Groundwater Flow Models of Northeastern Illinois
a case study for building MODFLOW models with GIS
PROJECT FRAMEWORK
Abstract The Chicago metropolitan area of northeastern Illinois is experiencing rapid population growth, particularly in the suburban counties surrounding the city of Chicago. The population of one of these suburban counties, Kane County, is predicted to increase by as much as 36 percent by the year 2020. At this growth rate, five townships within Kane County are projected to experience local water shortages by 2020, prompting Kane County to fund a joint study by the Illinois State Water Survey and Illinois State Geological Survey of the water resources of Kane County. Managers and policy makers require a study that organizes the data, evaluates the system, improves understanding, and establishes a framework for follow-up analyses. This effort is assisted by the use of GIS and database technologies, which provide efficient tools for building a knowledge base for this study and translating geohydrological data into model-ready formats. Examples are provided for several procedures to efficiently manage, interpolate, and reformat MODFLOW data using ArcGIS 8.X.
Figure 1. The model domain includes the active regional aquifer model (light gray) and local deep aquifer model (dark gray).
Figure 2. Conceptual geological model of aquifers in northeastern Illinois
Flux & Head Measurements
Climate Data
Watershed Boundaries
Withdrawal Data
Surface Water Properties
Existing Nondigital Maps
Existing Digital Maps
Mass Balance Models USGS Stream Coding System Pattern Recognition*
Well Cell & Transient Data
Stream to River Cell Conversion
Digitizing & Interpolation
Assembling & Smoothing
The management of water resources in northeastern Illinois is complicated by international and interstate agreements, hydraulically coupled aquifer systems, natural and anthropogenic contamination, surface water – groundwater interaction, and conjunctive use of multiple resources. The net effect of these complications is that withdrawals from Lake Michigan cannot be increased beyond current rates, and the withdrawals from the deep bedrock aquifers may already be at their maximum. Urban planners have concluded that the region’s growing population will need to investigate the shallow aquifers as a source of water to meet growing demands The initial efforts of the project have focused on assembling the data and information of previous studies in a unified format to provide the framework for analyses. A series of databases and GIS files are being created for storage, retrieval, and processing of these data (Figure 3). The end product of this initial step is a geologic and hydrologic data set extensive enough to permit modeling on an interstate basis, while detailed enough to support site-scale models. Rather than commit to a single model code or grid, this study has emphasized the use of GIS in cataloging, analyzing, and adapting the data, as well as procedures and scripts for translating geohydrological data into model-ready formats.
K/T/S Pattern & Rates
Recharge (R) Pattern & Rates
Boundary Conditions
Model Grid Structure
Groundwater Flow Model: MODFLOW 2000 Using GIS Platform (ArcGIS 8.x) Calibration Processing: UCODE No
Calibration Targets Head and Flux
Optimum Parameter Values (R/K/T/S)
Yes
Optimum GW Flow Model
Parameter Sensitivity
* Lin, Y-F., and M.P. Anderson, 2003. A Digital Procedure for Ground Water Recharge and Discharge Pattern Recognition and Rate Estimation, Ground Water 41(3), p.306-315
Figure 3. Flow chart of the regional modeling process: 1) gray boxes indicate the processing applications; 2) The dashed box contains processes and data in a GIS compatible format.
Yu-Feng Lin, Doug Walker, Scott Meyer
yflin@uiuc.edu, ddwalker@uiuc.edu, smeyer@uiuc.edu DATABASE INTEGRATION
Several procedures have been developed to efficiently manage, interpolate, and reformat hydrogeological data using ArcGIS 8.2 and 8.3. Figure 4 illustrates the process employed for integrating geologic information collected from various states in different formats. We then apply procedures and scripts for translating geohydrological data into digital and model-ready formats (e.g., Figure 5).
Silurian
Maquoketa St. Peter
Prairie du Chien St. Lawrence Ironton-Galesville
Eau Claire Mt. Simon Precambrian
Figure 4. An example of information unified under a GIS platform for re-interpolation based on MODFLOW grids. The bedrock elevation information includes GIS contours in different coordinate systems (IL and IN), contours manually digitized from paper maps (MI), USGS DEM data (SW WI) and interpolated point elevations from previous studies (SE WI and Lake Michigan).
Figure 5. SURFER interpolation of nine major bedrock structural surfaces as a preliminary step in compiling a geologic regional model.
GIS and other database tools have been used in organizing data and information of various types to create a knowledge base for this study and future studies. These data include an extensive database of groundwater withdrawals in Illinois, complete for the period from 1980 to present, presently being augmented with data for the pre-1980 period for the Kane County area (Figure 6). Low-flow data for 20 streams within the regional model domain have been assembled to assure that the calibrated regional model predicts reasonable rates of natural groundwater discharge to streams. Two hundred and two sets of shallow aquifer pumping test data in the Kane County area have been identified. High-quality pumping test data will be reinterpreted in later stages of the project to determine aquifer and aquitard hydraulic properties for use in local-scale groundwater flow modeling of the shallow aquifers. The above data will be in GIS and Microsoft Access compatible formats.
(a) (b) Figure 6. Groundwater withdrawal database integration using GIS application: (a) deep aquifer withdrawal locations, and (b) shallow aquifer withdrawal locations. The data sources were from various state agencies. The shallow aquifer accounting region was selected by the watershed areas contributing to Kane county and the adjacent watersheds to those areas.
Kane County Deep Aquifer Accounting Region Shallow Aquifer Accounting Region
Illinois State Water Survey
2204 Griffith Drive, Champaign, IL 61820, USA
GROUNDWATER MODEL
The groundwater flow models for this study will use a nested modeling strategy, with a regional-scale model of the entire set of aquifers; a local-scale model of only the deep bedrock aquifer system (Figure 3); and localscale models of the shallow aquifer system in selected areas of interest. The regional model provides the boundary conditions for constructing local-scale models in three dimensions, including local-scale models of shallow aquifers in Kane County. The regional model will be a multilayer groundwater flow model of the entire aquifer system from the Precambrian crystalline basement to ground surface. A local-scale, multilayer groundwater flow model will be constructed of the deep bedrock aquifer system. Local-scale shallow aquifer models will be constructed on an as-needed basis to support water resources planning at the county level. The models generally will consist of the unconsolidated glacial drift and the upper 50 to 100 feet of the fractured dolomite and shale bedrock; model domains will extend to natural hydrologic divides around Kane County.
(a)
Model Approximate Dimension • Regional deep aquifer model: • approximately 360 miles by 360 miles • grid spacing varies from 16 miles to 1/2 mile. • model size is approximately 44,000 cells for each layer or 880,000 cells for all 20 layers. • Local deep aquifer model • approximately 60 miles by 90 miles (6 counties) • grid spacing varying from 1/2 to 1/4 mile. • Local shallow aquifer models: • grid density of approximately 1/2 to 1/16 mile.
(b)
N N
Z Y X
X:Y:Z = 1:1:100
Figure 7. Grid structure of regional groundwater flow model in (a) whole domain, and (b) near field with finer grid resolution.
Key elements of database for surface water groundwater (SW-GW) interaction • Length, perimeter from USGS National Hydrography Dataset (NHD) • Stream width inferred from drainage area using GIS scripts. • Streambed conductivity from published estimates, and from well tests near streams • Determination of river type or drain type using stream seasonality from Q7,10 • Discharge from baseflow analysis of streams (flux targets for calibration)
Figure 8. Surface water features from NHD in shallow aquifer accounting region (Figure 6).
• Database of predevelopment head
Conclusion This project will provide a framework for managing groundwater resources in Kane County, Illinois through collection, management, and analysis of data; conceptual and numerical modeling of aquifers; and evaluation of aquifer yield. Evaluation of the consequences of management options will also be examined based on this framework. The studies will produce a series of models and databases that will be maintained for use in assessing management options and will provide a framework for model updates and additional studies. ArcGIS 8.X and MODFLOW were chosen as the data management platform and groundwater flow modeling package, respectively, to optimize the research process in different stages and to amalgamate data from various sources. Acknowledgment U.S. Geological Survey - Wisconsin District Office, Wisconsin Geological and Natural History Survey, Indiana Geological Survey, Indiana Department of Natural Resources, Michigan Department of Environmental Quality, Illinois State Geological Survey, and Kane County Water Resources Department.