range of concentration, 0.4 to 42 µg/L, and a very hot springs Sr geochemistry. Sr concentration and isotopic broad range of isotopic compositions—defining the isotopic composition end members for the quartz formations from a maximum of 0.7224 to the minimum of 0.7082. The shale 0.724 grouping, group 3, exhibits a relatively broad range of con- 0.722 End-member 1 Shale formations centrations (106 to 385 µg/L), and a narrow range of isotopic group Quartz formations compositions from 0.7085 to 0.7138. Samples from the hot 0.720 Hot springs springs exhibit very tight ranges for concentration and isotopic Western hot springs composition (96 to 108 µg/L and 0.712, respectively). The hot 0.718 springs samples are generally intermediate to the end member 0.716 1 and end member 2 groups in isotopic composition but have considerably greater concentrations (fig. 12 and fig. 8); hot 0.714 End-member 3 springs samples are intermediate to group 1 end member and 0.712 Greer group

group 3 end member for isotopic composition and Sr concen- STRONTIUM-87/STRONTIUM-86 tration. Isotopic composition-concentration relations indicate 0.710 Bratton that the geochemistry of hot springs in HSNP is controlled 0.708 End-member 2 predominantly by contribution from rock-water interaction group 0.706 with group 1 quartz formations and group 3 shale formations. 0 100 200 300 400 500 Sr chemistry and isotopic composition corroborate general STRONTIUM CONCENTRATION, IN MICROGRAMS PER LITER geochemistry and trace metals results. Figure 12. Strontium concentration and strontium isotope values Figure for12. groundwater Strontium samples concentration from the and study strontium area. isotope values for groundwater samples from the study area. Physical and Chemical Characteristics of Thermal Waters in the Ouachita Mountains 29 composition end-member values of 20 µg/L and 0.7185 Physical and Chemical Characteristics (quartz formation end member 1) and 265 µg/L and 0.71083 (shale formation end member) were input. The hot springs plot of Thermal Waters in the Ouachita on the hyperbolic mixing line derived from these quartz and shale formation end member Sr values—probably indicat- Mountains ing that simple dissolution processes predominate over more complex geochemical reactions, which would result in dis- This section describes the physical and chemical char- placement of the hot springs from the mixing line. The mixing acteristics of the various sources of thermal water outside of model analysis indicated that the Sr geochemistry of the hot the HSNP boundaries and provides a conceptual model for the springs results from an approximately 35-percent contribu- generation of thermal water at these various sites. In exam- tion from the shales and a 65-percent contribution from the ining the characteristics of thermal waters in the Ouachita quartz formations. Thus, as the recharge waters move out of Mountains, it is useful to review the key components of the the quartz formations and into shales along the thermal flow most widely accepted model accounting for the source, flow path, the original quartz formation geochemical signature is history, and hydrogeologic framework for the hot springs in overprinted to a considerable degree by the shales, resulting in HSNP. The combined thickness of the formations in the study a mixed signal. Evidence of this shale geochemical influence area ranging from the Bigfork Chert, which has been identi- also is seen in major ion and trace metal chemistry as dis- fied as the predominant formation recharging the hot springs cussed earlier, and similar mixing percentages to that calcu- flow system, upward to and including the youngest exposed lated for strontium isotopic data were shown on mixing curves formation in the study area, the Stanley Shale, is greater than for the trace metals Sr, Mn, and Li (fig. 10). 10,000 ft, not including the 1,500-ft Womble Shale. The The highly fractured nature of all formations exposed folding and extreme dips of the formations in the study area at the surface of the study area and the Ouachita Mountains additionally plunge these formations to great depths. Bedinger in general, creates a hydrologic environment favoring a high and others (1979) presented a model that showed a required degree of mixing and integration of different flow lines. depth of 4,500 ft to 7,500 ft at geothermal gradients of 0.6°C Additionally, these fractured strata present an extremely chal- to 1.0°C per 100 ft, respectively, to attain maximum projected lenging environment for drilling and, ultimately, sampling source temperatures of 63°C to 68°C at depth as calculated of isolated zones. This combination of effects results in great from silica concentrations in the hot springs. These depths difficulty in acquiring samples from discrete zones, such as is are easily achieved in view of the formation thicknesses and desired from the newly discovered thermal zones at the Brat- regional dips of 40 degrees and greater. With mean tempera- ton and Greer sites. It should be noted that optimally, isolated tures of approximately 62°C for the hot springs (Bedinger and samples of the thermal waters at the Bratton and Greer sites others, 1979), the water must rise fairly rapidly along faults would be the most desirable and most definitive for geochemi- to prevent excessive cooling on the waters path to the surface. cal signature interpretation; however, isolated thermal water The hot springs emerge at the nose of a southwest plunging samples proved impossible to collect within constraints of the anticline along a major mapped thrust fault (fig. 2). A review project. Thus, all of the samples available for interpretation of the geology outside the area of the hot springs revealed that were mixed samples. Interpretations using mixed samples similar geological conditions are replicated throughout the are informative and useful and are in fact a common part Ouachita Mountains. Similar to the hot springs, the Bratton of chemical hydrologic study, and, as such, interpretations and Greer well sites are on or near the nose of a southwest presented here are useful and sufficient to addressing the plunging anticline, and faulting is near to or directly bisect- questions posed, although initial objectives were to acquire ing the locations of the well sites (fig. 2). Hence, the model discrete-zone samples. for development of the hot springs in HSNP serves as an The newly discovered thermal water sites, Bratton and analog for explaining the occurrence of thermal water at these Greer, did not exhibit notable similarity with the hot springs locations. in HSNP with respect to Sr chemistry, but look geochemically like the groundwater of the local outcropping formations: the Bratton Site Characterization Stanley Shale in the case of Bratton and the Bigfork Chert for Greer. Median Sr concentration and isotopic composition Several exercises were conducted to investigate the for the two wells at the Bratton site is 327 µg/L and 0.7103, nature of the thermal component of water at the Bratton site, respectively. Median Sr concentration and isotopic compo- including pumping events and geophysical logging for the sition for the three wells at the Greer site are 4 µg/L and Bratton domestic well and a monitoring well installed near 0.7106, respectively (fig. 12); these waters differ markedly the domestic well. On March 13, 2006, following the initial from the hot springs in HSNP. The two very different Sr isoto- report of the discovery of thermal water in the domestic well, pic and geochemical signatures observed for the new thermal USGS personnel recorded well-water temperatures during a water sites do not provide evidence of any direct hydraulic 2-hour pumping event. A maximum of 31°C was measured connection with the springs in HSNP. initially, followed by a steady decrease in temperature with the 30 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs temperature never dropping below 21.7°C during the pumping increase to a temperature of 15.5°C (60°F) near a depth of 140 period. Mean temperature for shallow groundwater in the area to 145 ft; this is interpreted to be the zone of thermal water is approximately 17°C. This simple exercise provided evi- contribution. Temperatures then decreased to a minimum of dence of a weak, low-yield component of thermal water flow 15.0°C (59°F) at approximately 155 ft, where it remained to for which the thermal signature was dampened by a greater the completion depth of 165 ft. The increase in temperature cold-water contribution during pumping. The thermal water occurred at a similar depth as that recorded in the domestic likely moved through a discrete fracture set in communication well, and the monitoring well was screened from 130 to 150 ft with a deep flow path, while the cold-water component was in order to isolate the zone of interest. likely supplied by fracture sets representing short flow paths in Several short-term (less than 3 hours) pumping episodes direct communication with the surface. were conducted on the well with approximate yields of 1.5 Geophysical borehole logging was conducted on Janu- gal/min; pumping rates greater than 1.5 gal/min were not sus- ary 29, 2007, which included a temperature and flow profile tainable. Hydraulic connection between the domestic well and obtained using an electromagnetic (EM) flowmeter. The tem- the monitoring well were evident from water-level decreases perature profile indicated a uniform temperature of approxi- of greater than 1 ft observed in the monitoring well when the mately 17.8°C (64°F) to approximately 100 ft, at which point domestic well was being pumped. Although the domestic well temperatures began to increase to a maximum temperature of was not used as a continuous domestic supply, the owner did 21.1°C (70°F) at a depth of approximately 130 ft (fig. 13a). intermittently use the well, resulting in sudden drops in water Temperature rapidly decreased to 18.3°C (65°F) at a depth levels observed in the real-time data for the monitoring well. of approximately 150 ft, where it remained relatively con- Although the monitoring well was screened over the interval stant to the completion depth of 205 ft. Numerous fracture (130–150 ft) demonstrating a slight increase in tempera- sets were noted from review of the acoustic televiewer log ture, which correlated to a similar depth for the temperature (dark sinusoids on fig. 13a) to depths of approximately 160 increase in the domestic well, the real-time temperature mea- ft, at which point a lack of apparent fractures was noted to surements of water in the monitoring well never rose above the completion depth of 205 ft. The fact that the temperature 18.7°C for the period of measurement, considerably less than drops during pumping of the well suggests that the dominant the maximum temperatures observed for the Bratton domestic component of flow is a cold-water contribution, and that the well. However, these measurements were made approximately warm-water component is a weaker flow zone. A flow profile 20 ft below the water surface, which was the location of the was generated using an EM flowmeter under ambient and transducer that records water level. pumping conditions in the well. The EM flowmeter profile A 3-hour pumping event was conducted on the monitor- showed the greatest flow entering the well between 90 to 120 ing well on May 20, 2008, and revealed only slightly greater ft and a negative component of flow from 130 to 140 ft. As temperatures ranging from 19.1°C to 19.6°C. Although these such, individual flow components upward to 0.3 gal/min were temperatures were slightly higher than the real-time tempera- moving into the well bore from 90 to 120 ft; whereas below tures measured at the transducer approximately 10 ft below the this point, flow was moving out into the formation. A profile water surface for the period of measurement, the temperatures also was generated for specific conductance, which showed were far below the temperatures recorded in the domestic an increase in specific conductance within the same zone as well, and indicate a much weaker component of thermal flow the temperature increase. The increase in specific conductance to the monitoring well than in the domestic well. Despite the indicates that the thermal component of flow is enriched in proximity of the monitoring well to the domestic well (20 ft) dissolved solids relative to other flow zones in the well and and careful, deliberate drilling and completion, the monitor- provides evidence that this component of flow comes from a ing well was not as effectively connected to the thermal-water longer flow path. fracture sets providing thermal flow. Temperature of water in A monitoring well was drilled on June 27-28, 2007, the monitoring well was noted to increase from November approximately 20 ft from the original Bratton domestic well. 2007 through March 2008, contrary to the response of other The intention of drilling the monitoring well was to isolate shallow wells in the area, which exhibit a decrease in water the zone of thermal water contribution for estimating flow temperature during the cold season from November through and for sampling and analyzing water quality. The well was March (fig. 14). Because water levels are higher during the fall drilled to a depth of 165 ft. Driller-estimated flows of 90 gal/ and winter months, which coincides with the period of higher min occurred at approximately 80 ft with no appreciable gain water temperature, these data are interpreted to indicate that in flow below that depth. On June 28, 2007, geophysical log- higher heads in the recharge area for the deep flow system ging was performed on the monitoring well, which was an providing the thermal water moves a larger component of open-hole completion at the time of logging (fig. 13b). A high thermal water through the shallow aquifer system during the temperature of 21.1°C (70°F) near the surface was the result of wet season. Temperatures began to drop in the observation radiant warming with the high summer temperatures. Tem- well after March 2008 during the spring and summer months, perature dropped steadily to approximately 15.0°C (59°F) at only to begin to increase once again in November 2008. a depth of approximately 100 ft and remained at this tempera- These data indicate that during the wet season, water-level ture from 100 to 120 ft, at which point temperatures begin to increases and the greater pressure heads in the higher elevation Physical and Chemical Characteristics of Thermal Waters in the Ouachita Mountains 31

19.0 25

18.5 27

18.0 29

30 TEMPERATURE, IN DEGREES CELSIUS

17.5 31 DEPTH TO WATER, IN FEET BELOW LAND SURFACE

Depth to water Water temperature 32

17.0 33 Oct. 1 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1 May 1 June 1 July 1 Aug. 1 Sept. 1 Oct. 1 Nov. 1 Dec. 1 Jan. 1 Feb. 1 Mar. 1 Apr. 1 2007 2008 2009

Figure 14. Hydrograph of temperature and water level in the monitoring well at the Bratton site. thermal-water recharge areas on exposures of the Bigfork In groundwater systems, Fe occurs in one of two oxida- Chert and Arkansas Novaculite result in a rapid pressure tion states: reduced, very soluble divalent ferrous Fe (Fe+2) response through the system and increased flows in the down- or oxidized insoluble trivalent ferric Fe (Fe+3) (Appelo and gradient end of the flow path. The observations from study Postma, 1999). Under reducing conditions typical of deep of the Bratton site lead to an understanding that has broader flow-path waters, Fe+2 predominates, is very mobile, and is application and implications across the Ouachita Mountains carried with migrating groundwater. Under shallow groundwa- for the occurrence of thermal flow systems. ter conditions where free oxygen is available, Fe is oxidized to Temperature was monitored in the Bratton domestic well the insoluble Fe+3 form, creating Fe oxyhydroxide precipitates at the Bratton site during a 3-hour and 20-minute pumping (such as hematite and goethite). Precipitation of Fe oxides event on May 20, 2008. Continuous temperature monitoring and other minerals, including carbonate and silica minerals, yielded valuable information regarding the zone of thermal at these deep flow path/shallow flow path redox interfaces contribution in the domestic well. Water temperatures peaked is common for hot springs, and these precipitates often are at a value of 36.4°C, 4 minutes into the purging event, and observed to restrict flow, sometimes temporarily—ending began dropping thereafter to a low of 18.8°C after 1 hour and when plugs are pushed out by changes in head or other agita- 15 minutes of purging at 2 to 3 gal/min. This latter result was tion and sometimes permanently—resulting in repositioning in contrast with the earlier tests on March 13, 2006, when of spring emergence points, such as is common in HSNP and temperature never dropped below 21.7°C. Interestingly, dur- other thermal springs (Bryan, 1922; Yeatts, 2006). The slug of ing the May 20, 2008 test, after water temperature dropped Fe-precipitate laden water may have originated as a fracture- to the minimum of 18.8°C, a slug of red, Fe-precipitate rich, plugging Fe oxyhydroxide precipitate at the redox interface turbid water appeared in the collection bucket over a period of of the shallow subsurface of the well, temporarily sealing a couple of minutes, after which the water returned to a clear fractures conveying thermal water and reducing thermal-water condition. Following the breakthrough of the turbid water, flow and temperature of water in the well. The lowering of water temperatures began to rise to a maximum of 23.8°C and the water table and surging and concomitant turbulence in the remained above 22°C for the remainder of the purging event. well during pumping are thought to have unseated the weakly A red, Fe-oxyhydroxide precipitate was noted to have settled bound Fe-precipitate plug, resulting in an increase in the ther- at the bottom of the discharge monitoring bucket. mal flow component and increased temperatures. 32 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

Greer Site Characterization depths with no apparent confinement or isolation between the water-producing zones. The appearance of thermal water at the Greer site was Continuous monitoring of water level and temperature in discovered on June 6, 2007, during a well reconnaissance in the 155-ft well from October 2007 and March 2009 showed the early phase of the study. The site is approximately 3 miles only a slight increase in water temperature from 16.6°C to from the hot springs in HSNP. Water periodically is pumped 16.8°C. Continuous monitoring of the 185-ft domestic well from the 185-ft well into a large holding tank for domestic from June 2008 to March 2009 also showed little variation in use. However, a spigot available at the well head is used for temperature with only a slight increase from 15.8°C to 16.0°C. watering vegetation at the home during the spring and summer Purging of all three wells during a water-quality sampling months, and water is delivered directly from the pump into the event on September 5-7, 2007, resulted in final purged-water spigot and hose. The owner reported that during the first use of temperatures that ranged from approximately 18.0°C to the water from the spigot in May 2007 the water felt warm to 19.0°C for all three wells. Elevated temperatures have not the touch and remained warm during a long watering episode, been reported by the well owner since the original measure- though this phenomena had not been observed in the previous ments in June 2007, and continuous and periodic measure- 5 years use of the well. ments have not revealed elevated temperatures in any of the During this initial visit, USGS personnel pumped water three wells. from the spigot directly into a bucket and measured tem- There are several explanations for the appearance and perature over a 30-minute period. A maximum temperature temporal nature of the occurrence of a thermal component of of 26.6°C was attained during the early stage of pumping, water in the domestic well at the Greer site. An active quarry followed by a steady decline to 22.4°C after 5 minutes of with associated blasting is approximately 0.75 mile east of the pumping, with temperature fluctuating between 24°C and Greer site. Changes in the hydrologic system associated with 25°C throughout the remainder of the purging period. Vari- dewatering and removal of fine-grained material from frac- able temperatures during the purging event indicated mixing tures at the quarry may have altered groundwater flows near of both cold water and thermal water contributions to the the quarry. The generation of fractures greater than a few tens well bore, similar to that at the Bratton site. The Greer site of feet from blasting activities is not anticipated from the lit- presented numerous difficulties in determining the zone of erature review on the effects of blasting, and minor changes in contribution for the thermal component of water and the tem- flow regimes by dislodging sediment-plugged fractures during poral and variable nature of fluctuating temperature signatures blasting may become quickly sealed over time by fine-grained in the well. Geophysical logs showed the well to be cased to material and chemical precipitates. the completion depth of 185 ft—through the likely zones of Another explanation for the temporal nature of the occur- interest. As such, no information could be obtained from the rence of a thermal component of flow in the domestic well at geophysical logging to determine zones of flow, temperature the site may be that a large influx of cold water during high profiles, stratigraphic and geologic data, and other data that water-level periods associated with shallow recharge may might elucidate the thermal component of flow to the well. dampen the thermal component of flow. However, water-level Installation of a new, open-borehole monitoring well next increases and decreases over the period of continuous moni- to the Greer domestic well was attempted for the purpose of toring of two wells at the site did not result in changes in water providing access to the subsurface for borehole geophysical temperature. logging, determining the zone of thermal water contribution, Similar to the discovery of Fe-oxyhydroxide precipitation and providing for discrete zone sampling. Two well-instal- and apparent sealing of fractures within the thermal zone at lation attempts proved problematic as a result of continual the Bratton site, another explanation is that fractures associ- caving and collapse in the highly fractured Bigfork Chert. ated with transport of thermal water in the domestic well at the The borehole was advanced to a depth of 120 ft, whereupon a Greer site may have sealed with time. This explanation, how- constant infilling of rock material prevented further penetra- ever, does not fit with the timing as the thermal-zone fractures tion, even after numerous attempts to clean the borehole and would most likely be open when first encountered during the continue drilling. Decisions were made to complete the well drilling process, with sealing over time resulting from changes at a depth of 120 ft, and begin a new hole. A second borehole, in the redox conditions initiated by the presence of the well nearer to the domestic well, was advanced to a depth of 155 ft with the associated surface communication and continual before continual collapse of rock material forced drilling ter- influx of oxygen into the borehole. mination and completion at this depth. No attempt was made The most likely scenario is that the initial discovery of to log either open hole because of concerns of losing costly the apparent thermal water component at the Greer site coin- logging equipment in the unstable borehole. With comple- cided with a 3-year drought, more extreme than any in the past tion of the two additional wells, three wells were available for 50 years (which covers the period of the existence of the Greer monitoring of temperature and water quality at depths of 120 well) that ended as this study was starting (National Climatic ft, 155 ft, and 185 ft. Similar water levels in all three wells Data Center, 2009). Data from the National Climatic Data demonstrated the hydraulic connection between the various Center (NCDC) showed 26 of 36 months in drought status Implications and Conceptual Model for Thermal System Analogs in Ouachita Mountains 33 from the NCDC Palmer Drought Severity Index data with Table 4. Comparison of temperatures from thermal water in 12 months of moderate to extreme drought. This period of Ouachita Mountain region to shallow groundwater in study area. minimal local cold-water system recharge may have enabled a short-lived ability to detect the small thermal-water contribu- Approximate tion that was subsequently overridden by increased cold-water Water source temperature recharge as the drought ended. (degrees Celsius) Shallow groundwater 17 Thermal Springs Outside of Study Area Hot springs of Hot Springs National Park 62 Bratton site 34 A review of Ouachita Mountains region literature identi- Greer site 25 fied four thermal springs outside of the study area, three of which are in Montgomery County and one in Pike County (fig. Little Missouri Spring 23 15). These springs range from approximately 30 to 50 miles from the hot springs in HSNP. Three of the springs, Caddo Gap, Redland Mountain, and Little Missouri Springs, were Implications and Conceptual Model for listed in Bryan (1922). Temperatures for Caddo Gap and Little Thermal System Analogs in Ouachita Missouri Springs measured in 1915 were 34.5°C and 23.3°C (Bryan, 1922), respectively, and measurements made in June Mountains 2008 revealed similar temperatures of 33.8°C and 23.4°C, respectively. For comparison, surface-water temperatures in At the 1:500,000 scale represented by the Arkansas State the Caddo River and the Little Missouri River at the time of geologic map (Haley, 1976), thermal water generally is located sampling were 21.8°C and 19.5°C, respectively. Caddo Gap in similar geologic settings, although the thermal waters Spring issues from small fractures in the bedrock directly emanate from three different formations: the Stanley Shale beneath the surface of the Caddo River. Little Missouri Spring (Bratton site) including the Hot Springs Sandstone Member issues as a series of seeps in a sand and cobble bank approxi- of the Stanley Shale (the hot springs in HSNP), the Bigfork mately 5 ft from the Little Missouri River. Similar to measure- Chert (Greer site), and the Arkansas Novaculite (Caddo Gap, ments made at the hot springs in HSNP from the late 1800s to Redland Mountain, Barton, and Little Missouri Springs) (fig. the present date, which have showed no discernable long-term 15). These thermal waters tend to be found along the nose trend (Bedinger and others, 1979; Bell and Hays, 2007), the of plunging anticlines and closely align with mapped thrust measured temperatures at Caddo Gap and Little Missouri faults where detailed regional geologic maps were available. Springs have shown relatively stable temperature with time. Because shallow flow systems have been identified as being The lack of temperature variation in the hot springs in HSNP restricted to the regional surface watershed boundaries, the and thermal springs as far as 50 miles from the hot springs deep flow systems for thermal water are likely unique to the indicates that these springs arise from hydrogeologic systems local hydrologic and geologic setting and represent an analog that provide continuous and stable flow of thermal water and to the geologic framework model for propagation of ground- also indicates a degree of comparability between the hydro- water to the hot springs in HSNP. logic systems providing for these disparately located occur- Concerns related to pirating of water from the hot springs rences of thermal water. flow system in HSNP because of blasting near the Brat- A comparison of temperatures was made between the ton and Greer sites were not supported by the data gathered various thermal waters observed in this study, including the for this study. No physical hydraulic connection was found hot springs in HSNP, and shallow groundwater in the study between these thermal sites and the hot springs at HSNP, area to evaluate the spatial differences in temperature (table 4). and the waters exhibited different geochemical composition Although 30 miles from the hot springs in HSNP, Caddo Gap and different isotopic and trace metal signatures. The newly Spring is the same temperature as the water from the domestic discovered thermal sites (Bratton and Greer sites) appear to well at the Bratton site, which is only 5 miles from the hot be low-volume systems in comparison with the springs in springs in HSNP. Similarly, Little Missouri Spring is 50 miles HSNP, and the sites are separated by distance and geological from the hot springs in HSNP and has a similar temperature and hydrological boundaries. Thermal-water sites across the to water in the domestic well at the Greer site. The spatial Ouachita Mountains appear to represent discrete systems that distribution of similar temperatures in similar geologic set- are associated with a specific set of hydrologic and geologic tings (nose of plunging anticlines; fig. 15) indicates that other conditions, which occur at numerous locations across the thermal waters may exist throughout the Ouachita Mountains region and are viewed as analogs to the hot spring flow system within these settings that have yet to be discovered, at or near in HSNP rather than connected components of the same land surface. hydraulic system. Several variables, including permeability 34 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

(

( (

( Bratton

SALINE ! ( ( ea ! Obf (fig. 1) TION Smb 93° Greer

Study a r ( ( omble Shale

Hot springs

Geology modified from Haley, 1976 ( Stanley Shale Arkansas Novaculite Bigfork Chert W Missouri Mountain Shale and Blaylock Sandstone Other formations outside study area ! ! ! ! ! ! ! ! ! Ow Geology Anticline— Shows trace of crest plane and direction of plunge. Dashed where approximately located Inventoried thermal water EXPLAN A (

( ! Ms Ow ( Obf Smb MDa 93°10' MDa Ms MILES 10 KILOMETERS 10 5 93°20' MDa 5 SPRING GARLAND HO T River 0 Y 0 Ms CLARK

Caddo MONTGOME R 93°30' PIKE

( (

( Ow Caddo Gap Spring River

! e ( in al

S River ( study area study area Hot Springs ( Hot Springs ta chi ( Oua

! ( ( ARKANSAS ! Barton Spring ! 93°40' Spring Spring Smb Caddo Gap Ouachita Mountains Mountains Obf Map area Little Spring Location map Missouri Ms Redland Mountain Spring

River !

( River

93°50' Missouri ( ( ARD

Caddo MDa Little HO W Little Missouri Spring Obf Ms

ransverse Mercator projection POLK Geology associated with thermal sites in Ouachita Mountains region of western Arkansas. Base from U.S. Geological Survey digital data, 1:25,000 Universal T Zone 15 34°40' 34°30' 34°20' 34°10' Figure 15. Summary 35 along faults, downward and upward flow rates, pressure head surface runoff from the recharge area by installation of imper- at the recharge areas, plunging depth of the combined forma- meable surfaces (road cuts, parking lots, buildings, and other tions, and various other geologic and hydraulic factors affect manmade structures), removing soil and regolith, dewater- the amount and associated temperatures of thermal waters ing activities, changes in vegetation cover type and density, emerging at the surface in the various thermal system loca- and additionally by changing, blocking, or opening fractures tions. Although the mechanism for generation of thermal water carrying water to depth in the host rocks within the recharge appears to be very similar for the described thermal water area. Decreased recharge will result in decreased hot spring occurrences throughout the Ouachita Mountains, the tempera- discharge—decreasing the quantity of water available—and ture of these waters will vary at the land surface where they decrease average temperature of the springs. Because water issue as springs or along subsurface fractures intersected by flowing from the hot springs in HSNP includes a small, shallow wells. but important, component of cold-water recharge (surficial This conceptual model of the occurrence of thermal recharge proximal to the hot spring discharge area) (Bell and waters outside of HSNP is consistent with the observations Hays, 2007), any decrease in the rate of geothermal discharge and information gathered from data collected at the Bratton from the springs also would be accompanied by an immediate site. Although temperatures were lower in the monitoring well decrease in the temperature of the hot springs that would be than in the domestic well at the Bratton site, isolation of the much exacerbated during the wet season. fracture zones contributing the thermal component of flow to the monitoring well provided important data leading to implications regarding recharge in the deep flow system. During the wet season, water-level increases and the greater pressure Summary heads in the higher elevation thermal-water recharge areas on exposures of the Bigfork Chert and Arkansas Novaculite result A study was conducted by the U.S. Geological Survey in in a rapid pressure response through the system and increased cooperation with the Arkansas State Highway and Transporta- flows in the downgradient end of the flow path. This increased tion Department to characterize the contributing hydrogeol- contribution of hot water resulted in the increased water ogy of newly discovered thermal water in a private domestic temperature in the well. In exhibiting this type of response, well 5.5 miles east of Hot Springs National Park, Hot Springs, the Bratton site thermal flow system manifests as an impor- Arkansas, and to determine the degree of hydraulic connectiv- tant analog for the hot springs in HSNP, and observations and ity between the thermal water in the well and the hot springs conclusions for the Bratton system provide important input in Hot Springs National Park. The initial water temperature in for understanding and ultimately protecting and managing the the well, which was completed in the Stanley Shale, was mea- HSNP hot springs system. An important and useful difference sured at 33.9°C and never dropped below 21.7°C after 2 hours is that the Bratton system represents a smaller system with of pumping. A second well with a thermal water component more limited recharge and less integration and averaging of at a site 3 miles from the hot springs in Hot Springs National flow input. Because of this, the system is more easily per- Park was discovered during a reconnaissance of the area. The turbed and smaller or shorter-term effects on the system can well was completed in the Bigfork Chert and field measure- be more easily and quickly discerned than for the much larger ment of well water revealed a maximum temperature of HSNP system. Thus, seasonal influence of recharge-area head 26.6°C. Mean temperature for shallow groundwater in the area changes were observed in the Bratton system. It should be is approximately 17°C, and the occurrence of thermal water noted that discharge and temperature of individual springs in in these wells raised questions and concerns with regard to the HSNP have never been monitored in tandem on a long-term timing for the appearance of the thermal water. basis, making recognition of seasonal and other influences on Concerns regarding the possible effects of blasting the HSNP hot springs flow system very difficult. associated with highway construction near the first thermal Similar to the Bratton system, the hot springs in HSNP well necessitated a technical review on the effects of blast- are recharged by infiltrating rainwater through higher eleva- ing on shallow groundwater systems. Results from available tion exposures of Bigfork Chert and Arkansas Novaculite, studies indicate that propagation of new fractures near blasting and any changes in the recharge/runoff ratio in the recharge sites is of limited extent. Safe levels of ground vibration from area could affect the discharge and temperature at the hot blasting range from a peak particle velocity of 0.5 to 2.0 in/ springs. Changes to surface and subsurface physical proper- sec for residential structures, and this range in vibrations ties of hydraulic conductivity, porosity, storage, and fracture appears to be safe for protection of groundwater wells rang- and flow-path connectivity and orientation can cause changes ing from 300-5,100 ft from blasting sites based on a review in system hydraulic head and pressure. As such, alterations of of pertinent studies. Although a possibility of rock collapse the hydraulic characteristics of the land surface resulting from exists for uncased wells completed in highly fractured rock, land-use changes can affect changes in recharge and resultant as the vibrations can result in rock collapse, the propagation pressures in the deep flow system that will propagate quickly of newly formed large fractures that potentially could damage along the flow path with rapid effect on the flow at the hot well structures or result in pirating of water from production springs. This situation could arise as a result of a diversion of wells appears to be of limited possibility based on review of relevant studies. 36 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

Characteristics of storage and secondary fracture porosity hot spring flow path with the initial, low ionic strength charac- were interpreted from yields observed in individual wells com- ter inherited from shallow recharge through quartz formations pleted in the Bigfork Chert and Stanley Shale; from hydro- being modified by passage through the shales at depth. Mixing graphs produced from continuous measurements of water curves for lithium, manganese, and strontium concentrations, levels in wells completed in the Arkansas Novaculite, the which were greatest in the shale formations, revealed that the Bigfork Chert, and Stanley Shale; and from a potentiometric- hot springs represent an approximate 40-percent contribution surface map constructed using water levels in wells through- of water from the shale formations and 60-percent contribution out the study area. The data gathered from these three separate of water from the quartz formations. exercises corroborate findings from earlier reports that fracture Oxygen, hydrogen, and strontium isotopes were included porosity is much greater in the Bigfork Chert relative to that in in the geochemical analysis of study area groundwater. Char- the Stanley Shale. At the scale that the potentiometric map was acterization of strontium geochemistry and analysis of stron- developed, potentiometric contours tend to mimic topography tium isotopic signatures were conducted, comparing ground- with no discernable influence on shallow groundwater flow by water samples from the hot springs in Hot Springs National faulting or other geological structural features. The Bigfork Park, shallow groundwater samples in the study area from the Chert generally is located along the hinge line of the topo- quartz and shale formations, and samples from two western graphically higher anticlinal ridges, and the higher groundwa- hot springs 30 to 50 miles west of Hot Springs in Montgomery ter elevations from the potentiometric-surface map indicate County. Strontium isotopic data were used in concert with that exposures of Bigfork Chert serve as major recharge areas other geochemical data to discern geochemical signatures that for the regional shallow groundwater system in the study can provide information on recharge origin of the two geo- area. Flow lines developed on the potentiometric-surface map thermal waters (western hot springs and the hot springs in Hot indicated flow from these elevated recharge areas along and Springs National Park) and whether these groundwater sys- into streams within the synclinal valleys formed on exposures tems share any physical hydraulic connection. Mixing model of the Stanley Shale. Therefore, flow follows topography, analysis indicated that the strontium geochemistry of the hot and there is no evidence of interbasin transfer of groundwater springs results from an approximately 35-percent contribution within the shallow flow system. from the shales and a 65-percent contribution from the quartz A groundwater-sampling program was conducted to formations, similar to that found from trace-metal analysis determine water quality and geochemistry of shallow ground- mixing curves. The newly discovered thermal water sites water from the various exposed formations in the study area did not exhibit notable similarity with the hot springs in Hot for comparison to the geochemistry of the hot springs in Springs National Park with respect to strontium chemistry, but Hot Springs National Park. Values of pH and conductance look geochemically like the groundwater of the local outcrop- of groundwater were measured in 25 wells during an initial ping formations. The two very different isotopic and geochem- reconnaissance of the study area. The median pH was 7.3 for ical signatures observed for the new thermal water sites do not wells completed in shale formations and 4.5 for wells com- provide evidence of any direct hydraulic connection with the pleted in quartz formations (novaculite, chert, and sandstone hot springs in Hot Springs National Park. formations). Low specific-conductance values were measured Plotting all thermal sites on the State geologic map for wells completed in the quartz formations; all but one site showed that thermal water generally is located in similar exhibited specific-conductance values of less than 50 µS/ geologic settings, although the thermal waters emanate from cm, with a median of 30 µS/cm, whereas wells completed in three different formations: the Stanley Shale including the Hot shale formations ranged from 97 µS/cm to 490 µS/cm with a Springs Sandstone member of the Stanley Shale, the Bigfork median value of 290 µS/cm. Based on results from the recon- Chert, and the Arkansas Novaculite. The thermal waters tend naissance, 15 shallow wells and 2 cold-water springs were to be found along the nose of plunging anticlines and closely sampled to represent the water quality and geochemistry for align with mapped thrust faults where detailed geologic maps the shallow groundwater system. For the quartz formations, were available. Because shallow flow systems have been iden- total dissolved solids concentrations, as expected from the tified as being confined within the regional surface watershed low specific conductance values, were very low with a median boundaries, the deep flow systems for thermal water are likely concentration of 23 mg/L, whereas the median concentration unique to the local hydrologic and geologic setting and repre- for groundwater from the shale formations was 184 mg/L. Ten sent an analog to the geologic framework model for propaga- hot springs in Hot Springs National Park were sampled for tion of groundwater to the hot springs in Hot Springs National the study. Several chemical constituents for the hot springs, Park. Concerns related to pirating of water from the hot including pH, total dissolved solids, major cations and anions, springs in Hot Springs National Park because of blasting near and trace metals, show similarity with the shale formations in the thermal well sites were not supported by the data gathered exhibiting elevated concentrations, though mean and median for this study. No physical hydraulic connection is evidenced; concentrations for most analytes were slightly lower in the the waters exhibit different geochemical composition and dif- hot springs compared to water from the shale formations. The ferent isotopic and trace metal signatures; the newly discov- chemistry of the hot springs is likely a result of rock/water ered thermal sites appear to be low-volume systems; and the interaction in the shale formations in the deeper sections of the sites are separated by distance and geological and hydrological References Cited 37 boundaries. Thermal-water sites across the Ouachita Moun- Branner, J.C., 1892, Mineral waters of Arkansas: Arkansas tains appear to represent discrete systems that are associated Geological Survey, Annual Report, v. I, p. 18-27. with a specific set of hydrologic and geologic conditions, which occur at numerous locations across the region, and are Bryan, Kirk, 1922, The hot water supply of the hot springs, viewed as analogs to the hot spring flow system in Hot Springs Arkansas: Journal Geology, v. 30, p. 425-449. National Park rather than connected components of the same Bullen, T.D., Krabbenhoft, D., and Kendall C., 1996, Kinetic hydraulic system. Although the mechanism for generation of and mineralogic controls on the evolution of groundwater thermal water appears to be very similar for the described ther- chemistry and 87Sr/86Sr in a sandy silicate aquifer, north- mal water occurrences throughout the Ouachita Mountains, the ern Wisconsin: Geochim. et Cosmochim. Acta, v. 60, no. 10, temperature of these waters will vary at the land surface where p. 1807-1821. they issue as springs or along subsurface fractures intersected by shallow wells. Century Geophysical Corporation, 2006, 9721 Logging tool— E-M flowmeter: accessed November 3, 2009, at http://www. century-geo.com/9721-index.html Childress, C.J.O., Foreman, W.T., Conner, B.F., Maloney, T.J., References Cited 1999, New reporting procedures based on long-term method detection levels and some considerations of water-quality Appelo, C.A.J., and Postma, D., 1999, Geochemistry, ground- data provided by the U.S. Geological Survey National water and pollution: Brookfield, Vermont, A.A. Bakema Water Quality Laboratory: U.S. Geological Survey Open- Publishers, 536 p. File Report 99-193, 19 p. Albin, D.R., 1965, Water-resources reconnaissance of the Clark, I.D., and Fritz, P., 1997, Environmental isotopes in Ouachita Mountains, Arkansas: U.S. Geological Survey hydrogeology: Boca Raton, Florida, CRC Press, 328 p. Water-Supply Paper 1809-J, 14 p. Crecelius, E. , Trefry, J. , McKinley, J. , Lasorsa , B., and Tro- Arkansas Geological Survey, 2009, Arkansas’ geological map- cine , R., 2007, Study of barite solubility and the release of ping program: Arkansas Geological Survey, accessed May trace components to the marine environment: New Orleans 1, 2009 at http:// www.geology.ar.gov/geology/geo_map. (LA) U.S. Department of the Interior, Minerals Manage- htm ment Service, Gulf of Mexico Region, OCS Study MMS 2007–061, 147 p. Arndt, R.H., and Stroud, R.B., 1953, Thrust faulting near the hot springs, Hot Springs National Park, Arkansas, App. 3, Daniel B. Stephens and Associates, Inc., 2002, Comparative in Arndt, R.H., and Damon, P.E., Radioactivity of thermal study of domestic water well integrity to coal mine blasting waters and its relationship to the geology and geochemistry summary report: unpublished report prepared for Office of of uranium: Arkansas University of Institute of Science and Surface Mining Reclamation and Enforcement, Pittsburg, Technology Annual Progress Report to U.S. Atomic Energy Pennsylvania, 28 p. Commission, 28 p. Drimmie, R.J.; Heemskerk, A.R.; and Aravena, R., 1990, Dis- Bedinger, M.S., Pearson, F.J., Jr., Reed, J.E., Sniegocki, R.T., solved inorganic carbon (DIC): Technical Procedure 5.0, and Stone, C.G., 1979, The waters of Hot Springs National Revision 01, Environmental Isotope Laboratory, Depart- Park, Arkansas--Their nature and origin: U.S. Geological ment of Earth Sciences, University of Waterloo, 3 p. Survey Professional Paper 1044-C, 33 p. Duvall, W.I., and Fogelson, D.E., 1962, Review of criteria for Begor, K.F., Miller, M.A., and Sutch, R.W., 1989, Creation of estimating damage to residences from blasting activities: an artificially produced fracture zone to prevent contami- U.S. Department of Interior Bureau of Mines, Report of nated groundwater migration: Journal of Groundwater, v. Investigations 5968, 19 p. 27, no. 1, p. 57-65. Emerson Process Management, 2004, Reverse osmosis mea- Bell, R.W. and Hays, P.D., 2007, Influence of locally derived surements: Application Data Sheet 4950-83/rev.B, accessed recharge on the water quality and temperature of springs October 28, 2009 at http://www.emersonprocess.com/RAI- in Hot Springs National Park, Arkansas: U.S. Geological home/Documents/Liq_AppData_4950-83.pdf Survey Scientific Investigations Report 2007-5004, 45 p. Faure, G., 1986, Principles of isotope geology (3rd ed.): John Bergfelder, Bill, 1976, The origin of the thermal water at Hot Wiley and Sons, New York, 589 p. Springs, Arkansas: Columbia, Missouri, University of Mis- Fetter, C.W., 1988, Applied hydrology: Don Mills, Ontario, souri, Master of Arts thesis, 62 p. Macmillan Publishing Co., 592 p. 38 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs

Fishman, M.J., ed., 1993, Methods of analysis by the U.S. Lane, J.W., Haeni, F.P., Soloyanis, S., Placzek, G., Williams, Geological Survey National Water Quality Laboratory--De- J.H., Johnson, C.D., Buursink, M.L., Joesten, P.K., and termination of inorganic and organic constituents in water Knutson, K.D., 1996, Geophysical characterization of a and fluvial sediments: U.S. Geological Survey Open-File fractured-bedrock aquifer and blast-fractured contaminant- Report 93-125, 217 p. recovery trench, in Proceedings of the Symposium on the Application of Geophysics to Engineering and Environmen- Fournier, R.O., 1981, Application of water geochemistry to tal Problems: Environmental and Engineering Geophysical geothermal exploration and reservoir engineering: in Geo- Society, p. 429-441. thermal Systems: Principles and Case Histories, Rybach, L., and Muffler, L. (eds.) New York, John Wiley and Sons, p. National Atmospheric Deposition Program (NRSP-3), 2009, 109-144. 2008 Annual Summary: accessed April 9, 2009, at http:// nadp.sws.uiuc.edu/ Fournier, R.O., and Rowe, J.J., 1966, Estimation of under¬ground temperature from the silica content of water National Climatic Data Center, 2009, Climate – Radar Data from hot springs and wet steam wells: American Journal of Inventories; Central and Southwest Stations, Arkansas, Sci¬ence, v. 264, p. 685-697. accessed July 19, 2009, at http://www.ncdc.noaa.gov/oa/ climate/stationlocator.html Garbarino, J.R., 1999, Methods of analysis by the U.S. Geological Survey National Water Quality Laboratory -- North Dakota State University, 1992, Treatment systems Determination of dissolved arsenic, boron, lithium, sele- for household water supplies – reverse osmosis: NDSU nium, strontium, thallium, and vanadium using inductively Agriculture and University Extension, publication AE-1047, coupled plasma-mass spectrometry: U.S. Geological Survey accessed October 28, 2009, at http://www.ag.ndsu.edu/pubs/ Open-File Report 99-093, 31 p. h2oqual/watsys/ae1047w.htm Glasgow, W., Jr., 1860, Hot Springs of Arkansas, with their Ostlund, H.G., and Werner, E., 1962, The electrolytic enrich- travertine formation: St. Louis, Missouri, L. Gast Brothers ment of tritium and deuterium for natural tritium measure- and Co. ments: in Tritium in the physical and biological sciences, vol. 1, International Atomic Energy Agency, p. 95-104. Golder and Associates Ltd., 2004, Blasting impact assessment proposed Nelson Aggregate Nelson quarry extension: Owen, D.D., 1860, Second report of a geological reconnais- unpublished report submitted to Nelson Aggregate Co., sance of the middle southern counties of Arkansas: Phila- Burlington, Ontario, Report # 021-1238, 25 p. delphia, Pa., C. Sherman and Sons, p. 18-27. Halberg, H.N., Bryant, C.T., and Hines, M.S., 1968, Water Purdue, A.H., 1910, The collecting area of the waters of the resources of Grant and Hot Springs County, Arkansas: U.S. hot springs, Hot Springs, Arkansas: Journal of Geology, v. Geological Survey Water-Supply Paper 1857, 64 p. 18, p. 279-285. Haley, B.R., 1976 (revised 1993), Geologic map of Arkansas: Purdue A.H, and Miser, H.D., 1923, Hot Springs folio, Arkan- U.S. Geological Survey, 1 sheet. sas: U.S. Geological Survey Geologic Atlas, Folio no. 215, 12 p. Haywood, J.K., 1902, Report of analysis of the waters of the Hot Springs on the Hot Springs Reservation: U.S. 57th Révész, Kinga, and Coplen, T.B., 2008a, Determination of Con¬gress, 1st session, Senate Document 282, 78 p. the δ(2H/1H) of water: RSIL lab code 1574, chap. C1 of Révész, Kinga, and Coplen, T.B., eds., Methods of the Hem, J.D., 1989, Study and interpretation of the chemical Reston Stable Isotope Laboratory: U.S. Geological Survey characteristics of natural water (3rd ed.): U.S. Geological Techniques and Methods book 10, chap. C1, 27 p. Survey Water-Supply Paper 2254, 263 p. Révész, Kinga, and Coplen, T. B., 2008b, Determination of Kendall, C., and McDonnell, J.J., eds., 1998, Isotope tracers in the δ(18O/16O) of Water: RSIL Lab Code 489, in Révész, catchment hydrology: Amsterdam, Elsevier, 839 p. Kinga, and Coplen, T.B., eds., Methods of the Reston Stable Keys, W.S., 1997, A practical guide to borehole geophysics in Isotope Laboratory: U.S. Geological Survey Techniques and environmental investigations: Boca Raton, Fla., CRC-Lewis Methods, book 10, chap. C2, 28 p. Publishers, 176 p. References Cited 39

Robertson, D.A., Gould, J.A., Straw, J.A., and Dayton, M.A., 1980, Survey of blasting effects on groundwater supplies in Appalachia: unpublished report prepared for U.S. Depart- ment of Interior Bureau of Mines, Contract # J-0285029, 139 p. Rozanski, K., Araguas-Araguas, L., and Gonfiantini, R., 1993, Isotopic patterns in modern global precipitation, in Conti- nental Isotope Indicators of Climate: American Geophysical Union, 212 p. Siskind, D.E., Stagg, M.S., and Kopp, J.W., and Dowding, C.H., 1980, Structure response and damage produced by ground vibration from surface mine blasting: U.S. Depart- ment of Interior Bureau of Mines Report of Investigations 8507, 74 p. Stone, C.G., and Bush, W.V., 1984, General geology and mineral resources of the Caddo River watershed: Arkansas Geological Commission Information Circular 29, 32 p. Stroud, R.B., Arndt, R.H., Fulkerson, F.B. and Diamond, W.G., 1969, Mineral resources and industries of Arkansas: U.S. Bureau of Mines Bulletin 645, 418 p. Stuiver, M., and Polach, H, 1977. Discussion: Reporting of 14C data: Radiocarbon, v. 19, no. 3, p. 355-363 Thatcher, L. L., Janzer, V.J., and Edwards, K.W., 1977, Meth- ods for determination of radioactive substances in water and fluvial sediments: U.S. Geological Survey Techniques of Water-Resources Investigations, book 5, chap. A5, p. 79-81. Wagner, G.H., and Steele, K.F., 1985, Metal concentrations in the groundwater of the Ouachita Mountains, Arkansas, U.S.A.: Water, Air and Soil Pollution, v. 25, p. 339-347. Weed, W.H., 1902, Geological sketch of Hot Springs, Arkan- sas: U.S. 57th Congress, 1st session, Senate Document 282, p. 79-94. Wilde, F.D., 2008, Field measurements: U.S. Geological Sur- vey Techniques of Water-Resources Investigations, book 9, chap. A6, variously paginated. Wilde, F.D., Radke, D.B., Gibs, J., Iwatsubo, R.T., 1998, Cleaning of equipment for water sampling: U.S. Geological Survey Techniques of Water-Resources Investigations, book 9, chap. A3, variously paginated. Yeatts, D.S., 2006, Characteristics of thermal springs and the shallow ground-water system at Hot Springs National Park, Arkansas: U.S. Geological Survey Scientific Investigations Report 2006-5001, 35 p.

Appendixes 1-2

Appendix 41 -- 8.9 3.9 8.0 6.8 73.0 30.2 22.3 22.6 10.8 18.8 17.7 12.7 113.2 139.0 136.8 140.6 136.6 140.8 137.6 149.0 142.0 262.5 165.4 109.6 144.0 132.8 140.7 140.1 Total solids (mg/L) dissolved (calculated) ) ) 3 -- -- 2 0.1 5 1 4 1 53 96 21 16 12 68 85 12 130 130 130 130 130 130 140 230 140 140 120 133 130 CaCO (mg/L as Hardness 62.1 61.8 62.5 55.8 57.4 61.2 23.4 33.8 19.2 19.2 19.5 17.9 18.3 19.2 17.5 19.5 16.7 17.1 20.4 18.3 19.1 18.7 18.4 61.3 20.0 54.0 16.5 57.9 61.9 (°C) ature Water temper- 98 63 38 44 24 50 22 30 17 31 23 29 311 284 290 302 298 297 286 233 220 477 295 184 154 235 316 290 302 (field) (µS/cm) (µS/cm) Specific conductance 7.5 7.5 7.4 7.1 7.2 7.2 7.0 6.2 6.7 6.1 7.1 6.9 5.6 5.1 4.4 3.6 4.2 4.7 4.2 7.4 5.9 5.8 6.3 3.7 7.2 3.9 7.3 4.8 6.3 pH (field) ------4.0 1.5 0.5 0.5 0.9 0.7 0.5 5.7 0.5 8.3 0.6 0.5 0.9 1.0 6.8 5.5 6.8 g/L, micrograms per liter; pCi/L, picocuries per µ g/L, micrograms per liter; Dis- (mg/L) solved oxygen Sr, strontium-87/strontium-86; ams, accelerator mass spectrometry; PMC, percent Sr, 86 850 817 1411 1105 1045 1009 1555 1532 1100 1015 1100 1016 1300 1000 1000 1600 1700 1200 1500 1400 1400 1200 1300 1200 1400 1000 1500 1330 1530 col - Sr/ Time lected 87 , calcium carbonate; 3 Date 9/11/2007 9/11/2007 9/12/2007 9/12/2007 9/11/2007 9/10/2007 9/13/2007 9/10/2007 9/11/2007 6/10/2008 9/11/2007 6/18/2008 9/11/2007 9/13/2007 9/13/2007 9/07/2007 9/06/2007 9/07/2007 9/05/2007 9/07/2007 9/07/2007 9/06/2007 9/05/2007 9/10/2007 9/05/2007 9/13/2007 9/12/2007 9/10/2007 9/12/2007 collected C, carbon-13/carbon-12; 12 C/ 13 Formation Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Arkansas Novaculite Arkansas Novaculite Bigfork Chert Stanley Shale Womble Shale Womble Bigfork Chert Bigfork Chert Bigfork Chert Bigfork Chert Bigfork Chert Bigfork Chert Bigfork Chert Stanley Shale Stanley Shale Stanley Shale Stanley Shale Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Arkansas Novaculite O, oxygen-18/oxygen-16; Field name 16 O/ 18 Spring8 Spring9 Spring49 Spring25 Spring33 Spring46 Spring42 Little Missouri Spring Caddo Gap Spring Sleepy Valley Spring Valley Sleepy ArScenic Spring Rigsby Greer (MW1) Greer (MW2) Greer Ester Trusty-Carlin Ester King Gates Bratton4 Bratton2 (MW) Sharp Bratton Thornton Schnick Spring47 Appel Spring17 Spring6 Thomas Station name SPRING 8 SPRING 9 SPRING 49 HALE SPRING DISPLAY SPRING DISPLAY Fordyce Spring Quapaw Spring 04S27W08DDC1 04S24W19ABB1 02S19W27AAB1SP AR-SCENICSPRING 02S18W06ACD1 02S18W30CCB3 02S18W30CCB2 02S18W30CCB1 02S19W13ADC1 02S18W09DCD1 02S18W04DBD1 02S18W23ADB2 02S19W32DAB2 02S18W21CBD1 02S18W32DAB1 02S19W35ADD1 02S18W30DDD1 SPRING 47 02S18W22BBA1 02S19W33CBB6SP 02S19W33BCD2 02S19W24BCA1 H/H, deuterium/protium; 2 µ S/cm, microsiemens per centimeter at 25 degrees Celsius; °C, CaCO Site number Inorganic and isotopic water-quality data for samples collected in study area, Hot Springs, Arkansas. Inorganic and isotopic water-quality identification 343056093030901 343055093031301 343054093031201 343052093031301 343050093031201 343049093031301 343048093031301 343227093543501 342314093363301 343211093011501 343130093021901 343511092580901 343132092585402 343132092585501 343132092585401 343330092591101 343350092560701 343458092560201 343238092535201 343047092571201 343224092562301 343047092571101 343056093001701 343124092580301 343058093030901 343256092554201 343057093031301 343056093031101 343230093000701 5 3 1 91 90 89 88 87 86 85 84 83 61 60 53 52 51 49 46 45 41 40 34 30 24 16 94 93 92 Map number (figure 1) Appendix 1. [mg/L, milligrams per liter; liter; <, less than; E, estimated; modern carbon; --, not measured] 42 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs -- -- 0.39 0.43 0.40 0.44 0.38 0.39 0.32 0.38 6.88 0.38 0.30 1.70 1.60 1.40 2.09 1.10 3.30 7.20 0.49 0.47 0.13 0.52 <0.12 <0.12 <0.12 <0.12 <0.12 (µg/L) Arsenic Arsenic -- -- 0.40 0.55 0.07 0.25 0.81 0.24 0.15 0.09 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.20 <0.06 <0.06 <0.06 <0.06 E 0.04 E 0.06 E 0.04 (µg/L) Antimony Antimony -- -- 4.1 2.9 2.5 1.7 1.9 1.7 1.6 2.0 3.1 2.8 9.9 5.1 5.3 2.0 2.6 13.3 58.6 2.09 83.6 29.8 54.6 11.7 27.4 50.7 214.4 E 1.6 E 1.2 num num (µg/L) Alumi- -- -- 0.00006 0.00007 0.00008 0.01712 0.00004 0.00004 0.11572 0.00003 0.00049 0.19653 0.00005 0.00045 0.00006 0.00146 0.00116 0.00004 0.06659 0.02205 0.06215 0.242 0.04202 0.00765 0.00231 0.00012 0.00008 0.00076 0.0001 ion (mg/L) Hydrogen (calculated) (calculated) 40.4 41.3 39.7 10.7 40.8 39.8 9.08 39.6 7.76 39.7 31.3 39.7 15.8 31.3 18.3 7.23 8.34 8.38 11.8 13.5 10.8 10.7 11.1 12.5 22.2 11.2 39.8 39.89 12.33 Silica Silica (mg/L) 0.1 0.1 0.14 0.15 0.13 0.14 0.13 0.13 0.13 0.13 0.21 0.13 0.15 0.18 0.18 0.17 0.15 0.14 0.18 0.13 0.14 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 < 0.1 E 0.1 (mg/L) Fluoride Fluoride 1.86 1.82 1.48 1.83 1.84 2.92 1.77 1.84 2.12 1.86 1.87 2.86 1.84 4.73 2.57 0.98 1.49 1.52 3.00 1.83 1.71 1.22 9.95 2.23 2.28 3.18 1.63 1.93 11.2 Sr, strontium-87/strontium-86; ams, accelerator mass spectrometry; PMC, percent Sr, (mg/L) 86 Chloride Chloride Sr/ 87 -- 7.66 1.02 7.48 7.25 1.11 7.53 7.49 2.67 7.53 7.60 7.32 1.32 4.11 2.95 4.87 3.51 3.60 35.4 6.19 4.78 5.31 7.79 7.46 22.71 26.17 11.50 12.05 13.55 , calcium carbonate; µ g/L, micrograms per liter; pCi/L, picocuries 3 (mg/L) Sulfate ------0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3 0 0 0 0 0 0 0 0 (field) (mg/L) Carbonate C, carbon-13/carbon-12; 12 -- C/ 2 1 1 4 4 5 13 13 95 83 14 17 20 10 62 153 161 156 159 161 161 151 167 180 249 164 133 154 161 (field) (mg/L) Bicar- bonate ) 3 -- 0 0 0 3 0 4 8 10 78 68 11 14 16 51 126 132 128 131 132 132 124 138 149 205 135 109 127 132 (field) CaCO (mg/L as Alkalinity 3.7 1.00 3.92 3.75 1.53 3.88 3.88 1.60 3.91 4.04 3.91 0.69 0.91 1.03 1.12 9.17 1.46 1.68 1.45 4.16 1.61 3.95 4.01 11.3 14.8 14.3 11.5 O, oxygen-18/oxygen-16; 10.83 1.756 16 (mg/L) O/ Sodium 18 2.62 4.78 4.83 4.72 0.21 4.79 4.78 0.36 4.82 4.81 4.01 4.88 2.19 4.56 5.92 0.14 0.34 0.61 0.15 0.21 0.01 0.29 8.00 2.03 0.58 2.14 2.48 4.84 4.83 sium (mg/L) Magne - 0.3 4.1 0.9 0.31 0.12 0.81 0.39 0.01 6.11 7.57 45.2 45.3 44.4 44.7 44.8 45.3 42.0 45.9 12.4 19.5 44.7 78.1 51.1 34.7 17.3 45.2 45.4 27.16 0.078 (mg/L) Calcium Calcium H/H, deuterium/protium; 2 Station number Inorganic and isotopic water-quality data for samples collected in study area, Hot Springs, Arkansas.—Continued Inorganic and isotopic water-quality identification 343230093000701 343050093031201 343052093031301 343054093031201 343256092554201 343055093031301 343056093030901 343124092580301 343056093031101 343057093031301 343056093001701 343058093030901 343047092571101 343224092562301 343047092571201 343238092535201 343458092560201 343350092560701 343330092591101 343132092585401 343132092585501 343132092585402 343511092580901 343130093021901 343211093011501 342314093363301 343227093543501 343048093031301 343049093031301 1 3 5 87 88 89 90 91 92 93 16 94 24 30 34 40 41 45 46 49 51 52 53 60 61 83 84 85 86 Map number (figure 1) liter; <, less than; E, estimated; modern carbon; --, not measured] Appendix 1. Celsius; CaCO Celsius; °C, degrees [mg/L, milligrams per liter; µ S/cm, microsiemens centimeter at 25 degrees Appendix 43 -- -- 0.09 0.11 0.17 0.12 0.09 0.06 0.12 0.20 0.25 0.09 0.54 0.44 0.36 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 <0.08 (µg/L) Selenium Selenium -- -- 0.91 0.58 1.4 0.42 1.2 0.85 0.3 2.7 0.32 2.3 0.19 7.8 0.19 0.46 0.33 0.42 8.7 2.1 7.3 3.8 5.0 1.3 0.46 0.29 14 11.4 E 0.05 (µg/L) Nickel -- -- 0.2 0.2 0.3 0.2 1 0.1 0.2 0.3 0.2 0.4 0.2 0.4 0.5 2.3 0.7 0.7 0.2 0.2 0.2 <0.1 <0.1 <0.1 <0.4 <0.1 <0.1 <0.1 <0.1 (µg/L) Moly- bdenum -- -- 3.3 4.7 1.1 1.6 6.2 60.4 41.8 71.5 12 17.1 23.6 22.7 81.1 41.4 <0.2 <0.2 <0.2 229 226 140 213 810 212 144 223 215 1115 nese (µg/L) Manga - -- -- 6.8 5 0.7 4.4 4.9 4.7 4.7 2.9 4.8 4.2 6 4.7 0.8 0.6 1.4 4.7 5.8 3.6 4.8 4.7 16.7 18.3 <0.6 <0.6 <0.6 <0.6 E 0.4 (µg/L) Lithium -- -- 0.8 0.67 6.63 0.39 0.21 0.16 0.25 1.74 1.66 13.5 14.2 50.13 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 E 0.09 E 0.09 Lead µg/L) -- -- 8 Sr, strontium-87/strontium-86; ams, accelerator mass spectrometry; PMC, percent Sr, 38 25 29 20 89 96 12 21 18 14 <6 <6 <6 <6 <6 <6 86 611 158 798 E 4 E 5 E 4 216.4 696.6 2510 1220 Iron Iron Sr/ (µg/L) 87 , calcium carbonate; µ g/L, micrograms per liter; pCi/L, picocuries -- -- 3 2.1 7.2 0.4 1.7 2.1 0.7 0.9 0.9 3.4 0.9 0.7 1.2 31.6 42 15.6 89.7 19.7 <0.4 <0.4 <0.4 <0.4 <0.4 <0.4 446 E 0.4 E 0.4 E 0.23 (µg/L) (µg/L) Copper Copper -- -- 0.8 2.5 2.1 2.7 3.9 1.8 2.9 0.32 0.08 0.06 0.22 0.05 0.05 1.73 0.08 0.05 0.16 0.44 2.40 0.45 0.46 0.16 0.04 <0.04 <0.04 <0.04 <0.04 (µg/L) Cobalt Cobalt C, carbon-13/carbon-12; 12 C/ -- -- 13 0.09 0.08 0.12 0.22 0.13 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 <0.12 E 0.08 (µg/L) Chromium Chromium -- -- 0.05 0.23 0.05 0.36 0.04 0.12 0.08 0.108 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 (µ(g/L) Cadmium Cadmium 9 -- -- 12 10 11 10 11 11 28 11 11 10 15 11 56 11 11 <8 <8 <8 <8 E 5 E 6 E 5 E 4 E 6 O, oxygen-18/oxygen-16; E 6 16 E 4.7 (µg/L) Boron Boron O/ 18 -- -- 0.11 0.06 0.09 0.18 0.10 0.09 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 <0.06 E 0.04 E 0.03 E 0.05 (µg/L) Beryllium Beryllium -- -- 0.7 5 2 2 9 8 27 20 17 75 86 18 81 18 21 51 13 143 141 111 138 108 146 138 139 138 E 0.05 (µg/L) Barium Barium H/H, deuterium/protium; 2 Station number Inorganic and isotopic water-quality data for samples collected in study area, Hot Springs, Arkansas.—Continued Inorganic and isotopic water-quality identification 343230093000701 343256092554201 343052093031301 343124092580301 343054093031201 343056093001701 343055093031301 343056093030901 343047092571101 343056093031101 343224092562301 343057093031301 343047092571201 343058093030901 343238092535201 343458092560201 343350092560701 343330092591101 343132092585401 343132092585501 343132092585402 343511092580901 343130093021901 343211093011501 342314093363301 343227093543501 343048093031301 343049093031301 343050093031201 1 3 5 88 89 16 90 91 24 92 30 93 34 94 40 41 45 46 49 51 52 53 60 61 83 84 85 86 87 Map number (figure 1) liter; <, less than; E, estimated; modern carbon; --, not measured] Appendix 1. Celsius; CaCO Celsius; °C, degrees [mg/L, milligrams per liter; µ S/cm, microsiemens centimeter at 25 degrees 44 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs Sr 86 0.7119 0.7119 0.7119 0.7119 0.7121 0.7082 0.7103 0.71192 0.71108 0.71107 0.71196 0.71189 0.71192 0.71196 0.71191 0.71211 0.70937 0.70851 0.71064 0.71258 0.72242 0.70969 0.71319 0.71092 0.71383 0.71032 0.71435 0.71812 0.72205 Sr/ 87 ------36.6 78.6 85.2 77.8 36.2 40 36.3 23.7 38.3 37 66.1 79.5 44.8 35.7 36.9 35.6 (PMC) Carbon-14 ------0.2 0.4 0.4 0.4 0.2 0.2 0.2 0.2 0.2 0.2 0.3 0.4 0.3 0.2 0.2 0.2 ams error) (percent Carbon-14 ------C 12 -13.3 -12.7 -13.8 -14.03 -22.14 -23.09 -23.49 -13.51 -13.87 -13.24 -17.37 -19.43 -14.02 -14.73 -13.42 -13.86 C/ (ratio 13 per mil) ------O 16 -5.52 -5.58 -5.11 -5.44 -5.32 -5.63 -5.58 -5.60 -5.41 -5.52 -5.61 -5.48 -5.47 -5.81 -5.56 -5.56 -5.41 -5.52 -4.88 -5.55 -5.57 -5.54 -5.73 -5.48 -5.60 -5.49 O/ (ratio 18 per mil) ------2.1 0.10 0.13 0.49 0.45 0.09 0.06 0.040 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 E 0.03 E 0.03 (µg/L) Uranium Uranium ------2 7 8 8 7 9 1 0 9 0 7 9 12 Sr, strontium-87/strontium-86; ams, accelerator mass spectrometry; PMC, percent Sr, 86 (pCi/L) Tritium Tritium Sr/ 87 ------3.0 3.2 3.2 3.2 3.2 3.2 3.2 3.2 3.2 3.2 3.2 3.2 3.2 , calcium carbonate; µ g/L, micrograms per liter; pCi/L, picocuries 3 pCi/L Tritium Tritium 2-sigma 2-sigma ------28.9 -29.4 -29.5 -29.9 -28.7 -31.3 -30.6 -30.2 -29.3 -30.4 -29.5 -29.6 -29.4 -32.2 -28.6 -28.3 -29.9 -28.2 -28.1 -29.7 -30.7 -29 -31.3 -30.2 -31 -28.6 H/H 2 (ratio per mil) C, carbon-13/carbon-12; -- -- 12 1.2 8.9 5.1 0.6 4.3 7.3 6.7 6.4 6.5 198 114 41.1 26.7 49.9 10.5 39.1 31.6 <0.6 <0.6 <0.6 C/ 33.42 10.23 E 0.48 E 0.37 E 0.49 E 0.47 E 0.42 Zinc 13 (µg/L) -- -- 0.06 0.09 0.21 0.15 1.5 0.56 0.15 0.06 1.342 0.22 0.44 0.06 0.05 0.05 0.04 0.05 0.06 0.05 <0.04 <0.10 <0.04 <0.04 E 0.03 E 0.03 E 0.03 E 0.03 E 0.03 (µg/L) (µg/L) Vanadium -- -- 0.04 0.04 0.23 0.1 0.19 0.31 0.14 0.05 0.003 0.04 0.04 0.05 0.4076 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 <0.04 E 0.03 E 0.03 E 0.04 E 0.03 E 0.04 E 0.02 E 0.04 O, oxygen-18/oxygen-16; (µg/L) 16 Thallium Thallium O/ 18 3.02 4.03 5.07 1.87 6.63 1.81 2.33 37.1 83.0 40.5 19.1 10.7 98.8 <0.4 108 106 208 287 106 102 105 324 104 328 236 107 209 106 101 (µg/L) Strontium -- -- <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.2 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 <0.1 < 0.1 Silver (µg/L) H/H, deuterium/protium; 2 Site number Inorganic and isotopic water-quality data for samples collected in study area, Hot Springs, Arkansas.—Continued Inorganic and isotopic water-quality identification 343049093031301 343048093031301 343227093543501 342314093363301 343211093011501 343130093021901 343511092580901 343132092585402 343132092585501 343330092591101 343132092585401 343350092560701 343058093030901 343458092560201 343057093031301 343238092535201 343056093031101 343047092571201 343056093030901 343047092571101 343224092562301 343055093031301 343056093001701 343054093031201 343124092580301 343052093031301 343256092554201 343050093031201 343230093000701 5 3 1 86 85 84 83 61 60 53 52 51 46 49 45 94 41 93 40 92 34 91 24 30 90 16 89 88 87 Map number (figure 1) liter; <, less than; E, estimated; modern carbon; --, not measured] Appendix 1. Celsius; CaCO Celsius; °C, degrees [mg/L, milligrams per liter; µ S/cm, microsiemens centimeter at 25 degrees Appendix 45 Sr 86 0.71197 0.71199 0.71197 0.71197 0.71198 0.71434 0.7137 0.71534 0.7144 0.71343 0.71265 0.71505 0.71417 0.71446 0.71505 0.72082 0.71017 0.71221 0.71583 0.71484 0.71574 0.71648 0.71367 0.71458 0.71042 0.71311 0.71298 0.71347 Sr/ 87 6 2 5 5 9 3 4 6 5 4 6 (ug/L) 11 98 96 14 10 13 20 12 22 42 101 106 101 106 381 282 385 Strontium 845 915 950 820 915 945 905 945 930 1110 1115 1155 1135 1000 1035 1510 1255 1325 1218 1505 1445 1615 1415 1340 1500 1210 1025 1040 Time collected Time g/L, micrograms per liter; pCi/L, picocuries µ g/L, micrograms 5/29/2008 5/29/2008 5/29/2008 5/29/2008 4/29/2008 5/29/2008 4/29/2008 4/29/2008 4/28/2008 4/29/2008 4/29/2008 4/29/2008 4/29/2008 4/29/2008 4/29/2008 4/28/2008 4/30/2008 4/28/2008 4/28/2008 4/28/2008 4/28/2008 4/28/2008 4/28/2008 4/28/2008 4/30/2008 4/29/2008 4/30/2008 4/30/2008 Sr, strontium-87/strontium-86; ams, accelerator mass spectrometry; Sr, 86 Date collected Sr/ , calcium carbonate; 3 87 Formation C, carbon-13/carbon-12; 12 Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Hot Springs Sandstone Stanley Shale Hot Springs Sandstone Bigfork Chert Arkansas Novaculite Arkansas Novaculite Bigfork Chert Bigfork Chert Bigfork Chert Arkansas Novaculite Arkansas Novaculite Arkansas Novaculite Missouri Mountain Shale Stanley Shale Bigfork Chert Bigfork Chert Bigfork Chert Missouri Mountain Shale Hot Springs Sandstone Missouri Mountain Shale Arkansas Novaculite Stanley Shale Stanley Shale Stanley Shale Stanley Shale C/ 13 Field name Spring 17 Spring 8 Spring 49 Spring 25 Denise Spring Spring 46 - Fordyce Thousand Spring Novastone Spring Montanus Spring Sam Spring Seth Spring Panther Valley Panther Simmerman Spring Covenant Spring Powerline Spring Tree Fall Spring Tree Amphitheater Spring Thresher South Thresher North Between Spring Java Spring Jerryrig Spring Sangster Two Sangster Sangster Lane Spring Artesian Shinkle Bodi Smith O, oxygen-18/oxygen-16; 16 O/ 18 Station name 02S19W33CBB6SP SPRING 8 SPRING 49 HALE SPRING 02S18W19CCC1 Fordyce Spring 02S18W30BCB1 02S19W26BCA1 02S19W27ADB1 02S18W30ABB2 02S18W30BAA1 02S18W30BDA1 02S18W30DCC1 02S18W30DCA1 02S18W30DAD1 02S18W22CCD1 02S18W32ACC1 02S18W22CDB1 02S18W22CBD1 02S18W27BCA1 02S18W27BDA1 02S18W28ADA1 02S18W27BBA1 02S18W27BBC1 02S19W36CBB1 02S19W25BCD1 02S18W28BCB1 02S18W31DDB1 H/H, deuterium/protium; 2 µ S/cm, microsiemens per centimeter at 25 degrees Celsius; °Celsius, CaCO Site number identification Inorganic and isotopic water-quality data for samples collected in study area, Hot Springs, Arkansas.—Continued Inorganic and isotopic water-quality 343057093031301 343056093030901 343054093031201 343052093031301 343218092585501 343049093031301 343158092585501 343159093005301 343156093012101 343211092582301 343209092583101 343155092583301 343126092582601 343128092581901 343134092580501 343209092553601 343057092572101 343214092553501 343219092553601 343152092554101 343149092553101 343150092555301 343203092553901 343159092554601 343050093000101 343150092594901 343150092565101 343034092581601 9 6 93 91 89 88 82 86 81 80 79 78 77 75 74 73 72 71 69 68 67 65 66 64 63 62 33 19 Map number (figure 1) Appendix 1. [mg/L, milligrams per liter; liter; <, less than; E, estimated; PMC, percent modern carbon; --, not measured] 46 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs 800 710 580 990 490 550 510 600 530 525 530 540 530 520 550 520 470 600 480 550 720 685 595 560 500 520 655 685 625 1,025 Land- datum surface altitude (NGVD 29) (NGVD Date toried inven- 4/04/2007 6/02/2007 4/02/2007 2/28/2007 4/05/2007 4/02/2007 4/02/2007 4/02/2007 4/02/2007 2/27/2007 2/27/2007 2/27/2007 2/27/2007 2/28/2007 2/28/2007 4/03/2007 4/03/2007 4/04/2007 4/03/2007 4/05/2007 6/02/2007 6/02/2007 6/02/2007 6/02/2007 11/08/2006 11/08/2006 11/16/2006 11/08/2006 11/14/2006 11/14/2006 ------4.77 5.91 4.73 3.70 7.40 7.67 6.24 6.35 5.84 6.98 7.33 7.68 6.44 pH ------29 62 24 31 97 338 354 419 235 181 155 392 490 202 tance (µS/cm) conduc Specific ------20.1 20.6 18.5 17.0 17.7 22.8 19.1 18.0 18.9 18.0 18.0 17.9 18.1 (°C) ature Temper------753 702 580 754 487 523 482 580 522 518 517 526 526 507 541 501 466 583 480 547 716 667 472 501 618 619 (feet level above Water- altitude NGVD 29) ------7.63 3.00 7.73 7.35 4.28 9.16 3.68 0.00 3.49 3.83 6.24 46.82 26.98 28.45 20.34 12.59 14.01 12.65 19.10 17.00 17.56 28.42 18.85 66.66 236.16 Artesian land (feet level Water Water below surface) ------35 90 74 75 48 150 300 100 300 130 253 100 130 120 200 197 130 Well Well (feet) depth Site number identification 343230093000701 343115092582001 343256092554201 343043092565801 343124092580301 343034092581601 343051092575401 343058092574001 343150092565101 343025092584601 343023092584801 343030092572201 343119092551701 343204092585501 343026092570901 343056093001701 343031093001901 343124093000101 343150092594901 343022092590301 343229092585201 343231092532001 343210092535201 343047092571101 343035092572801 343029092593001 343039092580701 343039092560601 343245092560801 343224092562301 S/cm, microsiemens per centimeter at 25 degrees Celsius; --, not measured; 330ARKS, Arkansas Novaculite; 330HSPG, Hot Springs S/cm, microsiemens per centimeter at 25 degrees Celsius; --, not measured; 330ARKS, Longitude µ 93.002000 92.972194 92.928250 92.949333 92.967583 92.971222 92.964917 92.961111 92.947583 92.979556 92.979972 92.956278 92.921444 92.982028 92.952528 93.004639 93.005361 93.000222 92.996833 92.984194 92.981028 92.888750 92.897833 92.953167 92.957694 92.958333 92.968694 92.935139 92.935694 92.939778 Latitude 34.541750 34.520944 34.548806 34.511917 34.523278 34.509417 34.514194 34.516167 34.530500 34.507056 34.506306 34.508222 34.521889 34.534556 34.507167 34.515500 34.508611 34.523417 34.530472 34.506167 34.541361 34.542028 34.536167 34.513000 34.509778 34.507917 34.510861 34.544222 34.545861 34.540000 Well/ spring Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Formation 330ARKS 330ARKS 330HSPG 330HSPG 330HSPG 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL 330STNL Field name Measurement site characteristics for Hot Springs well and spring reconnaissance.—Continued Thomas House for sell Appel Stubbs Schnick Smith Lantz Remont Bodi USNR (north) USNR (south) Uzick Wheeler Jones UNK (basement) Thornton Edwards Daniel Shinkle UNK (MHSL) McDougal Blackwell Rest area Bratton 186 Bratton Cabinet shop Faye Jackson McCoy Sharp

1 2 3 4 5 6 7 8 9 11 10 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Map number (figure 1) Appendix 2. Datum of 1929: °C, degrees Celsius; Vertical [NGVD 29, National Geodetic Shale] Womble 364BGFK, Bigfork Chert; 367WMBL, Sandstone; 330STNL, Stanley Shale; 350SLRN, Missouri Mountain Shale and Polk Creek Shale, undifferentiated; Appendix 47 -- 540 560 485 595 710 715 700 680 750 790 550 760 735 620 610 670 740 580 705 790 700 705 540 560 530 480 445 490 555 Land- datum surface altitude (NGVD 29) (NGVD Date toried inven- 6/02/2007 6/02/2007 6/02/2007 7/03/2007 6/02/2007 7/10/2007 7/10/2007 2/27/2007 2/27/2007 2/27/2007 2/28/2007 2/28/2007 4/04/2007 4/03/2007 6/02/2007 7/10/2007 7/03/2007 7/03/2007 2/28/2007 2/28/2007 4/05/2007 4/05/2007 4/05/2007 9/13/2007 11/16/2006 11/16/2006 11/08/2006 10/26/2006 12/12/2006 12/12/2006 ------7.03 7.07 7.39 6.72 4.18 4.66 4.21 3.62 4.4 5.12 5.64 6.92 7.10 pH ------17 30 23 22 41 32 24 44 38 190 171 324 295 238 298 477 273 tance (µS/cm) conduc Specific ------16.9 17.5 18.9 17.5 18.0 17.3 16.6 22.4 18.1 23.8 17.2 19.2 (°C) ature Temper------544 485 564 667 665 747 795 543 625 671 609 610 644 540 573 609 738 605 608 532 552 520 471 432 490 547 (feet level above Water- altitude NGVD 29) ------2.92 7.08 7.43 8.08 7.64 9.53 8.80 7.74 16.05 30.77 42.97 50.48 63.80 10.93 25.58 95.95 51.64 94.79 96.89 12.62 134.80 200.40 Artesian Artesian Artesian land (feet level Water Water below surface) ------45 89 83 50 54 165 106 380 208 300 185 180 120 155 100 165 150 200 Well Well (feet) depth Site number identification 343118092572901 343114092574001 343050093000101 343047092571201 343305092593501 343304092593501 343312092592301 343129092583201 343240092535801 343238092535201 343458092560201 343208092582601 343252093002501 343441092553101 343350092560701 343330092591101 343146092584701 343153092582401 343132092585401 343253092535401 343132092585501 343132092585402 343511092580901 343445092584801 343523092575501 343407092491601 343424092492701 343614092554001 343450092522801 343130093021901 S/cm, microsiemens per centimeter at 25 degrees Celsius; --, not measured; 330ARKS, Arkansas Novaculite; 330HSPG, Hot Springs S/cm, microsiemens per centimeter at 25 degrees Celsius; --, not measured; 330ARKS, Longitude µ 92.958028 92.960972 93.000389 92.953250 92.993000 92.992944 92.989694 92.975528 92.899472 92.897861 92.933917 92.973944 93.006861 92.925361 92.935194 92.986278 92.979639 92.973472 92.981556 92.898306 92.981944 92.981528 92.969278 92.979889 92.965333 92.821222 92.824111 92.927722 92.874389 93.038611 Latitude 34.521639 34.520472 34.513972 34.512972 34.551472 34.551056 34.553417 34.524583 34.544472 34.544000 34.582694 34.535417 34.547806 34.578056 34.564000 34.558250 34.529333 34.531361 34.525472 34.547917 34.525472 34.525444 34.586333 34.579222 34.589722 34.568583 34.573250 34.603750 34.580583 34.525000 Well/ spring Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Well Spring Formation 330STNL 330STNL 330STNL 330STNL 350SLRN 350SLRN 350SLRN 350SLRN 350SLRN 350SLRN 364BGFK 364BGFK 364BGFK 364BGFK 364BGFK 364BGFK 364BGFK 364BGFK 364BGFK 350SLRN 364BGFK 364BGFK 367WMBL 367WMBL 367WMBL 367WMBL 367WMBL 367WMBL 367WMBL 330STNL Field name Measurement site characteristics for Hot Springs well and spring reconnaissance.—Continued Chamness Stallones Artesian Bratton2 (MW) Thomas (USGS1) Thomas (USGS2) Abrasives Smith Simmerman Bratton3 Bratton4 Gates Stanage Belvedere C.C. Weston King Trusty-Carlin Ester Weddock Valley Panther Greer Bratton5 Greer2 (MW1) Greer3 (MW2) Rigsby UNK (for rent) Church Loprinzi Lyon Fountain Lake Edds ArScenic Spring

31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Map number (figure 1) Appendix 2. Datum of 1929: °C, degrees Celsius; Vertical [NGVD 29, National Geodetic Shale] Womble 364BGFK, Bigfork Chert; 367WMBL, Sandstone; 330STNL, Stanley Shale; 350SLRN, Missouri Mountain Shale and Polk Creek Shale, undifferentiated; 48 Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters of Hot Springs -- 640 805 740 865 670 600 660 640 570 680 930 740 715 565 555 630 620 760 860 590 535 590 1,060 Land- datum surface altitude (NGVD 29) (NGVD Date toried inven- 4/28/2008 4/28/2008 5/05/2008 5/05/2008 5/05/2008 5/05/2008 5/05/2008 5/06/2008 5/06/2008 5/06/2008 5/07/2008 5/07/2008 5/07/2008 5/07/2008 5/07/2008 5/07/2008 4/29/2008 4/29/2008 4/28/2008 4/29/2008 4/29/2008 4/29/2008 6/18/2008 6/10/2008 ------6.10 4.09 4.46 4.13 4.05 6.72 6.17 pH ------63 23 25 19 20 20 20 22 20 10 16 28 13 21 19 18 98 232 222 147 233 tance (µS/cm) conduc Specific ------9.4 9.5 11.5 10.2 10.9 10.6 15.6 16.9 10.1 13.9 14.2 14.6 15.9 10.8 15.5 12.3 15.8 16.5 33.6 23.5 (°C) ature Temper------(feet level above Water- altitude NGVD 29) ------land (feet level Water Water below surface) ------Well Well (feet) depth Site number identification 343211093011501 343159092554601 343203092553901 343150092555301 343152092554101 343149092553101 343219092553601 343214092553501 343057092572101 343105092572001 343209092553601 343134092580501 343128092581901 343126092582601 343155092583301 343157092583701 343209092583101 343211092582301 343156093012101 343159093005301 343158092585501 343218092585501 342314093363301 343227093543501 S/cm, microsiemens per centimeter at 25 degrees Celsius; --, not measured; 330ARKS, Arkansas Novaculite; 330HSPG, Hot Springs S/cm, microsiemens per centimeter at 25 degrees Celsius; --, not measured; 330ARKS, Longitude µ 93.020833 92.92955 92.9274806 92.9312806 92.9280611 92.9253194 92.9266806 92.9262611 92.9557222 92.9559639 92.9267611 92.9679639 92.9720694 92.9738 92.9759389 92.9769056 92.9753611 92.9730556 93.0225278 93.0146389 92.9821111 92.9818611 93.6092444 93.9097222 Latitude 34.536389 34.533181 34.534100 34.530639 34.531150 34.530319 34.538600 34.537239 34.515706 34.516325 34.535831 34.526153 34.524481 34.523864 34.531900 34.532492 34.535694 34.536306 34.532250 34.532917 34.532778 34.538194 34.387161 34.407403 Well/ spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Spring Formation 364BGFK 330HSPG 364BGFK 350SLRN 364BGFK 364BGFK 364BGFK 364BGFK 350SLRN 330ARKS 330ARKS 330ARKS 364BGFK 364BGFK 364BGFK 364BGFK 330ARKS 330ARKS 364BGFK 330STNL 330ARKS 330ARKS 330ARKS 330ARKS Field name Spring Measurement site characteristics for Hot Springs well and spring reconnaissance.—Continued Sleepy Valley Spring Valley Sleepy Sangster Lane Two Sangster Jerryrig Spring Between Spring Java Spring Thresher North Thresher South Amphitheater Spring McClendon Spring Fall Spring Tree Powerline Spring Covenant Spring Simmerman Spring Valley Panther Spring Witness Seth Spring Sam Spring Montanus Spring Novastone Spring Thousand Spring Denise Spring Caddo Gap Spring Little Missouri

61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 Map number (figure 1) Appendix 2. Datum of 1929: °C, degrees Celsius; Vertical [NGVD 29, National Geodetic Shale] Womble 364BGFK, Bigfork Chert; 367WMBL, Sandstone; 330STNL, Stanley Shale; 350SLRN, Missouri Mountain Shale and Polk Creek Shale, undifferentiated; Publishing support provided by: Lafayette and Rolla Publishing Service Centers

For more information concerning the research described in the report:

U.S. Geological Survey Arkansas Water Science Center 401 Hardin Road Little Rock, AR 72211-3528 (501) 228-3600 http://ar.water.usgs.gov Kresse and Hays— Geochemistry, Comparative Analysis, and Physical and Chemical Characteristics of the Thermal Waters East of Hot Springs—SIR 2009-5263