Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated To

Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to

Conditions through 2002

e v Riverton i Salt Draper R Lake Valley eek Cr

Dry

15 Jordan Alpine Mountains River Fork

Highland American Traverse W Camp asatch Williams Cedar Hills Northern American Mt Timpanogos Utah 11,750 feet Eagle Lehi Fork Mountain Valley 89 Pleasant Grove Saratoga Springs River Lindon Provo Range

Vineyard Orem Cedar Valley

Mountains Provo

Utah Lake

Lake

Provo Bay

Springville

Hobble 15 Creek

Spanish

Spanish Fork Fork

est River

Mountain 89

Scientific Investigations Report 2006–5064

Prepared in cooperation with the Central Utah Water Conservancy District; Jordan Valley Water Conservancy District representing Draper City; Highland Water Company; Utah Department of Natural Resources, Division of Water Rights; and the municipalities of Alpine, American Fork, Cedar Hills, Eagle Mountain, Highland, Lehi, Lindon, Orem, Pleasant Grove, Provo, Saratoga Springs, and Vineyard

U.S. Department of the Interior U.S. Geological Survey Cover: View looking east toward American Fork Canyon in Utah County, Utah, March 2004. Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

By Susan A. Thiros

Scientific Investigations Report 2006–5064

U.S. Department of the Interior U.S. Geological Survey U.S. Department of the Interior Gale A. Norton, Secretary U.S. Geological Survey P. Patrick Leahy, Acting Director

U.S. Geological Survey, Salt Lake City, Utah: 2006

For additional information write to: U.S. Geological Survey Director, USGS Utah Water Science Center 2329 W. Orton Circle Salt Lake City, UT 84119-2047 Email: [email protected] URL: http://ut.water.usgs.gov/ 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. iii

Contents

Abstract...... 1 Introduction...... 1 Purpose and Scope...... 3 Description of the Study Area...... 3 Ground-Water Hydrology...... 3 Description of the Model...... 5 Update of the Model to Conditions through 2002...... 8 Performance of the Updated Model ...... 8 Potential Revisions and New Data to Improve the Updated Model ...... 15 Summary...... 19 References Cited...... 19

Figures

Figure 1. Location of northern Utah Valley study area, Utah...... 2 Figure 2. Generalized block diagram showing the basin-fill deposits and ground-water system in northern Utah Valley, Utah...... 4 Figure 3. Grid for the model of the ground-water system in northern Utah Valley, Utah...... 6 Figure 4. Location of cells simulating recharge and discharge in the model of the ground-water system in northern Utah Valley, Utah...... 7 Figure 5. Specified ground-water discharge from wells in the updated model of the ground-water system in northern Utah Valley, Utah, 1947–2002...... 9 Figure 6. Location of cells simulating flowing and pumping wells in 1981 and 2002 in the updated model of the ground-water system in northern Utah Valley, Utah...... 10 Figure 7. Annual streamflow in the American Fork and Provo Rivers and specified recharge to the ground-water system in the updated model of northern Utah Valley, Utah, 1947–2002...... 11 Figure 8. Location of wells with measured and computed water-level changes in the updated model of the ground-water system in northern Utah Valley, Utah...... 12 Figure 9. Measured water-level change for selected wells during 1981–2003 and computed water-level change for the corresponding model cell and layer in the updated model of the ground-water system in northern Utah Valley, Utah...... 13 Figure 10. Computed water-level decline from the end of 1980 to the end of 2001 in layer 5 of the updated model of the ground-water system and measured water-level decline from March 1981 to March 2002 at 14 wells that correspond to layer 5 in northern Utah Valley, Utah...... 16 Figure 11. Measured water-level altitude for selected wells during 1981–2003 and computed water-level altitude for the corresponding cell and layer in the updated model of the ground-water system in northern Utah Valley, Utah...... 17 Figure 12. Simulated and specified ground-water recharge and discharge in the updated model of the ground-water system in northern Utah Valley, Utah, 1947–2002...... 18 iv

Table

Table 1. Conceptual ground-water budget for the basin-fill aquifer system in northern Utah Valley, Utah...... 4

Conversion Factors and Datums

Multiply By To obtain Length foot (ft) 0.3048 meter (m) mile (mi) 1.609 kilometer (km)

Area square mile (mi2) 2.590 square kilometer (km2)

Volume acre-foot (acre-ft) 1,233 cubic meter (m3)

Flow rate acre-foot per year (acre-ft/yr) 1,233 cubic meter per year (m3/yr)

Vertical coordinate information is referenced to the North American Vertical Datum of 1929 (NAVD 29).

Horizontal coordinate information is referenced to the North American Datum of 1983 (NAD 83).

Altitude, as used in this report, refers to distance above the vertical datum. Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

By Susan A. Thiros

Abstract The ability of the model to adequately simulate climatic extremes such as the wetter-than-normal conditions of This report evaluates the performance of a numerical 1982–84 and the drier-than-normal conditions of 1999–2002 model of the ground-water system in northern Utah Valley, indicates that the annual variation of recharge to the ground- Utah, that originally simulated ground-water conditions during water system based on streamflow entering the valley, which 1947–1980 and was updated to include conditions estimated in turn is primarily dependent upon precipitation, is appropri- for 1981–2002. Estimates of annual recharge to the ground- ate but can be improved. The updated transient-state model water system and discharge from wells in the area were added of the ground-water system in northern Utah Valley can be to the original ground-water flow model of the area. improved by making revisions on the basis of currently avail- The files used in the original transient-state model able data and information. of the ground-water flow system in northern Utah Valley were imported into MODFLOW-96, an updated version of MODFLOW. The main model input files modified as part of Introduction this effort were the well and recharge files. Discharge from pumping wells in northern Utah Valley was estimated on an Ground water is the primary source of drinking water in annual basis for 1981–2002. Although the amount of average northern Utah Valley, Utah, and withdrawals for public supply annual withdrawals from wells has not changed much since have increased because of rapid population growth. Increased the previous study, there have been changes in the distribution withdrawals coupled with drought conditions during 1999– of well discharge in the area. Discharge estimates for flowing 2004 caused water levels in many wells in the area to decline wells during 1981–2002 were assumed to be the same as those to their lowest recorded levels (Burden and others, 2004, p. used in the last stress period of the original model because of a 40–41). Water-level declines may affect the ability of water lack of new data. Variations in annual recharge were assumed managers to withdraw water from public-supply wells or may to be proportional to changes in total surface-water inflow to affect the discharge to springs, drains, streams, Utah Lake, and northern Utah Valley. Recharge specified in the model during flowing wells in lower parts of the valley. The effects of with- the additional stress periods varied from 255,000 acre-feet in drawals and modifications to the current hydrologic system are 1986 to 137,000 acre-feet in 1992. not known, but need to be understood in order to manage and The ability of the updated transient-state model to match protect the ground-water resource. hydrologic conditions determined for 1981–2002 was evalu- The U.S. Geological Survey began a 4-year study of the ated by comparing water-level changes measured in wells to ground-water system in northern Utah Valley, Utah (fig. 1), in those computed by the model. Water-level measurements made 2003 in cooperation with the Central Utah Water Conservancy in February, March, or April were available for 39 wells in the District; Jordan Valley Water Conservancy District represent- modeled area during all or part of 1981–2003. In most cases, ing Draper City; Highland Water Company; Utah Depart- the magnitude and direction of annual water-level change ment of Natural Resources, Division of Water Rights; and the from 1981 to 2002 simulated by the updated model reason- municipalities of Alpine, American Fork, Cedar Hills, Eagle ably matched the measured change. The greater-than-normal Mountain, Highland, Lehi, Lindon, Orem, Pleasant Grove, precipitation that occurred during 1982–84 resulted in period- Provo, Saratoga Springs, and Vineyard. The objectives of this of-record high water levels measured in many of the observa- study are to develop a better understanding of the ground- tion wells in March 1984. The model-computed water levels at water system and to provide information to help determine the end of 1982–84 also are among the highest for the period. potential effects of withdrawals on water levels, water quality, Both measured and computed water levels decreased during and natural ground-water discharge in northern Utah Valley. the period representing ground-water conditions from 1999 to A new aerially expanded model of the ground-water system 2002. Precipitation was less than normal during 1999–2002. in northern Utah Valley is currently (2006) being constructed. 2 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

112°0' 111°35'

Riverton Salt Draper River Lake 40°30' Valley Creek

Dry

15 Jordan Alpine Mountains River Fork

Highland American Traverse Wasatch Camp Williams Cedar Hills Northern American Mt Timpanogos Utah 11,750 feet Eagle Lehi Fork Mountain Valley 89 Pleasant Grove Saratoga Springs River Lindon Provo Range

Vineyard Orem Cedar Valley

Mountains Provo

Utah Lake

Lake

Provo Bay

Springville

Hobble 15 Creek

Spanish

Spanish Fork Fork

River 40°5'

West

Mountain 89

0 2 4 6 8 10 Miles Great Salt 0 2 4 6 8 10 Kilometers Lake

EXPLANATION Study area Boundary of active cells in the model (Clark, 1984) of the ground-water system in northern Utah U T A H Valley

River Approximate boundary of basin-fill deposits

Colorado

Figure 1. Location of northern Utah Valley study area, Utah. Introduction

The original model (Clark, 1984) was updated and evalu- control gates for diversions are required to be fully opened ated as part of this study in order to determine how well it (compromise level), to about 5,160 ft at the highest level of simulated ground-water conditions during periods of climatic prehistoric Lake Bonneville deposits along the mountain sides. extremes and to use this information in the construction of the Mount Timpanogos in the adjacent Wasatch Range reaches an new model. Ground-water flow models can be used by water altitude of 11,750 ft and provides runoff to two major streams managers to help understand impacts to the ground-water that enter northern Utah Valley, the American Fork and Provo system from increased development, changing water use, and Rivers. changes in recharge. The population in northern Utah Valley increased from 170,000 in 1980 to 282,000 in 2000, a 66 percent change, and land is rapidly being converted from agricultural to urban uses Purpose and Scope to accommodate this growth. Prior to these recent changes in land use, the area had water-use patterns associated with This report evaluates the performance of a numerical agricultural diversions from streams and withdrawals from model of the ground-water system in northern Utah Valley, wells. Changing water-use patterns have the potential to Utah, that originally simulated conditions during 1947–1980 affect ground-water quantity and quality because the location (Clark, 1984) and was updated to include ground-water condi- of withdrawals may change and mountain-front streamflow tions estimated for 1981–2002. Estimates of annual recharge previously used for irrigation will likely be used for munici- to the ground-water system and discharge from wells in the pal supply. Recharge to the ground-water system will likely area were added to the original ground-water flow model of be reduced because of less seepage from irrigated fields and the area. The ability of the updated transient-state model to canals as agricultural areas are converted to residential areas. match hydrologic conditions determined for 1981–2002 was Mountain-front streams could become piped and diverted evaluated by comparing water-level changes measured in wells upstream from the valley for public supply, also resulting in to those computed by the model. This period includes both a less recharge to the ground-water system. wet and a dry sequence of years that resulted in the highest and lowest water levels measured in most wells with long periods of record in northern Utah Valley. Numerical models of Ground-Water Hydrology ground-water systems are constructed on the basis of available information and data, and if new data or information become The ground-water system in northern Utah Valley con- available, testing the performance of a model against those sists of aquifers contained in unconsolidated sediments of data will result in a better understanding of the model and the Tertiary and Quaternary age that have filled the basin between ground-water system (Konikow and Bredehoeft, 1992). The the surrounding mountains and in consolidated rock in the performance of the updated model provides information that mountains. The Wasatch Range and the Traverse and Lake can be used to guide the study for updating and improving the Mountains, which are composed primarily of fractured quartz- model and for additional data collection. This assessment also ite, limestone, and shale, receive varying amounts of recharge provides valuable information to users of the original model from precipitation. Almost all of the wells in the valley are while the new model is being developed. completed in the basin-fill deposits; therefore, it is considered the principal ground-water aquifer in the study area. Several wells have been recently completed in the consolidated rock Description of the Study Area along the margins of the valley to provide water for public supply. The study area covers the northern part of Utah Valley in Clark and Appel (1985) described the principal ground- the north-central part of Utah and corresponds to the extent water system in northern Utah Valley as consisting of three of basin-fill deposits (fig. 1). It includes the northern part of generally distinct aquifers consisting of predominantly Utah Lake, a natural, large (about 150 mi2), shallow (9.5 ft coarser-grained sediment separated by confining layers of clay average depth) lake in the lowest part of the valley. Northern (fig. 2). The first major aquifer is the shallow artesian aquifer Utah Valley is bounded on the north by the Traverse Moun- in deposits of Pleistocene age and is typically overlain by tains, on the east by the Wasatch Range, and on the west by blue clay about 50–100 ft below the valley surface that thins the Lake Mountains. The boundary between the northern and and pinches out near the valley margins. Another fine-grained southern parts of Utah Valley is arbitrarily located near Provo sequence separates the shallow from the deep artesian aquifer Bay. Ground water occurring north of this boundary generally of Pleistocene age at about 150 ft below land surface. A few discharges in northern Utah Valley and ground water occurring wells penetrate to depths greater than 500 ft below land sur- south of this boundary generally discharges in southern Utah face in the valley and are completed in the underlying deposits Valley. The boundary is poorly defined through Utah Lake of Quaternary/Tertiary age. Water levels in wells generally because of lack of data. indicate an upward gradient between the confined aquifers. The land-surface altitude of the basin-fill deposits in The confining layers become thin or discontinuous near the the area ranges from 4,489 ft, the level of Utah Lake set by mountain fronts, resulting in the basin-fill aquifers not being court decree as the maximum legal storage level, above which 4 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002 differentiated by depth and ground water occurring under Table 1. Conceptual ground-water budget for the basin-fill unconfined conditions. A shallow unconfined aquifer overlies aquifer system in northern Utah Valley, Utah (from Clark and the uppermost artesian aquifer and can be within a few feet of Appel, 1985, table 18) the land surface in the lower parts of the valley. Budget component Acre-feet Recharge to the basin-fill aquifers is from subsurface per year inflow from the surrounding consolidated rocks, in addition to losses from streams and canals (table 1) in the primary Recharge recharge area (fig. 2). Ground-water discharge is primarily to Seepage from natural channels and irrigation canals 73,000 wells, drains, and springs in and around Utah Lake. The differ- Seepage from irrigated fields, lawns, gardens, and 15,000 ence between the totals for recharge and discharge is mainly direct precipitation the result of insufficient data (Clark and Appel, 1985, p. 85) Subsurface inflow from mountains 112,000 and changes in the amount of ground water in storage. The value listed for each budget component represents an average Recharge total 200,000 annual amount. Data were not available to calculate average Discharge values for all components for the same time period. Wells 68,000 Drains, springs, and waterways 135,000 Diffuse seepage to Utah Lake 7,000 Evapotranspiration 8,000 Subsurface outflow through Jordan Narrows 2,000

Discharge total 220,000

EXPLANATION Direction of ground-water movement

Primary recharge area Wasatch Secondary recharge area Losing stream Discharge area Range

Utah Lake Spring Phreatophytes

Gaining stream Benches

Valley lowland Consolidated rock Potentiometric surface Confining layers

Unconfined aquifer in pre-Lake Bonneville Sediments of Deep deposits Quaternary/Tertiary Shallow artesian aquifer age artesian aquifer

Figure 2. Generalized block diagram showing the basin-fill deposits and ground-water system in northern Utah Valley, Utah. Description of the Model 5

Description of the Model charging at the land surface by evapotranspiration (fig. 4) with the Evapotranspiration Package (McDonald and Harbaugh, The U.S. Geological Survey developed a numerical flow 1988, p. 10–1). Discharge by evapotranspiration is dependent model in the early 1980s to simulate the ground-water system on depth to water and is, therefore, a head-dependent process. and flow in northern Utah Valley (Clark, 1984). The finite-dif- The model begins with steady-state conditions that were ference three-dimensional model was constructed using the assumed to exist in 1947, before the construction of large- MODFLOW program (McDonald and Harbaugh, 1988) and yielding pumped wells. Model-computed water levels were consisted of 7 layers, 36 rows, and 19 columns (fig. 3). The compared to measured water levels for 1947 and included 63 largest active cells are 0.85 mi2 and generally included areas water levels in layer 3, 48 in layer 5, and 9 in layer 7 (Clark, where data were sparse. The smallest cells are 0.25 mi2 and 1984, p. 22). Most of the model-computed water levels are generally included areas with wells, ground-water withdraw- within 5 ft of those measured in wells; however, in areas with als, or historic water-level measurements. few data, the difference may be as much as 10 ft. Impermeable boundaries were used to simulate the The transient-state calibration of the model consisted contact between the valley and mountains in the model area. primarily of varying aquifer properties along with recharge The impermeable boundaries are represented in the model as estimates and discharge data from wells for 1947–80 and inactive cells (fig. 3). The approximate location of the contact comparing the model-computed water levels with water levels between the basin-fill deposits and consolidated rock is along measured during the late winter-early spring of 1948–81. The the eastern Traverse Mountains and the western flank of the transient period was divided into seven stress periods repre- Wasatch Range. Inactive cells also were placed about one cell senting intervals of time when total discharge from wells was lakeward from the eastern shoreline of Utah Lake and beneath fairly constant. The periods with similar pumping rates are layer 7, on the assumption that there is no upward flow from 1947–50, 1951–55, 1956–62, 1963–65, 1966–73, 1974–77, below. and 1978–80. Simulating fluctuations in well discharge dur- Simulated recharge to the ground-water system included ing model calibration from one multiyear stress period to the seepage from streams, irrigation canals, irrigated fields, next did not result in a reasonable match between water-level lawns, and gardens; infiltration of precipitation; and subsur- changes computed in the model and water-level changes mea- face inflow from consolidated rocks in the primary recharge sured during the calibration period (Clark, 1984). Therefore, area (fig. 2), a narrow strip of land adjacent to the mountain recharge also was varied to provide a better match between fronts (Clark and Appel, 1985, fig. 9) that is not underlain by model-computed and measured water-level changes. thick layers of fine-grained material that impede the down- Variations in annual recharge were assumed to be propor- ward movement of water. Recharge was distributed to the tional to changes in total surface-water inflow to northern Utah cells near the eastern extent of the valley according to an Valley. The initial rate of recharge obtained from the steady- initial rate determined for steady-state conditions (fig. 4) and state calibration was multiplied by one-half of the percentage was specified by Clark (1984) as rates input to the Recharge change from the average surface-water inflow during a given Package (McDonald and Harbaugh, 1988, p. 7–1). Ground- time period (Clark, 1984, p. 27) and applied to all of the model water recharge from subsurface inflow along the mountain cells that simulate recharge. front calculated by the model using constant-head cells under steady-state conditions was about 100,000 acre-ft/yr. The Multiplier = [(SW - SWave) / SWave] x 0.5 (1) constant-head cells were replaced with specified recharge rates where: SW is the estimated annual surface-water inflow at the end of the steady-state calibration. The total amount of to the model area, in acre-ft and recharge to the ground-water system simulated under steady- SWave is the estimated average surface-water state conditions was 190,000 acre-ft/yr. inflow to the model area, in acre-ft. Constant-head cells were used along the Jordan River and the eastern shoreline of Utah Lake to simulate ground-water For example, if surface-water inflow for a given time discharge by upward leakage from the artesian aquifers (fig. period was 20 percent above average, then recharge for that 4). They are placed in layers 1, 2, and 3 at approximately the time period was assumed to be 10 percent above the ini- river or the compromise lake-surface altitude. Ground-water tial rate. Recharge to the ground-water system may change discharge from springs, seeps, and drains was simulated with substantially from one year to the next, and such changes are the Drain Package (McDonald and Harbaugh, 1988, p. 9–1). a major cause for variations of water levels in northern Utah The Drain Package allows water to discharge from the aquifer Valley. at a rate proportional to the difference between the head in the Clark (1984, fig. 16) compared measured and model- aquifer and some fixed altitude when the head in the aquifer is computed water-level changes for 16 observation wells with above that altitude. Discharge stops if the head falls below the data for some of or all the pumping periods. The computed specified altitude for that cell. water levels were close to the measured levels at most of the Some water moving from model layers 3, 5, and 7, wells. At wells where the computed levels did not match the representing the artesian aquifers, to layer 1, which represents measured levels, the magnitude of the water-level changes the upper unconfined aquifer, was simulated as eventually dis- from one pumping period to the next were generally about the same. 6 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

18 17 16 15 14 13 12 11 10 9 8 COLUMNS7 6 5 4 3

9 10

18 ROWS 19

28 EXPLANATION 29

30 Cells inactive in all layers 31 Cells active in layers 1–7 32

Cells active in layers 3–7 33

34 Cells active in layers 4–7 35

Utah Lake 36

Figure . Grid for the model of the ground-water system in northern Utah Valley, Utah. Description of the Model

18 17 16 15 14 13 12 11 10 9 8 COLUMNS7 6 5 4 3

9 10

18 ROWS 19

25 EXPLANATION 26 27

Cells simulating recharge—Amount of 28

recharge specified at the end of steady-state 29 calibration, in hundreds of acre-feet per year 2 to 5 30 6 to 10 31 11 to 20 32 21 to 50 33 34 51 to 100 35 100 to 142 36 Cells simulating discharge Constant-head boundary Drains Evapotranspiration Cells active in one or more layers—No recharge applied Utah Lake

Figure 4. Location of cells simulating recharge and discharge in the model of the ground-water system in northern Utah Valley, Utah. Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

Update of the Model to Conditions shown in figure 5. The location of model cells containing the wells simulated in 1981 and 2002 is shown in figure 6. through 2002 Discharge to wells is represented as a specified-flux boundary using the Well Package (McDonald and Harbaugh, The files used in the transient-state model (Clark, 1984) 1988, p. 8–1). Required input data for the Well Package are the were imported into MODFLOW-96 (Harbaugh and McDon- model cell (row and column) and layer that correspond to the ald, 1996) for use in updating the model to conditions through location and depth of the open intervals (perforated, screened, 2002. MODFLOW-96 is an updated version of MODFLOW or open hole) of the well. The distribution of discharge from a (McDonald and Harbaugh, 1988). The main model input files pumped well open to more than one model layer representing modified as part of this effort were the well and recharge files. an artesian aquifer was the same as that used in the original Other parts of the updated model described in the remainder model. Discharge from wells drilled after 1980 that were open of this report are the same as in the original model (1947–80), to more than one model layer simulating artesian aquifers except for factors related to the extended simulation period. was divided evenly between the layers on the basis of similar The number of stress periods listed in the Basic Package file distribution of transmissivity across the layers (Clark, 1984, p. was changed from 7, representing the period from 1947–80 15–17). in the original model, to 29, which includes an additional 22 Recharge estimates for 1981–2002 were included in annual stress periods representing the period from 1981–2002. the updated model on the basis of annual change from aver- Other input files were adjusted to account for the additional age surface-water inflow to the area according to the method stress periods. Hydraulic properties of the aquifers and con- described by Clark (1984, p. 27). Annual streamflow totals fining layers, the distribution of recharge, the use of head- determined for the American Fork River (U.S. Geological Sur- dependent cells to simulate ground-water discharge to springs, vey gaging station 10164500) and the Provo River below Deer drains, and evapotranspiration, and the use of constant-head Creek Dam (U.S. Geological Survey gaging station 10159500) cells to simulate discharge to Utah Lake, the Jordan River, were added to estimates of flow from the other streams enter- and subsurface outflow through the Jordan Narrows were not ing the valley. Streamflow estimates for the ungaged streams changed in the updated model. were calculated by correlating flow with that of gaged streams Discharge from pumping wells in northern Utah Valley along the Wasatch Range in Salt Lake Valley to the north. The was estimated on an annual basis for 1981–2002. Annual with- amount of streamflow determined by this method does not drawals from irrigation wells were estimated from unpublished distinguish between water that was diverted to or from the data in the files of the U.S. Geological Survey Utah Water Sci- drainage basin and water that originated in or drained to the ence Center office in Salt Lake City, Utah. Public-supply and basin. Recharge specified in the model during the additional industrial well withdrawal data were obtained from the Utah stress periods varied from 255,000 acre-ft in 1986 to 137,000 Division of Water Rights (written commun., 2003). Withdraw- acre-ft in 1992 (fig. 7). als from wells in the area were estimated to be about 72,500 acre-ft/yr in 2002 and averaged about 72,800 acre-ft/yr during 1999–2002. Although the amount of average annual withdrawals from wells has not changed much from what was reported Performance of the Updated Model in the previous study (table 1), there have been changes in the distribution of well discharge in the area. Since 1981, more The ability of the updated transient-state model to match than 40 large-yield wells have been constructed to supply the hydrologic conditions determined for 1981–2002 was evalu- growing residential areas while many flowing wells used for ated by comparing water-level changes measured in wells to irrigation, stock, and domestic use have been destroyed or are those computed by the model. Model performance was not no longer in use. evaluated for fluctuations in natural ground-water discharge Records of well discharge prior to 1963 are less com- because few observations of changes in flow were available plete than for the later periods and flowing-well discharge is for 1981–2002. Water-level measurements made in Febru- assumed to be a major part of the discharge from wells before ary, March, or April were available for 39 wells in the mod- 1963 (Clark, 1984, p. 29). Discharge estimates for flowing eled area during all or part of 1981–2003 (fig. 8). Most of wells during 1981–2002 were assumed to be the same as those the water-level measurements were made during the month used in the last stress period of the original model because of of March, a time of the year when there is less stress on the a lack of new data. Because discharge from flowing wells for ground-water system from pumping, evapotranspiration, snow- irrigation likely has decreased since 1980 as a result of of the melt runoff, or irrigation. This water-level measurement was change from agricultural to residential land use in the area, used to represent the cumulative effects from the preceding this assumption results in a higher-than-actual discharge esti- year on ground-water conditions at each well. mate for flowing wells during 1981–2002. Additional study of The performance of the updated model can be assessed land-use changes would likely provide an improved estimate. by comparing how closely it simulated the water-level change Estimated ground-water withdrawals from wells used in the since 1981. The measured water-level change at a well updated model for each stress period from 1947 to 2002 are was determined by subtracting the water level measured in Performance of the Updated Model

80,000

Total specified discharge from wells Specified discharge from flowing wells 70,000 Specified discharge from pumping wells

60,000

50,000

40,000

30,000 ACRE-FEET PER YEAR ACRE-FEET PER

20,000

10,000

0 1950 1960 1970 1980 1990 2002 1955 1965 1975 1985 1995 2000 1947 Figure 5. Specified ground-water discharge from wells in the updated model of the ground-water system in northern Utah Valley, Utah, 1947–2002.

March 1981, or the first available March measurement after in large increases in model-computed water levels during the 1981, from water levels measured in the well during Febru- wet years of 1982–84. The greater-than-normal precipitation ary, March, or April (preferably March) 1981–2003. These that occurred during 1982–84 resulted in period-of-record measured water-level changes were compared to water-level high water levels measured in many of the observation wells changes computed at the end of the preceding calendar year in March 1984. The model-computed water levels at the end by the model for the cell and layer that correspond to the well of 1982–84 also are among the highest for the period and (fig. 8). The model-computed water-level change was deter- are only exceeded by the water levels computed at the end of mined by subtracting the water level computed at the end of 1986, the year with the most recharge simulated. 1980, which corresponds to the water-level measurement made The water level measured in March 1987 is lower than in March 1981 (or the computed water level that corresponded the water level measured in March 1984 in all but one of the to the first available March measurement after 1981), from wells with measurements (well 27). The high model-com- water levels computed by the model at the end of 1980–2002. puted water levels at the end of 1986 indicate some error in For example, the measured water-level change from March the amount of recharge simulated in 1986. The total stream- 1981 to March 2003 was compared to the model-computed flow measured in the American Fork River in 1986 is similar change from the end of 1980 to the end of 2002. This method (within 8 percent) to the average of what was measured during focuses on water-level change, so that estimated water levels 1982–84 (fig. 7). The American Fork River drains the Wasatch used in the calibration of the original model in areas with few Range adjacent to the northern part of Utah Valley. The total or no data do not overwhelm the comparison. streamflow measured in the Provo River below Deer Creek In most cases, the magnitude and direction of annual Dam was 36 percent more in 1986 than the average total water-level change from 1981 to 2002 computed by the streamflow during 1982–84 (fig. 7). Streamflow at the Provo updated model reasonably matched the measured change River below Deer Creek Dam is regulated and includes water (fig. 9). The similar response of annual model-computed diverted from outside of the Wasatch Range drainage basin. water-level change and measured water-level change with Recharge to the ground-water system specified to the model is time indicates that the variation in the amount of recharge influenced by factors that are not considered in the simulated specified annually is generally a good approximation of the correlation between annual recharge and total surface-water actual variation in recharge. Recharge estimates determined inflow (eq. 1). from total annual surface-water inflow to the valley resulted 10 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

18 17 16 15 14 13 12 11 10 9 8 COLUMNS7 6 5 4 3

9 10

18 ROWS 19

29 EXPLANATION 30 Cells active in one or more layers 31 32 Cells simulating flowing wells from 1981 through 2002 33 Utah Lake 34 Boundary of cells simulating flowing wells from 1981 35

through 2002 36 Wells pumping from layer 3 in 1981 Wells pumping from layer 5 in 1981 Wells pumping from layer 7 in 1981 Wells pumping from layer 3 in 2002 Wells pumping from layer 5 in 2002 Wells pumping from layer 7 in 2002

Figure 6. Location of cells simulating flowing and pumping wells in 1981 and 2002 in the updated model of the ground-water system in northern Utah Valley, Utah. Performance of the Updated Model 11

600,000

American Fork River (U.S. Geological Survey streamflow gaging station 10164500) Provo River below Deer Creek Dam (U.S. Geological Survey streamflow gaging station 10159500, period of record begins in 1954) Specified ground-water recharge 500,000

400,000

300,000 ACRE-FEET PER YEAR ACRE-FEET PER 200,000

100,000

0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 1947 2002

Figure . Annual streamflow in the American Fork and Provo Rivers and specified recharge to the ground-water system in the updated model of northern Utah Valley, Utah, 1947–2002.

Both measured and model-computed water levels are not as large as measured water-level declines from March decreased during the period representing ground-water condi- 2000 to March 2002, a period with drier-than-normal conditions from 1999 to 2002. The water levels measured in March tions, in wells 2, 13, 14, 15, and 16 (fig. 9). This is a discharge 2003 were the lowest measured at that time of the year for the area where wells have historically flowed, but pumping has period of record for many of the wells. Precipitation measured increased substantially from when the original model was at Silver Lake near Brighton, Utah, about 10 mi northeast calibrated. Also, more recharge from the constant-head cells of the study area and at an altitude of about 8,740 ft in the occurs during periods with less overall recharge because the Wasatch Range, was less than normal during 1999–2002. This simulated water level drops below that of the constant-head drought is comparable in length and magnitude to previous boundary. Model parameters used to simulate aquifer proper- droughts since 1965; however, with population growth and ties, such as storage coefficient and transmissivity, discharge increased demand for water in the area, the general effect is and recharge from constant-head cells, and the specified dis- more severe. The water-level declines at the end of 2002 were charge from flowing wells may need adjustment in this area. about the largest computed and were similar to those at the Measured water-level changes in the eastern part of the end 1992 (fig. 9), the year with the least amount of applied modeled area, in or near where recharge was specified and recharge in the updated model. ground water is unconfined, generally did not match computed Water-levels declines computed at the end of 2001 and water-level changes as well as in other areas. Differences compared to water-level declines measured at 37 wells in between measured and computed water-level change exceeded March 2002 ranged from 11.3 ft less than declines measured 20 ft in wells 3, 4, and 29 at times in the 1990s (fig. 9). These at well 14 to 5.5 ft greater than declines measured at well wells are near American Fork Canyon and the measured 33 (fig. 9). Model-computed water-level changes at the end water-level changes are likely affected by the timing of runoff of 2001 were within 5 ft of measured water-level changes at and ground-water recharge occurring at and near the moun- 70 percent of the wells with water-level measurements from tain front. Model-computed water-level changes in the area March 2002. Measured water-level declines in many of the are affected more by the large influence of annual streamflow wells in the northwestern part of the modeled area near Utah in the Provo River on the recharge multiplier and less on the Lake are larger than those computed by the model for a corre- actual variation in annual streamflow in the American Fork sponding cell (fig. 10). Model-computed water-level declines River (fig. 7). 12 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

18 17 16 15 14 13 12 11 10 9 8 COLUMNS7 1 6 5 4 4 3 3

2 5

1 7

6 8 29 11 1 13 12 2 10 9 2 18 23 14 24 31 17 22 30 3 16 15 19 20 25 21 28 4 26

5 6 27 7

9 10

11 33

12 32 13

18 35 34 ROWS 19

22 37 36 23 38 24 39

29 EXPLANATION 30 Cells active in one or more layers 31 32 Utah Lake 33 Wells with annual water-level measurements available 34 for all or part of 1981–2003—number is the well number shown on figures 9 and 11 35 Well depth corresponds to layer 3 36 Well depth corresponds to layer 5 Well depth corresponds to layer 7

Figure . Location of wells with measured and computed water-level changes in the updated model of the ground-water system in northern Utah Valley, Utah. Performance of the Model 1

10 20 1* 9 0 10 0 -10 -10 -20 Measured (C-4-1)25dcb-1 -20 Measured (D-5-1)16ccb-4 Computed Row 1, column 8, layer 7 Computed Row 8, column 7, layer 7 -30 -30 5 20 0 2 10 10

-5 0

-10 -10

WATER-LEVEL CHANGE, IN FEET WATER-LEVEL -15 Measured (C-5-1)13dac-5 -20 Measured (D-5-1)17acb-5 Computed Row 5, column 4, layer 5 Computed Row 7, column 7, layer 5 -20 -30 30 20 20 10 11 10 3* 0 0 -10 -20 -10 -30 Measured (D-4-1)36cab-1 -20 Measured (D-5-1)17adc-12 -40 Computed Row 7, column 16, layer 7 Computed Row 7, column 7, layer 7 -50 -30

30 10 20 4* 5 12 10 0 0 -5 -10 -10 -20 -15 WATER-LEVEL CHANGE, IN FEET WATER-LEVEL -30 Measured (D-4-2)31abd-1 Measured (D-5-1)18acb-1 -20 -40 Computed Row 8, column 18, layer 7 Computed Row 6, column 5, layer 5 -50 -25 10 10 5* 5 13 0 0 -5 -10 -10 -15 -20 Measured (D-5-1)6bcd-1 Measured (D-5-1)18cab-2 Computed Row 3, column 7, layer 7 -20 Computed Row 6, column 4, layer 7 -30 -25 10 10 6 5 14 0 0 -5 -10 -10 -20 -15 WATER-LEVEL CHANGE, IN FEET WATER-LEVEL Measured (D-5-1)8dcc-1 Measured (D-5-1)19bcb-2 Computed Row 6, column 7, layer 5 -20 Computed Row 7, column 4, layer 7 -30 -25 20 10 10 Measured (D-5-1)10bdd-3 7* 5 Computed Row 8, column 11, layer 5 15* 0 0 -5 -10 -10 -20 -15 -30 Measured (D-5-1)19ccb-1 -20 Computed Row 8, column 3, layer 5 -40 -25 20 10 Measured (D-5-1)13aaa-2 10 5 Computed Row 12, column 14, layer 5 8* 16 0 0 -5 -10 -10 -20 -15 WATER-LEVEL CHANGE, IN FEET WATER-LEVEL -30 Measured (D-5-1)19ccc-1 -20 Computed Row 7, column 3, layer 5 -40 -25 1981 1985 1990 1995 2000 2003 1981 1985 1990 1995 2000 2003

* water-level altitude for this well is shown in figure 11 Measured water-level change at well (See individual graph) Computed water-level change at model cell (See individual graph)

Figure . Measured water-level change for selected wells during 1981–2003 and computed water-level change for the corresponding model cell and layer in the updated model of the ground-water system in northern Utah Valley, Utah. 14 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

20 10 10 17 5 25 0 0 -5 -10 -10 -20 -15 Measured—(D-5-1)20aba-1 Measured—(D-5-1)24ddd-4 -30 Computed—Row 8, column 6, layer 7 -20 Computed—Row 15, column 12, layer 3 -40 -25 20 10 10 18 5 26 0 0 -5 -10 -10 -20 -15 WATER-LEVEL CHANGE, IN FEET WATER-LEVEL Measured—(D-5-1)20aba-2 -30 Measured—(D-5-1)25aaa-3 Computed—Row 8 column 6, layer 5 -20 Computed—Row 15, column 12, layer 3 -40 -25 20 10 10 19 5 27 0 0 -5 -10 -10 -20 -15 Measured—(D-5-1)25ccd-1 -30 Measured—(D-5-1)20cbc-1 Computed—Row 9, column 4, layer 5 -20 Computed—Row 16, column 11, layer 5 -40 -25 10 10 5 20 5 28 0 0 -5 -5 -10 -10

WATER-LEVEL CHANGE, IN FEET WATER-LEVEL -15 -15 Measured—(D-5-1)21dba-2 Measured—(D-5-1)27aac-1 -20 Computed—Row 10, column 7, layer 3 -20 Computed—Row 13, column 8, layer 3 -25 -25 20 20 10 21 10 29* 0 0 -10 -10 -20 -20 Measured—(D-5-1)21dda-2 Measured—(D-5-2)18aba-1 -30 -30 Computed—Row 11, column 7, layer 5 Computed—Row 13, column 16, layer 5 -40 -40 10 20 5 22 10 30 0 0 -5 -10 -10 -20 -15

WATER-LEVEL CHANGE, IN FEET WATER-LEVEL Measured—well (D-5-1)22bcc-1 -30 Measured—(D-5-2)19dca-1 -20 Computed—Row 10, column 8, layer 3 Computed—Row 15, column 14, layer 3 -25 -40 20 20 10 Measured—(D-5-1)22bcc-2 10 Computed—Row 10, column 8, layer 5 23 31 0 0 -10 -10 -20 -20 -30 -30 Measured—(D-5-2)21cba-1 Computed—Row 16, column 16, layer 7 -40 -40 20 20 10 24 10 32* 0 0 -10 -10 -20 -20

WATER-LEVEL CHANGE, IN FEET WATER-LEVEL -30 Measured—(D-5-1)23dab-3 -30 Measured—(D-6-2)9ccc-1 Computed—Row 13, column 11, layer 7 Computed—Row 22, column 11, layer 5 -40 -40 1981 1985 1990 1995 2000 2003 1981 1985 1990 1995 2000 2003

* water-level altitude for this well is shown in figure 11 Measured water-level change at well (See individual graph) Computed water-level change at model cell (See individual graph)

Figure . Measured water-level change for selected wells during 1981–2003 and computed water-level change for the corresponding model cell and layer in the updated model of the ground-water system in northern Utah Valley, Utah—Continued. Potential Revisions and New Data to Improve the Updated Model 15

20 5 Measured—(D-6-2)12bdb-1 10 33* 37 Computed—Row 24, column 17, layer 7 0 0 -5 -10 -10 -20 Measured—(D-7-2)4cbc-1 Computed—Row 29, column 5, layer 5 -30 -15 20 5 34* 38 10 0 0 -5 -10

WATER-LEVEL CHANGE, IN FEET WATER-LEVEL -10 -20 Measured—(D-6-2)26cca-1 Measured—(D-7-2)10adc-3 Computed—Row 28, column 11, layer 3 Computed—Row 31, column 7, layer 3 -30 -15 10 5

5 35 39 0 0 -5 -5 -10 -10 Measured—(D-6-2)28bad-1 Measured—(D-7-2)11caa-1 Computed—Row 25, column 9, layer 3 Computed—Row 32 column 8, layer 3 -15 -15 1981 1985 1990 1995 2000 2003 10 5 36 Measured water-level change at well (See individual graph) 0 Computed water-level change at model cell (See individual graph)

-5 WATER-LEVEL CHANGE, IN FEET WATER-LEVEL * water-level altitude for this well is shown in figure 11 -10 Measured—(D-7-2)3cca-1 Computed—Row 30, column 7, layer 3 -15 1981 1985 1990 1995 2000 2003

Eleven wells with water levels measured during and constant-head cells and decreased withdrawals from wells. 1981–2003 did not have water-level measurements made in The difference between total recharge and total discharge for March prior to 1981 (fig. 11) and, therefore, were not used a year is water added to or removed from storage. Predic- to calibrate the original model. The updated model computed tive simulations of interest would include a period with less a higher water-level altitude for the cell and layer that corre- recharge coupled with increased withdrawals from wells sponded to wells 1, 5, 29, and 33 than was measured, with off- greater than the amounts to date. sets generally more than 20 ft (fig. 11). These wells are located near model boundaries and indicate that the model needs some refinement in these areas. Water levels measured at wells 3, 4, 7, 8, 15, 32, and 34 generally were similar to those computed Potential Revisions and New Data to by the updated model for the period representing ground-water Improve the Updated Model conditions during 1981–2002, indicating that the starting heads used in the original model were a close approximation. The updated transient-state model of the ground-water The ability of the model to adequately simulate cli- system in northern Utah Valley can be improved by making matic extremes such as the wetter-than-normal conditions of revisions on the basis of currently available data and informa- 1982–84 and the drier-than-normal conditions of 1999–2002 tion. The addition of many wells completed at different depths indicates that the annual variation of recharge to the ground- in the study area provides water levels in areas where informa- water system based on streamflow entering the valley, which tion was not available prior to 1981. Revised water levels for in turn is primarily dependent upon precipitation, is appropri- the model representing steady-state conditions would reduce ate but can be improved. The model’s response to years with the offset between measured and computed water levels. Tran- less specified recharge and increased withdrawals from wells sient-state calibration to water levels measured during peri- is a decrease in discharge from drains simulating springs, ods of climatic extremes, such as during the 1980s and early constant-head cells simulating seepage to the Jordan River and 2000s, and to increased withdrawals in some areas, would Utah Lake, and evapotranspiration (fig. 12). Years with more increase the range in which the model was tested. recharge generally correspond to more discharge from drains 16 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

18 17 16 15 14 13 12 11 10 9 8 COLUMNS7 6 5 4 3

5 6

9 10

18 ROWS 19

EXPLANATION 29

Computed water-level decline—From the end of 30 1980 to the end of 2001 in model layer 5, in feet 2.1 to 9.9 31 10.0 to 19.9 32 20.0 to 29.9 33 34 30.0 to 35.0 35

Utah Lake 36 Measured water-level decline—From March 1981 to March 2002 in wells that correspond to model layer 5, in feet 12.3 to 18.9 21.5 to 28.9

Figure 10. Computed water-level decline from the end of 1980 to the end of 2001 in layer 5 of the updated model of the ground-water system and measured water-level decline from March 1981 to March 2002 at 14 wells that correspond to layer 5 in northern Utah Valley, Utah. Potential Revisions and New Data to Improve the Updated Model 1

4,600 4,600 1 15 4,580 4,580 Measured (D-5-1)19ccb-1 Computed Row 8, column 3, layer 5 4,560 4,560

4,540 4,540

4,520 Measured (C-4-1)25dcb-1 4,520 Computed Row 1, column 8, layer 7 4,500 4,500 4,650 4,650

4,630 3 4,630 29

4,610 4,610

4,590 4,590

4,570 Measured (D-4-1)36cab-1 4,570 Measured (D-5-2)18aba-1 Computed Row 7, column 16, layer 7 Computed Row 13, column 16, layer 5 4,550 4,550 4,670 4,600 4,650 4 4,580 32 4,630 4,560

4,610 4,540

WATER-LEVEL ALTITUDE, IN FEET ALTITUDE, WATER-LEVEL 4,590 Measured (D-4-2)31abd-1 4,520 Measured (D-6-2)9ccc-1 Computed Row 8, column 18, layer 7 Computed Row 22, column 11, layer 5 4,570 4,500

4,600 4,600

4,580 5 4,580 33

4,560 4,560

4,540 4,540

4,520 Measured (D-5-1)6bcd-1 4,520 Measured (D-6-2)12bdb-1 Computed Row 3, column 7, layer 7 Computed Row 24, column 17, layer 7 4,500 4,500 4,650 4,600

4,630 7 4,580 34

4,610 4,560

4,590 4,540

4,570 Measured (D-5-1)10bdd-3 4,520 Measured (D-6-2)26cca-1 Computed Row 8, column 11, layer 5 Computed Row 28, column 11, layer 3 4,550 4,500 4,650 1947 1950 1960 1970 1980 1990 2000 2003 4,630 8

4,610

4,590 Measured water-level altitude at well (See individual graph) Computed water-level altitude at model cell (See individual graph) Measured (D-5-1)13aaa-2 WATER-LEVEL ALTITUDE, IN FEET ALTITUDE, WATER-LEVEL 4,570 Computed Row 12, column 14, layer 5 4,550 1947 1950 1960 1970 1980 1990 2000 2003

Figure 11. Measured water-level altitude for selected wells during 1981–2003 and computed water-level altitude for the corresponding cell and layer in the updated model of the ground-water system in northern Utah Valley, Utah.

A refinement to the model that could improve its rep- average recharge would increase discharge from flowing wells resentation of the conceptual ground-water system is to vary in some areas. Simulating fluctuation in the discharge from discharge from flowing wells on the basis of land use and flowing wells on the basis of the amount of irrigated land and water-level changes over time. Discharge from flowing wells on water levels could provide a more reasonable estimate of likely has decreased since 1980 because of changes in land use discharge than the original method of increasing flowing-well from agricultural to residential and a corresponding decrease discharge proportionally with withdrawals from pumping in water use from flowing irrigation wells. Many flowing wells wells to satisfy demands for irrigation. used for irrigation are no longer used or have been abandoned The ground-water flow model of northern Utah Valley because of an ongoing decrease in agricultural acreage. Dis- could be improved by updating it to operate with newer simu- charge from flowing wells also is affected by the water level in lation methods, such as MODFLOW-2000 (Harbaugh and the aquifer: increased withdrawals from pumping wells have others, 2000; Hill and others, 2000). The parameter estimation lowered the artesian pressure resulting in the potentiometric package of MODFLOW-2000 uses a more efficient and objec- surface dropping below land surface during the irrigation tive method of calibration and provides a quantitative assess- season in some areas. Water-level rises caused by greater-than- ment of model uncertainty. 1 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

300,000 Simulated discharge from evapotranspiration Simulated discharge from constant-head cells Simulated discharge from drains Total specified discharge from wells 250,000 Specified recharge Simulated recharge from constant-head cells

200,000

150,000 ACRE-FEET PER YEAR 100,000

50,000

0 1947 1950 1965 1985 1955 1960 1970 1975 1980 1990 1995 2000 2002

Figure 12. Simulated and specified ground-water recharge and discharge in the updated model of the ground-water system in northern Utah Valley, Utah, 1947–2002.

Ground-water discharge data in addition to water-level The original ground-water flow model did not include data are important to constrain ground-water flow models. the basin-fill deposits west of the Jordan River and Utah Lake. Information on discharge from springs, seeps, and drains near Ground-water development is occurring in that area, includ- Utah Lake measured during the modeled transient period and ing production wells completed in the consolidated rock west incorporated into the calibration process, would further con- of and underneath the basin-fill deposits, outside of the extent strain the model. of active cells in the original model. These wells, along with More wells have been completed in the aquifer contained recent geologic mapping of the area, can provide more infor- within the Quaternary/Tertiary-age sediments (model layer mation on the extent and thickness of the model layers west 7) in the modeled area since the completion of the original of the Jordan River and Utah Lake. The extent of the original ground-water flow model. Information about aquifer proper- model grid would have to be expanded to include this informa- ties such as hydraulic conductivity, transmissivity, and storage tion. coefficient estimated from these wells and used to simulate The current study (2003–06) of the ground-water system layer 7 could improve the model. in northern Utah Valley is developing a new, aerially expanded Environmental-tracer data, such as tritium, helium-3, and ground-water flow model for the area that incorporates these helium-4 concentrations, collected as part of the larger study, suggestions and data. This new model will serve as a tool are being used to estimate residence time in the ground-water to help understand the effects of increased development and system along selected flow paths. The simulated time of travel changing water use on the ground-water resources. to a cell and layer can be compared to estimates of ground- water ages for selected locations and aquifers in the calibration process using MODPATH (Pollock, 1994). MODPATH is a particle-tracking postprocessing package that was developed to compute three-dimensional flow paths using output from steady-state or transient ground-water flow simulations by MODFLOW. Summary 1

Summary used to represent the cumulative effects from the preceding year on ground-water conditions at each well. The U.S. Geological Survey began a 4-year study of the In most cases, the magnitude and direction of annual ground-water system in northern Utah Valley, Utah, in 2003 in water-level change from 1981 to 2002 simulated by the cooperation with the Central Utah Water Conservancy District; updated model reasonably matched the measured change. The Jordan Valley Water Conservancy District representing Draper greater-than-normal precipitation that occurred during 1982– City; Highland Water Company; Utah Department of Natural 84 resulted in period-of-record high water levels measured Resources, Division of Water Rights; and the municipalities of in many of the observation wells in March 1984. The model- Alpine, American Fork, Cedar Hills, Eagle Mountain, High- computed water levels at the end of 1982–84 also are among land, Lehi, Lindon, Orem, Pleasant Grove, Provo, Saratoga the highest for the period. Both measured and computed water Springs, and Vineyard. This report is one component of the levels decreased during the period representing ground-water study that evaluates the performance of a numerical model of conditions from 1999 to 2002. Precipitation was less than the ground-water system in northern Utah Valley that origi- normal during 1999–2002. nally simulated ground-water conditions during 1947–1980 The ability of the model to adequately simulate cli- and was updated to include conditions estimated for 1981– matic extremes such as the wetter-than-normal conditions of 2002. Estimates of annual recharge to the ground-water system 1982–84 and the drier-than-normal conditions of 1999–2002 and discharge from wells in the area for 1981–2002 were indicates that the annual variation of recharge to the ground- added to the original ground-water flow model of the area. water system based on streamflow entering the valley, which The files used in the original transient-state model of in turn is primarily dependent upon precipitation, is appropri- the ground-water flow system in northern Utah Valley were ate but can be improved. The updated transient-state model imported into MODFLOW-96, an updated version of MOD- of the ground-water system in northern Utah Valley can be FLOW. The main model input files modified as part of this improved by making revisions on the basis of currently avail- effort were the well and recharge files. Other parts of the able data and information. updated model are the same as in the original model, except for factors related to the extended simulation period. Dis- charge from pumping wells in northern Utah Valley was esti- References Cited mated on an annual basis for 1981–2002. Although the amount of average annual withdrawals from wells has not changed Burden, C.B., and others, 2004, Ground-water conditions much since the previous study, there have been changes in the in Utah, spring of 2004: Utah Department of Natural distribution of well discharge in the area. Discharge estimates Resources Cooperative Investigations Report No. 45, 120 p. for flowing wells during 1981–2002 were assumed to be the same as those used in the last stress period of the original Clark, D.W., 1984, The ground-water system and simulated model because of a lack of new data. effects of ground-water withdrawals in northern Utah Val- Variations in annual recharge were assumed to be pro- ley, Utah: U.S. Geological Survey Water-Resources Investi- portional to changes in total surface-water inflow to northern gations Report 85–4007, 56 p. Utah Valley. The initial rate of recharge obtained from the steady-state calibration was 190,000 acre-ft/yr. This was multi- Clark, D.W., and Appel, C.L., 1985, Ground-water resources plied by one-half of the percentage change from the average of northern Utah Valley, Utah: Utah Department of Natural surface-water inflow during a given time period and applied Resources Technical Publication No. 80, 115 p. to all of the model cells simulating recharge. Recharge to the Harbaugh, A.W., Banta, E.R., Hill, M.C., and McDonald, ground-water system may change substantially from one year M.G., 2000, MODFLOW-2000, the U.S. Geological Survey to the next, and such changes are a major cause for variations modular ground-water model—User guide to modulariza- of water levels in northern Utah Valley. Recharge specified tion concepts and the ground-water flow process: U.S. in the model during the additional stress periods varied from Geological Survey Open-File Report 00–92, 121 p. 255,000 acre-ft in 1986 to 137,000 acre-ft in 1992. The ability of the updated transient-state model to match Harbaugh, A.W., and McDonald, M.G., 1996, User’s docu- hydrologic conditions estimated for 1981–2002 was evalu- mentation for MODFLOW-96, an update to the U.S. Geo- ated by comparing water-level changes measured in wells logical Survey modular finite difference ground-water flow to those computed by the model. Water-level measurements model: U.S. Geological Survey Open-File Report 96–485, made in February, March, or April were available for 39 wells 56 p. in the modeled area during all or part of 1981–2003. Most of the water-level measurements were made during the month of March, a time of the year when there is less stress on the ground-water system from pumping, evapotranspiration, snow- melt runoff, or irrigation. This water-level measurement was 20 Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002

Hill, M.C., Banta, E.R., Harbaugh, A.W., and Anderman, E.R., 2000, MODFLOW-2000, the U.S. Geological Survey modular ground-water model—User guide to the observation, sensitivity, and parameter-estimation process and three post-processing programs: U.S. Geological Survey Open- File Report 00–184, 209 p. Konikow, L.F., and Bredehoeft, J.D., 1992, Ground-water models cannot be validated: Advances in Water Resources, v. 15, no. 1, p. 75–83. McDonald, M.G., and Harbaugh, A.W., 1988, A modular three-dimensional finite-difference ground-water flow model: U.S. Geological Survey Techniques of Water- Resources Investigations, book 6, ch. A1, 586 p. Pollock, D.W., 1994, User’s guide for MODPATH/MOD- PATH-PLOT, Version 3: A particle tracking post-processing package for MODFLOW, the U.S. Geological Survey finite-difference ground-water flow model: U.S. Geological Survey Open-File Report 94–464.

Thiros—Evaluation of the Ground-Water Flow Model for Northern Utah Valley, Utah, Updated to Conditions through 2002—SIR 2006–5064