Link to USGS home page
Thumbnail of cover

A Precipitation-Runoff Model for the Analysis of the Effects
of Water Withdrawals and Land-Use Change on Streamflow
in the Usquepaug–Queen River Basin, Rhode Island

By Phillip J. Zarriello and Gardner C. Bent

Prepared in cooperation with the Rhode Island Water Resources Board

Scientific Investigations Report 2004-5139

ABSTRACT

The 36.1-square-mile Usquepaug–Queen River Basin in south-central Rhode Island is an important water resource. Streamflow records indicate that withdrawals may have diminished flows enough to affect aquatic habitat. Concern over the effect of withdrawals on streamflow and aquatic habitat prompted the development of a Hydrologic Simulation Program–FORTRAN (HSPF) model to evaluate the water-management alternatives and land-use change in the basin.

Climate, streamflow, and water-use data were collected to support the model development. A logistic-regression equation was developed for long-term simulations to predict the likelihood of irrigation, the primary water use in the basin, from antecedent potential evapotranspiration and precipitation for generating irrigation demands. The HSPF model represented the basin by 13 pervious-area and 2 impervious-area land-use segments and 20 stream reaches. The model was calibrated to the period January 1, 2000 to September 30, 2001, at three continuous streamflow-gaging stations that monitor flow from 10, 54, and 100 percent of the basin drainage area. Hydrographs and flow-duration curves of observed and simulated discharges, along with statistics compiled for various model-fit metrics, indicate a satisfactory model performance.

The calibrated HSPF model was modified to evaluate streamflow (1) under no withdrawals to streamflow under current (2000–01) withdrawal conditions under long-term (1960–2001) climatic conditions, (2) under withdrawals by the former Ladd School water-supply wells, and (3) under fully developed land use. The effects of converting from direct-stream withdrawals to ground-water withdrawals were evaluated outside of the HSPF model by use of the STRMDEPL program, which calculates the time delayed response of ground-water withdrawals on streamflow depletion.

Simulated effects of current withdrawals relative to no withdrawals indicate about a 20-percent decrease in the lowest mean daily streamflows at the basin outlet, but withdrawals have little effect on flows that are exceeded less than about 90 percent of the time. Tests of alternative model structures to evaluate model uncertainty indicate that the lowest mean daily flows ranged between 3 and 5 cubic feet per second (ft3/s) without withdrawals and 2.2 to 4 ft3/s with withdrawals. Changes in the minimum daily streamflows are more pronounced, however; at the upstream streamflow-gaging station, a minimum daily flow of 0.2 ft3/s was sustained without withdrawals, but simulations with withdrawals indicate that the reach would stop flowing part of a day about 5 percent of the time.

The effect on streamflow of potential ground-water withdrawals of 0.20, 0.90, and 1.78 million gallons per day (Mgal/d) at the former Ladd School near the central part of the basin were evaluated. The lowest daily mean flows in model reach 3, the main stem of the Queen River closest to the pumped wells, decreased by about 50 percent for withdrawals of 0.20 Mgal/d (from about 0.4 to 0.2 ft3/s) in comparison to current withdrawals. Reach 3 would occasionally stop flowing during part of the day at the 0.20-Mgal/d withdrawal rate because of diurnal fluctuation in streamflow. The higher withdrawal rates (0.90 and 1.78 Mgal/d) would cause reach 3 to stop flowing about 10 to 20 percent of the time, but the effects of pumping rapidly diminished downstream because of tributary inflows. Simulation results indicate little change in the annual 1-, 7-, and 30-day low flows at the 0.20 Mgal/d pumping rate, but at the 1.78 Mgal/d pumping rate, reach 3 stopped flowing for nearly a 7-day period every year and for a 30-day period about every other year. At the 0.90 Mgal/d pumping rate, reach 3 stopped flowing about every other year for a 7-day period and about once every 5 years for a 30-day period.

Land-use change was simulated by converting model hydrologic-response units (HRUs) representing undeveloped areas to HRUs representing developed areas on the basis of development suitability and town zoning. About 55 percent of the basin is suitable for development; this area would accommodate about 4,300 new low-density residential homes under current zoning. Increases in storm volume and peak flows, and decreases in base flow, typically associated with urbanization, were not evident in buildout simulations because the effective impervious area was assumed to increase by only 2 percent. Under fully developed conditions, withdrawals from self-supply wells were estimated to reach 1.2 Mgal/d. Potential increases in water withdrawals for a fully developed basin have only a minor impact on the main stem streamflow, but the effects of urbanization could be more pronounced in localized areas where development is concentrated.

Streamflow-depletion rates were calculated for varying distances of a pumped irrigation well from a stream. For the irrigation rates and aquifer conditions tested, streamflow depletion, relative to the pumping rate, decreases rapidly as the pumped well was moved away from the stream. Streamflow depletion, relative to the peak withdrawal rate, decreased by about 60, 80, and 90 percent by locating the pumped well 500, 1,000, and 1,500 feet from the stream, respectively.

CONTENTS

Abstract

Introduction

Purpose and Scope

Previous Investigations

Description of the Basin

Time-Series Data

Climate

Water Withdrawals

Data Collection

Logistic-Regression Equation to Predict Irrigation Withdrawals

Distribution of Daily Irrigation Withdrawals

Streamflow

Continuous-Record Stations

Partial-Record Stations

Precipitation-Runoff Model

Functional Description of Hydrologic Simulation Program–FORTRAN

Database

Representation of the Basin

Development of Hydrologic Response Units (HRUs)

Impervious Areas (IMPLNDs)

Pervious Areas (PERLNDs)

Stream Reaches (RCHRES)

Model Calibration

Tributary Streams

Ground-Water Underflow

Simulated Hydrologic Budgets and Flow Components

Sensitivity Analysis

Model Uncertainty and Limitations

Hydrologic Effects of Water Withdrawals and Land-Use Change

Current Withdrawals and No Withdrawals

Converting from Surface- to Ground-Water Withdrawals

Potential Withdrawals at the Former Ladd School Water-Supply Wells

Land-Use Change

Summary

Acknowledgments

References Cited

Appendix 1: Hydrologic Simulation Program–FORTRAN User Control File (uci) Input for Pervious-Area Model Variables

Appendix 2: Model-Fit Statistics Computed with the PEST Surface-Water Utilities Program

FIGURES

1–6. Maps of the Usquepaug–Queen River Basin showing:

1.Location, climate stations, and index streamflow-gaging stations used to develop continuous records at partial record stations, Rhode Island

2. Continuous- and partial-record streamflow-gaging stations

3. Generalized surficial geology

4. Generalized wetlands

5. Generalized land use and land cover

6. Principal withdrawal locations

7. Graph showing observed average daily irrigation patterns for A, individual turf- farms withdrawals; and B, combined withdrawals

8. Schematic showing simplified representation of the Hydrologic Simulation Program–FORTRAN (HSPF) inflows and outflows to a stream

9. Graph showing area as a percentage of the intervening basin area between streamflow-gaging stations Queen River at Exeter (QRPB), Queen River at Liberty (QRLY), and Usquepaug River near Usquepaug (USQU), for hydrologic response units (HRUs) in the Hydrologic Simulation Program–FORTRAN (HSPF) model

10. Map showing model reaches and subbasin boundaries developed for the Hydrologic Simulation Program–FORTRAN (HSPF) model

11–28. Graphs showing:

11. Precipitation at A, FBWR, and daily mean discharge simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) and observed discharge at streamflow-gaging stations B, Queen River at Exeter (QRPB); C, Queen River at Liberty (QRLY); and D, Usquepaug River near Usquepaug (USQU), Usquepaug–Queen River Basin, January 2000 through September 2001

12. Relation of Hydrologic Simulation Program–FORTRAN (HSPF) simulated discharge to observed discharge at streamflow-gaging stations A, Queen River at Exeter (QRPB); B, Queen River at Liberty (QRLY); and C, Usquepaug River near Usquepaug (USQU), January 2000 through September 2001

13. Flow-duration curves of daily mean discharges simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) and observed discharges at streamflow- gaging stations A, Queen River at Exeter (QRPB); B, Queen River at Liberty (QRLY); and C, Usquepaug River near Usquepaug (USQU), January 2000 through September 2001

14. Simulated and estimated daily mean discharge at Queens Fort Brook (model reach 8), January 2000 through September 2001

15. Calibration-period water budget by component from each hydrologic response unit simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) model, in inches A, per acre; and B, over the entire basin, January 2000 through September 2001

16. Wet-month water budget by component from each hydrologic response unit simulated by the Hydrologic Simulation Program –FORTRAN (HSPF) model, in inches A, per acre; and B, over the entire basin, March 2001

17. Dry-month water budget by component from each hydrologic response unit simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) model, in inches A, per acre; and B, over the entire basin, October 2000

18. Summary of water budgets by component averaged over all hydrologic response units simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) model for A, the calibration period—January 1, 2000, through September 30, 2001; B, wet month—March 2001; and C, dry month—October 2000

19. Sensitivity of selected Hydrologic Simulation Program–FORTRAN (HSPF) pervious area (PERLNDs) hydrologic variables to four indices of model fit for A, hourly discharge; and B, total monthly runoff

20. Daily mean discharge A, hydrographs; and B, flow-duration curves at the outlet (USQU) for observed flows and flows simulated with the calibrated Hydrologic Simulation Program–FORTRAN (HSPF) and an alternative HSPF model calibrated to low flows, January 2000 through September 2001

21. Irrigation withdrawals measured and simulated by the logistic-regression equation and the mean daily distribution pattern for A, June and July 2000; and B, 2000–01 irrigation season, reach 20

22. Flow-duration curves of daily mean discharge at the outlet (USQU) simulated with withdrawals and without withdrawals by the calibrated Hydrologic Simulation Program–FORTRAN (HSPF) and alternative models, 1960–2001

23. Hourly discharge at the outlet (USQU) observed and simulated by the A, calibrated Hydrologic Simulation Program–FORTRAN (HSPF); and B, an alternative HSPF model calibrated to low flow, August and September 1995

24. Flow-duration curves of simulated A, daily mean discharge; and B, minimum daily discharge at streamflow-gaging stations—Queen River at Exeter (QRPB), Queen River at Liberty (QRLY), and Usquepaug River near Usquepaug (USQU)— made with the Hydrologic Simulation Program–FORTRAN (HSPF) with and without withdrawals, 1960–2001

25. Magnitudes and recurrence intervals of A, 1-day; B, 7-day; and C, 30-day minimum daily mean streamflows determined by Log-Pearson Type III analysis of Hydrologic Simulation Program–FORTRAN (HSPF) simulated flows (1960–2001) and observed flows (1976–2002) at the outlet (USQU)

26. Observed mean daily irrigation withdrawals (2000–01) and corresponding streamflow depletions computed with the STRMDEPL program for a well 2,500 feet from the river

27. Observed mean daily irrigation withdrawals (2000) and corresponding streamflow depletions computed with the STRMDEPL program for hypo- thetical wells 50 to 2,500 feet from the river

28. Decrease in peak-streamflow depletion as a percentage of the peak withdrawal rate computed with the STRMDEPL program for an irregularly pumped well offset by distances of 50 to 2,500 feet from the river

29. Map showing approximate boundary of the former Ladd School facility, locations of water-supply wells, and model-reach numbers

30–33. Graphs showing:

30.Daily mean and hourly flow-duration curves of streamflow simulated with the Hydrologic Simulation Program–FORTRAN (HSPF) under no withdrawals (QUUS-NoW), current withdrawals (QUUS-lgW), and withdrawals of 0.20, 0.90, and 1.78 million gallons per day (QUUS-P1, QUUS-P2, and QUUS-P3, respectively) at the former Ladd School water-supply wells (fig. 29) for model A, reach 3; B, reach 9; and C, reach 20, 1960–2001

31. Magnitudes and recurrence intervals of A, 1-day; B, 7-day; and C, 30-day low flows at model reaches 3 and 9 by Log Pearson Type III analysis of discharge simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) under current withdrawals (QUUS-lgW), and simulated withdrawals at the former Ladd School water-supply wells of 0.20, 0.90, and 1.78 million gallons per day, (QUUS-P1, QUUS-P2, and QUUS-P3, respectively), 1960–2001

32. Flow-duration curves of simulated daily mean discharges under current (2000–01) withdrawals and buildout with various withdrawals at streamflow- gaging stations A, Queen River at Exeter (QRPB); B, Queen River at Liberty (QRLY); and C, Usquepaug River near Usquepaug (USQU), 1960–2001

33. Flow-duration curves of simulated daily mean discharges under current (2000–01) withdrawals and buildout with various withdrawals at Sherman Brook (reach 17), 1960–2001

TABLES

1.Climate data collected and compiled for the Usquepaug–Queen River Basin, Rhode Island

2.Continuous streamflow-gaging stations and partial-record stations in the Usquepaug–Queen River Basin

3.Summary of the relation between streamflow measurements at partial-record stations and mean daily discharge at nearby continuous streamflow-gaging stations (index stations) for computing a continuous record, Usquepaug–Queen River Basin

4.Organization and description of Data Set Numbers (DSNs) in the Watershed Data Management (WDM) database developed for the Usquepaug–Queen River Basin

5.Watershed Data Management (WDM) system constituent attribute (IDCONS) values for the Usquepaug–Queen River Basin

6.The percentage of developed area initially estimated as effective impervious area by land-use classification in the Usquepaug–Queen River Basin

7.Stream reaches (RCHRES) in the Hydrologic Simulation Program–FORTRAN (HSPF) model of the Usquepaug–Queen River Basin

8.Model-fit statistics computed by the Parameter Estimation (PEST) Surface Water Utilities program for flows simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) and observed flows at three continuous streamflow-gaging stations in the Usquepaug–Queen River Basin, January 2000 through September 2001

9.Model-fit statistics computed by the HSPEXP program for flows simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) and observed flows at three continuous streamflow-gaging stations in the Usquepaug–Queen River Basin, January 2000 through September 2001

10.Summary of daily and monthly model-fit statistics for flows simulated by the Hydrologic Simulation Program–FORTRAN (HSPF) and observed flows at three continuous streamflow-gaging stations in the Usquepaug–Queen River Basin, January 2000 through September 2001

11.Summary of model-run files (uci) and target data-set numbers (DSN) for Hydrologic Simulation Program–FORTRAN (HSPF) simulations of the Usquepaug– Queen River Basin

12.Magnitudes and recurrence intervals of 1-day, 7-day, and 30-day minimum mean daily discharges at streamflow-gaging stations—Queen River at Exeter (QRPB), Queen River at Liberty (QRLY), and Usquepaug River near Usquepaug (USQU)—by Log-Pearson type III analysis of Hydrologic Simulation Program–FORTRAN (HSPF) simulated flows with no withdrawals, Usquepaug–Queen River Basin, 1960–2001

13.Town zoning (2003), estimated developable area, and potential number of new homes within the Usquepaug–Queen River Basin

14.Summary of self-supply withdrawals for the period 1995–99 and the potential withdrawals by model reach (RCHRES) in the Usquepaug–Queen River Basin



AVAILABILITY

This report is presented in Portable Document Format (PDF).

To view and print report you will need to use PDF Viewer or Adobe Acrobat Reader (available as freeware)

Users with visual disabilities can visit this site for conversion tools and information to help make PDF files more accessible


Printable tabloid cover (20.8 MB)--1 page

Left inside cover (14 KB)--1 page

Right inside cover (22.3 KB)--1 page

Whole report (8.1 MB)--86 pages

 

The citation for this report, in USGS format, is as follows:

Zarriello, P.J., and Bent, G.C., 2004, A precipitation-runoff model for the analysis of the effects of water withdrawals and land-use change on streamflow in the Usquepaug–Queen River Basin, Rhode Island: U.S. Geological Survey Scientific Investigations Report 2004-5139, 75 p.


 For more information about USGS activities in Massachusetts-Rhode Island District, visit the USGS Massachusetts-Rhode Island Home Page.




FirstGov button  Take Pride in America button