Water Resources Management And Modeling Purna
Nayak download
https://ebookbell.com/product/water-resources-management-and-
modeling-purna-nayak-4109700
Explore and download more ebooks at ebookbell.com
Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Geospatial Information Handbook For Water Resources And Watershed
Management Volume 2 Methods And Modelling John G Lyon
https://ebookbell.com/product/geospatial-information-handbook-for-
water-resources-and-watershed-management-volume-2-methods-and-
modelling-john-g-lyon-46538182
Water Resources Management And Sustainability Solutions For Arid
Regions Mohsen Sherif
https://ebookbell.com/product/water-resources-management-and-
sustainability-solutions-for-arid-regions-mohsen-sherif-49435410
Public Participation In The Governance Of International Freshwater
Resources Water Resources Management And Policy Carl Bruch
https://ebookbell.com/product/public-participation-in-the-governance-
of-international-freshwater-resources-water-resources-management-and-
policy-carl-bruch-2181242
Water Resources Management Innovative And Green Solutions De Gruyter
Stem 2nd Edition Brears
https://ebookbell.com/product/water-resources-management-innovative-
and-green-solutions-de-gruyter-stem-2nd-edition-brears-56327240
Water Resources Management Innovative And Green Solutions Robert C
Brears
https://ebookbell.com/product/water-resources-management-innovative-
and-green-solutions-robert-c-brears-50336436
Enhancing Participation And Governance In Water Resources Management
Conventional Approaches And Information Technology Libor Jansky
https://ebookbell.com/product/enhancing-participation-and-governance-
in-water-resources-management-conventional-approaches-and-information-
technology-libor-jansky-1384502
Delta Waters Research To Support Integrated Water And Environmental
Management In The Lower Mississippi River 1st Edition National
Research Council Division On Earth And Life Studies Water Science And
Technology Board Committee On Strategic Research For Integrated Water
Resources Management
https://ebookbell.com/product/delta-waters-research-to-support-
integrated-water-and-environmental-management-in-the-lower-
mississippi-river-1st-edition-national-research-council-division-on-
earth-and-life-studies-water-science-and-technology-board-committee-
on-strategic-research-for-integrated-water-resources-
management-51874184
Water Resources Planning And Management R Quentin Grafton Karen Hussey
https://ebookbell.com/product/water-resources-planning-and-management-
r-quentin-grafton-karen-hussey-2324186
Water Resources Planning And Management Draft R Quentin Grafton Karen
Hussey Eds
https://ebookbell.com/product/water-resources-planning-and-management-
draft-r-quentin-grafton-karen-hussey-eds-4157192
WATER RESOURCES
MANAGEMENT
AND MODELING
Edited by Purna Nayak
Water Resources Management and Modeling
Edited by Purna Nayak
Published by InTech
Janeza Trdine 9, 51000 Rijeka, Croatia
Copyright © 2012 InTech
All chapters are Open Access distributed under the Creative Commons Attribution 3.0
license, which allows users to download, copy and build upon published articles even for
commercial purposes, as long as the author and publisher are properly credited, which
ensures maximum dissemination and a wider impact of our publications. After this work
has been published by InTech, authors have the right to republish it, in whole or part, in
any publication of which they are the author, and to make other personal use of the
work. Any republication, referencing or personal use of the work must explicitly identify
the original source.
As for readers, this license allows users to download, copy and build upon published
chapters even for commercial purposes, as long as the author and publisher are properly
credited, which ensures maximum dissemination and a wider impact of our publications.
Notice
Statements and opinions expressed in the chapters are these of the individual contributors
and not necessarily those of the editors or publisher. No responsibility is accepted for the
accuracy of information contained in the published chapters. The publisher assumes no
responsibility for any damage or injury to persons or property arising out of the use of any
materials, instructions, methods or ideas contained in the book.
Publishing Process Manager Marina Jozipovic
Technical Editor Teodora Smiljanic
Cover Designer InTech Design Team
First published March, 2012
Printed in Croatia
A free online edition of this book is available at www.intechopen.com
Additional hard copies can be obtained from orders@intechopen.com
Water Resources Management and Modeling, Edited by Purna Nayak
p. cm.
ISBN 978-953-51-0246-5
Contents
Preface IX
Part 1 Surface Water Modeling 1
Chapter 1 Tools for Watershed Planning – Development
of a Statewide Source Water Protection System (SWPS) 3
Michael P. Strager
Chapter 2 Strengths, Weaknesses, Opportunities
and Threats of Catchment Modelling with Soil
and Water Assessment Tool (SWAT) Model 39
Matjaž Glavan and Marina Pintar
Chapter 3 Modelling in the Semi Arid Volta Basin of West Africa 65
Raymond Abudu Kasei
Chapter 4 Consequences of Land Use
Changes on Hydrological Functioning 87
Luc Descroix and Okechukwu Amogu
Chapter 5 Fuzzy Nonlinear Function Approximation
(FNLLA) Model for River Flow Forecasting 109
P.C. Nayak, K.P. Sudheer and S.K. Jain
Chapter 6 San Quintin Lagoon Hydrodynamics Case Study 127
Oscar Delgado-González, Fernando Marván-Gargollo,
Adán Mejía-Trejo and Eduardo Gil-Silva
Chapter 7 Unsteady 1D Flow Model of Natural Rivers
with Vegetated Floodplain – An Application
to Analysis of Influence of Land Use on Flood
Wave Propagation in the Lower Biebrza Basin 145
Dorota Miroslaw-Swiatek
VI Contents
Chapter 8 Hydrology and Methylmercury
Availability in Coastal Plain Streams 169
Paul Bradley and Celeste Journey
Chapter 9 Contribution of GRACE Satellite
Gravimetry in Global and Regional
Hydrology, and in Ice Sheets Mass Balance 191
Frappart Frédéric and Ramillien Guillaume
Part 2 Groundwater Modeling 215
Chapter 10 Simplified Conceptual Structures and Analytical Solutions
for Groundwater Discharge Using Reservoir Equations 217
Alon Rimmer and Andreas Hartmann
Chapter 11 Integration of Groundwater Flow Modeling and GIS 239
Arshad Ashraf and Zulfiqar Ahmad
Chapter 12 Percolation Approach
in Underground Reservoir Modeling 263
Mohsen Masihi and Peter R. King
Chapter 13 Quantity and Quality Modeling of Groundwater
by Conjugation of ANN and Co-Kriging Approaches 287
Vahid Nourani and Reza Goli Ejlali
Preface
Water Resources Management intends to optimize the available water resources,
which consists of the optimal utilization of surface water and groundwater, to satisfy
the requirements of domestic, agricultural and industrial needs. In some parts of the
world, there is abundance of water, while in other parts of the world the resources are
scanty particularly in developing and under-developed countries. Floods and
droughts continue to threaten and affect the livelihoods of most of the population in
these countries. Therefore, there is an urgent need for optimal utilization of water
resources. With ever increasing population, particularly in developing and under-
developed countries, there is an urgent need to cater the needs of the population,
where water is the basic requirement. Groundwater and surface water play a pivotal
role in agriculture, and an increasing portion of extracted groundwater is used for
irrigating agriculture fields. It is estimated that at least 40% of the world's food grains
are produced using groundwater, by irrigated farming, both in countries with low
GDP as well as in high GDP countries. In arid and semi-arid areas, the dependency on
groundwater for water supply is much higher in comparison to other areas.
This book is designed to address some of the real issues concerning water resource
management, with some illustrative and good case studies relevant to the topic and
up-to-date. We hope that the chapters in the book will be of great use to postgraduate
and research scholars, providing them with current research trends and applications
of water resources for better management. This book consists of two sections: surface
water and groundwater. Surface water section covers watershed planning, impact of
climate change on Volta Basin, rainfall-runoff and sediment modeling using SWAT
model, flood forecasting using fuzzy logic approach, effect of land use changes on
hydrology, hydrodynamics and unsteady flow modeling, water quality modeling,
information on GRACE satellite and on wetland hydrology. Analytical solutions to
groundwater discharge are discussed in groundwater section followed by
groundwater flow modeling using MODFLOW, percolation approach, quality and
quantity modeling using ANN and Kriging approach. Different modeling approaches
are described followed by examples of case studies. The materials presented in this
book should help a wide range of readers to apply different simulation techniques to
resolve real life problems and issues concerned with water resource management.
X Preface
I am highly grateful to Intech, Open Access Publisher, for giving me the opportunity
to contribute as the Book Editor to this valuable book. In Particular, I would like to
thank Ms. Marina Jozipovic, the Publishing Process Manager, for her constant support
and cooperation during the preparation of this book. I also acknowledge the support
of my colleague Mr. D. Mohana Rangan for his assistance in reviewing the chapters
and helping in improving the quality of the contents.
Purna Nayak
National Institute of Hydrology,
Kakinada, AP,
India
Part 1
Surface Water Modeling
1
Tools for Watershed Planning –
Development of a Statewide Source
Water Protection System (SWPS)
Michael P. Strager
Division of Resource Management, West Virginia University,
USA
1. Introduction
A Surface Water Protection System (SWPS) was developed to bring spatial data and surface
water modeling to the desktop of West Virginia Bureau of Public Health (WVBPH), Office of
Environmental Health Services (OEHS), Environmental Engineering Division (EED). The
SWPS integrates spatial data and associated information with the overall goal of helping to
protect public drinking water supply systems.
The SWPS is a specialized GIS project interface, incorporating relevant data layers with
customized Geographic Information Systems (GIS) functions. Data layers have been
assembled for the entire state of West Virginia. Capabilities of the system include map
display and query, zone of critical concern delineation, stream flow modeling, coordinate
conversion, water quality modeling, and susceptibility ranking. The system was designed to
help meet the goals of the Surface Water Assessment and Protection (SWAP) Program.
The goal of the SWAP program is to assess, preserve, and protect West Virginia’s source
waters that supply water for the state’s public drinking water supply systems. Additionally,
the program seeks to provide for long term availability of abundant, safe water in sufficient
quality for present and future citizens of West Virginia. The SWPS was designed to help
meet this goal by addressing the three major components of the SWAP program: delineating
the source water protection area for surface and groundwater intakes, cataloging all
potential contamination sources, and determining the public drinking water supply
system’s susceptibility to contamination.
This chapter outlines the functions and capabilities of the SWPS and discusses how it
addresses the needs of the SWAP program. The following sections discuss the application
components. The components consist of:
1. A customized interface for study area selection
2. Integration of the EPA WHAEM and MODFLOW models
3. Delineation of groundwater public supply systems
4. Watershed delineation and zone of critical concern delineation for surface water sites
5. Stream flow model from multivariate regression
6. The environmental database
Water Resources Management and Modeling
4
7. UTM latitude/longitude conversion utility
8. Statewide map/GIS data layers
9. Water quality modeling capability
10. Groundwater and surface water susceptibility model
Component 1. A customized interface for study area selection
Using customized programming we were able to create a GIS interface to allow users to
quickly find locations or define study areas for further analysis in the state. The locations
may be selected in three ways: by geographical extent (e.g. county, watershed, 1:24,000 quad
map, major river basin), by area name or code (e.g. abandoned mine land problem area
description number, stream or river name, WV Division of Natural Resource (WVDNR)
stream code, public water identification number or name), or by typing in the latitude and
longitude coordinates. Once the study area is defined, the system zooms automatically to
the extent of the selected feature and all available spatial data layers are then displayed. A
discussion of the spatial data layers included is discussed in Component 8 of this
document.
Component 2. Integrating EPA WHAEM and MODFLOW models
The SWPS application has the ability to read output from either EPA WHAEM or
MODFLOW models. It does this by importing dxf file formats directly into SWPS from a
pull down menu choice. Data can also be converted to shapefile format from SWPS to be
read directly into WHAEM and MODFLOW. The data being read into SWPS needs to be in
the UTM zone 17, NAD27 projection (with map units meters) for the new data to overlay on
the current data existing within SWPS. Consequently, any data exported from SWPS will
automatically be in the UTM zone 17 NAD27 coordinate system.
Component 3. Delineation of groundwater public supply systems
A fixed radius buffer zone was created around each groundwater supply site based on the
pumping rate. If the pumping rate was less than or equal to 2,500 gpd, a radius of 500 feet
was used. If the pumping rate was greater than 2,500 gpd but less than or equal to 5,000
gpd, a radius of 750 feet was used. If the pumping rate was greater than 5,000 gpd and less
than or equal to 10,000 gpd, a radius of 1,000 feet was used. If the pumping rate was greater
than 10,000 gpd and less than or equal to 25,000 gpd, then a radius of 1,500 feet was used.
There were two exceptions to this fixed radius buffer procedure. The first was for any
groundwater site less than or equal to 25,000 gpd that was in a Karst or mine area. These
locations regardless of their pumping rate less than 25,000 gpd were buffered 2,000 feet. The
second exception was for sites over 25,000 gpd. For these sites, hydro geologic and/or
analytical mapping delineations will be done by personnel at the Bureau of Public Health.
These were only identified in SWPS as being a well location and are left to more
sophisticated groundwater modeling software.
To perform buffers automatically, the user can use the GIS to create buffers dialog within
SWPS susceptibility ranking menu option. The automatic fixed radius buffering requires
knowledge about the pumping rate and fixed radius distance. This information is provided
in a pulldown text information box within the susceptibility ranking menu option.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 5
Component 4. Watershed delineation and zone of critical concern delineation for surface
water sites
The ability to interactively delineate watersheds and zones of critical concern is built into
SWPS. In this section, the watershed delineation tool is discussed, followed by the zone of
critical concern delineation tool.
Watershed Delineation
SWPS allows the user to delineate a watershed for any mapped stream location in the state.
The watershed is delineated based on the user-clicked point and it is added to the current
view’s table of contents as a new theme or map layer labeled “Subwatershed.” The drainage
area is reported back to the user as well. If only drainage area is requested, a separate tool
allows for quick query of stream drainage area in acres and square miles, without waiting
for the watershed boundary to be calculated.
The watershed delineation is driven by a hydrologically correct digital elevation model
(DEM). The DEM is corrected using stream centerlines for all 1:24,000 streams. The stream
centerlines are converted to raster cells and DEM values are calculated for each cell. All off-
stream DEM cells are raised by a value of 20 meters to assure the DEM stream locations are
the lowest cells in the DEM. This step is necessary to assure of more accurate watershed
delineations especially at the mouth of the watersheds. After the DEM is filled of all
spurious sinks, flow direction and flow accumulation grids are calculated. These grids help
determine the direction of flow and the accumulated area for each cell in the landscape.
These grids were necessary for watershed delineation to occur and are important inputs for
finding the zones of critical concern for surface water intakes.
Surface Water Zones of Critical Concern
Stream velocity is the driving factor for determining a five-hour upstream delineation for
each surface water intake in WV. Only with stream velocity calculated was it possible to
include factors such as high bank-full flow, average flow, stream slope, and drainage area all
at once. The velocity equation used in this study came from a report titled “Prediction of
Travel Time and Longitudinal Dispersion in Rivers and Streams” (US Geological Survey,
Water-Resources Investigations Report 96-4013, 1996). In this report, data were analyzed for
over 980 subreaches or about 90 different rivers in the United States representing a wide
range of river sizes, slopes, and geomorphic types. The authors found that four variables
were available in sufficient quantities for a regression analysis. The variables included the
drainage area (Da), the reach slope (S), the mean annual river discharge (Qa), and the
discharge at the section at time of the measurement (Q). The report defines peak velocity
as:
V’
p = VpDa/Q
The dimensionless drainage area as:
D’
a = Da
1.25 * sqrt(g) / Qa
Where g is the acceleration of gravity. The dimensionless relative discharge is defined as:
Q’a = Q/Qa
Water Resources Management and Modeling
6
The equations are homogeneous, so any consistent system of units can be used in the
dimensionless groups. The regression equation that follows has a constant term that has
specific units, meters per second. The most convenient set of units for use with the equation
are: velocity in meters per second, discharge in cubic meters per second, drainage area in
square meters, acceleration of gravity in m/s2, and slope in meters per meter.
The equation derived in the report and the equation used in this study for peak velocity in
meters per second was the following:
Vp = 0.094 + 0.0143 * (D’
a)0.919 * (Q’
a)-0.469 * S 0.159 * Q/Da
The standard error estimates of the constant and slope are 0.026 m/s and 0.0003,
respectively. This prediction equation had an R2 of 0.70 and a RMS error of 0.157 m/s.
Once a velocity grid was calculated as described above, it was used as an inverse weight
grid in the flowlength ArcGIS (ESRI, 2010) command. The flowlength command calculates a
stream length in meters. If velocity is in meters per second, the inverse velocity as a weight
grid will return seconds in our output grid. This calculation of seconds would track how
long water takes to move from every cell in the state where a stream is located to where it
leaves the state. The higher values will exist in the headwater sections of a watershed. By
querying the grid, it is possible to add the appropriate travel time to the cell value and this
will the time of travel for an intake. All cells above an intake by 18,000 seconds (5 hours) will
be the locations in which water would take to reach the intake.
To use this methodology, GIS data layers had to be calculated for drainage area, stream
slope, annual average flow, and bank-full flow for all of WV. The sections below describe
how each of these grids was created.
Drainage area
To obtain a drainage area calculation for every stream cell in the state required a
hydrogically correct DEM. The process of creating a hydrologically correct DEM was
covered in the watershed delineation component described earlier. Essentially, from the
DEM the flow direction and flow accumulation values for each stream cell are derived. The
output of the flow direction request is an integer grid whose values range from 1 to 255. The
values for each direction from the center are:
32 64 128
16 X 1
8 4 2
For example, if the direction of steepest drop were to the left of the current processing cell,
its flow direction would be coded as 16. If a cell is lower than its 8 neighbors, that cell is
given the value of its lowest neighbor and flow is defined towards this cell (ESRI, 2010).
The accumulated flow is based upon the number of cells flowing into each cell in the output
grid. The current processing cell is not considered in this accumulation. Output cells with a
high flow accumulation are areas of concentrated flow and may be used to identify stream
channels. Output cells with a flow accumulation of zero are local topographic highs and
may be used to identify ridges. The equation to calculate drainage area from a 20-meter cell
sized flow accumulation grid was:
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 7
(cell value of flow accumulation grid + 1) * 400 = drainage area in meters squared
Stream slope
Stream slope was calculated for each stream reach in the state. A stream reach is not necessarily
an entire stream but only the section of a stream between junctions. The GIS command
streamlink was first used to find all unique streams between stream intersections or junctions.
For each of these reaches, the length was calculated from the flowlength GIS command.
Having the original DEM allowed us to find the maximum and minimum values for each of
the stream reaches. The difference in the maximum and minimum elevations for the stream
reach divided by the total reach length gave us our stream reach slope in meters per meter.
Annual average flow
Annual average flow for each stream cell location was found based on a relationship
between drainage area and gauged stream flow. For 88 gauging stations in WV, covering
many different rainfall, geological, and elevation regions, we assembled a table of drainage
area for the gauges versus the historic annual stream flow for the gauge. After fitting a
linear regression line for this data set, we found the following equation for annual stream
flow setting the y intercept to zero.
Annual stream flow in cfs = 2.05 * drainage area in square miles
This equation had a corrected R2 of .9729. The XY plot and equation are shown in Figure 1.
Fig. 1. Annual stream flow from gauged stations and drainage area at the gauges
Since drainage area is already calculated for each stream cell location, this equation
incorporated the drainage area grid to compute a separate grid layer of annual stream flow.
This would be another input for the velocity calculation.
Bank-full flow
The last input for the velocity equation was the bank-full flow measure. Just as with annual
average flow, this required a modeled value for every raster stream cell in WV. Using the
drainage area vs. mean flow y = 2.05x
R2
= 0.9729
-500
0
500
1000
1500
2000
2500
3000
3500
0.00 200.00 400.00 600.00 800.00 1000.00 1200.00 1400.00 1600.00
drainage area sq miles
mean
flow
in
cfs
Water Resources Management and Modeling
8
same approach to regressing drainage area to gauged stream flow as performed to find an
annual average flow equation, this equation was used to find bank-full flow. Bank-full flow
as defined by the Bureau of Public Health, is 90% of the annual high flow. To find the 90% of
high flow for each gauging station, all historic daily stream flow data was downloaded for
each of the 88 gauging stations. This data was then sorted lowest to highest and then
numbered lowest to highest after removing repeating values. The value of flow at the 90% of
the data became the bank-full flow value for that gauge. These values were then regressed
against drainage area at the gauge. The linear regression equation for bank-full stream flow
setting the y intercept to zero is listed below.
Bank-full stream flow in cfs = 4.357 * drainage area in square miles
This equation had a corrected R2 of .9265. The XY plot and equation are shown in Figure 2.
Fig. 2. Bank-full stream flow from gauged stations and drainage area at the gauges
This equation could be applied to the drainage area grid to calculate the bank-full flow for
any stream cell in the state. It was the final input needed in the velocity calculation.
The interactive zone of critical concern ability of SWPS delineates the upstream contributing
area for a surface water intake in the following way. First, the user locates the surface water
intake and makes sure the intake is on the raster stream cell. A button on the interface then
initiates the model. The model will query the time of travel value for the intake and then add
18,000 seconds (5 hours) to the queried value upper range. All cells which fit this range are
identified and the stream order attribute retrieved for those cells. All cells that are on the main
stem stream where the intake existed are buffered 1000 feet on each side of the stream. All
tributaries to the main stem are buffered 500 feet on each side of the stream. Next, a watershed
boundary for the location of the intake is delineated and used to clip any areas of the buffer
that may extend beyond ridgelines. And lastly, the surface water intake is buffered 1000 feet
and combined with the clipped buffer to include areas 1000 feet downstream of the intake.
drainage area vs. 90% of high flow
y = 4.357x
R
2
= 0.9265
-1000
0
1000
2000
3000
4000
5000
6000
7000
0 500 1000 1500 2000
drainage area in sq miles
90%
of
high
flow
in
cfs
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 9
This interactive ability allows zones of critical concern to be delineated for any river or stream
in WV. Only large rivers which border WV, such as the Ohio, Tug, and Potomac can not be
interactively delineated using this method. This is due to unknown drainage areas for these
bordering rivers and unknown tributaries to these major rivers coming from the bordering
states. This is the major limitation of this modeling approach for WV at this time and in the
next version of this watershed tool will account for all outer drainage influences.
The Ohio River Sanitation Commission (ORSANCO) is responsible for delineating zones of
critical concern for the intakes along the Ohio River. ORSANCO uses uniform 25-mile
upstream distances for zones of critical concern for intakes along the Ohio River. This same
approach could be applied to other rivers such as the Tug and Potomac in WV.
For reservoirs and lakes within the watershed delineation area, a set of standards was set by
the Bureau of Public Health and was used in this study. For a reservoir, a buffer of 1000 feet
on each bank and 500 feet on each bank of the tributaries that drain into the lake or reservoir
was used. When a lake or reservoir is encountered within the five-hour time of travel, a
specific delineation was used. If the length of the lake/reservoir was less than or equal to the
five hour calculated time of travel distance from the intake, then the entire water body was
included. If the length of the lake/reservoir was greater than the calculated five hour time of
travel distance from the intake, then the section of water body within the five hour time of
travel distance was used to establish the zone of critical concern.
Component 5. Stream flow model from multivariate regression
Overview
This project component for SWPS used multivariate techniques to evaluate stream flow
estimation variables in West Virginia. The techniques included correlation analysis, multiple
regression, cluster analysis, discriminant analysis and factor analysis. The major goal was to
define watershed scale factors to estimate the stream flow at recorded USGS gauges. To do
this, the contributing area upstream of each gauge was first delineated. Next, annual
averages of precipitation and temperature and landscape based variables for the
contributing upstream area were calculated and regressed against 30-year average annual
flow at the USGS gauge. Results from the statistical analysis techniques found the most
important variables to be upstream drainage area, 30-year annual maximum temperature,
and stream slope. While this analysis was limited by the availability of data and
assumptions to predict stream flow, the results indicate that stream flow can be modeled
with reasonably good results.
The following sections include a review of the literature on stream flow estimation
techniques, a description of the variables used in this study to predict stream flow, the
multivariate statistical methods, and a discussion of results and limitations of the study.
Literature Review
The intent of this literature review was to determine variables that were used to estimate
stream flow in other studies, identify different statistical procedures, and to find limitations
in this study based on other papers.
The impact of land-use, climate change and groundwater abstraction on stream flow was
examined by Qerner et al. (1997). They analyzed the effects of these factors using physical
Water Resources Management and Modeling
10
models BILAN, HBVOR, MODFLOW and MODGROW. The models were used to simulate
the impact of afforstation, climate warming by 2 and 4 degrees Celsius in combination with
an adoption of the precipitation changes in groundwater recharge and groundwater
abstractions on stream flow droughts. The authors found that all the physical models can be
used to assess the impacts of human activities on stream flow. They also concluded that
based on some climate change scenarios they followed out, that the deficit volume of water
is very sensitive to both an increase in temperature and a change in precipitation. Even in
basins with abundant precipitation, the warming of 2 degrees Celsius would result in a rise
in the deficit volume of water by 20 percent. Their findings also acknowledge the
importance of using precipitation, temperature, groundwater recharge and groundwater
abstractions along with water storage holding capacity of watersheds.
Timofeyeva and Craig (1998) used Monte Carlo techniques to estimate month by month
variability of temperature and precipitation for drainage basins delineated by a digital
elevation model. They also used a runoff grid from the digital elevation model to estimate
discharge at selected points and compared this to known gauge station data. The variance of
temperature was modeled as the standard error of the regression from the canonical
regression equation. For precipitation, they modeled the variance as the standard error of
the prediction. This was done to achieve unbiased estimators. When comparing the climate
and resulting runoff and stream flow estimators calculated by Monte Carlo estimation, to
the observed flow, the simulated results were within the natural variability of the record
(Timofeyeva and Craig, 1998).
Long-range stream flow forecasting using nonparametric regression procedures was
developed by Smith, (1991). The forecasting procedures, which were based solely on daily
stream flow data, utilized nonparametric regression to relate a forecast variable to a
covariate variable. The techniques were adopted to develop long-term forecasts of minimum
daily flow of the Potomac River at Washington, D.C. Smith’s key finding was that to
implement nonparametric regression requires the successful specification of “bandwidth
parameters.” The bandwidth parameters are chosen to minimize the integrated mean square
error of forecasts. Basically, his stream flow technique focussed on examining past history of
stream flow and making nonparametric regression forecasts based on what is likely to occur
in the future. No additional variables besides historic flow were used to model future
conditions.
Another nonparameteric approach to stream flow simulation was done by Sharma et al.
(1997). They used kernal estimates of the joint and conditional probability density functions
to generate synthetic stream flow sequences. Kernal density estimation includes a weighted
moving average of the empirical frequency distribution of the data (Sharma, et al. 1997). The
reason for this method is to estimate a multivariate density function. This is a nonparametric
method for the synthesis of stream flow that is data driven and avoids prior assumptions as
to the form of dependence (linear or non linear) and the form of the probability density
function. The authors main finding was that the nonparametric method was more flexible
for their study than the conventional models used in stochastic hydrology and is capable of
reproducing both linear and nonlinear dependence. In addition, their results when applied
to a river basin indicated that the nonparametric approach was a feasible alternative to
parametric approaches used to model stream flow.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 11
Garren (1992) noted that although multiple regression has been used to predict seasonal
stream flow volumes, typical practice has not realized the maximum accuracy obtainable
from regression. The forecasting methods he mentions which can help provide superior
forecasting include: (1) Using only data known at forecast time; (2) principal components
regression; (3) cross validation; and (4) systematically searching for optimal or near-optimal
combinations of variables. Some of the variables he used included snow water equivalent,
monthly precipitation, and stream flow. The testing of selection sites for a stream flow
forecasting study, he feels should be based on data quality, correlation analyses, conceptual
appropriateness, professional judgement, and trial and error. The use of principal
components regression provides the most satisfactory and statistically rigorous way to deal
with intercorrelation of variables. He concluded that the maximum forecast accuracy gain is
obtained by proper selection of variables followed by the use of principal components
regression and using only known data (no future variables).
The results of a multiple-input transfer function modeling for daily stream flow using
nonlinear inputs was studied by Astatkie and Watt (1998). They argue that since the
relationship between stream flow and its major inputs, precipitation and temperature, are
nonlinear, the next best alternative is to use a multiple input transfer function model
identification procedure. The transfer function model they use includes variables such as
type of terrain, drainage area, watercourse, the rate of areal distribution of rainfall input,
catchment retention, loss through evapotranspiration and infiltration into the groundwater,
catchment storage, and melting snow. When comparing their modeling technique for stream
flow to that of a nonlinear time series model, they found their transfer function model to be
direct and relatively easy for modeling multiple inputs. They also found it more accurate in
head to head tests against the nonlinear time series model.
Since stream flow modeling is an outcome of many runoff estimation models, the literature
for deriving runoff grids is applicable to stream flow studies. Anderson and Lepisto (1998)
examined the links between runoff generation, climate, and nitrate leaching from forested
catchments. One of the things they sought out to prove in their study was that climate will
influence the amount of nitrate that can be leached from the soil and the water flow that will
transport it to the streams. They found that a negative correlation existed between stream
flow and temperature. Significant positive correlation between modeled surface runoff and
concentrations of nitrate was found when they considered periods of flow increases during
cold periods. Their study identified the importance for identifying and calculating the
surface runoff fraction, daily dynamics of soil moisture, groundwater levels, and extensions
of saturated areas when doing a contaminant transport or flow estimation study.
In another study, Moore (1997) sought to provide an alternative to the matching strip,
correlation, and parameter-averaging methods for deriving master recession characteristics
from a set of recession segments. The author then choose to apply the method to stream
flow recession segments for a small forested catchment in which baseflow is provided by
drainage of the saturated zone in the shallow permeable soil. The plots indicated the
recessions were non-linear and that the recessions did not follow a common single valued
storage outflow relation. The final decision was a model with two linear reservoirs that
provided substantially better fit than three single reservoir models, indicating that the form
of the recession curve probably depends on not just the volume of subsurface storage, but
also on its initial distribution among reservoirs.
Water Resources Management and Modeling
12
Gabriele et al. (1997) developed a watershed specific model to quantify stream flow,
suspended sediment, and metal transport. The model, which estimated stream flow,
included the sum of three major components: quick storm flow, slow storm flow, and long-
term base flow. Channel components were included to account for timing effects associated
with waters, sediments, and metals coming from different areas. Because of relatively good
results from the modeling process, the conceptualizations supported that the study area
river was strongly influenced by three major components of flow: quick storm flow, slow
storm flow, and long-term base flow. Therefore, sediment inputs can be associated with
each of those stream flow components and assign metal pollution concentrations to each
flow and sediment input.
From this review of other studies, variables were determined that have been used
successfully in stream flow estimation. Examining the limitations of other studies has also
provided insight into data layers that may not be able to include. Of the statistical
techniques used, the multivariate approach, in which components are added or subtracted
to achieve the best fit possible, is a sound statistical procedure. In addition to this approach,
testing the correlations between variables is another way of finding a model for estimating
stream flow in WV.
Methodology
The first step in assembling data for this study was to delineate the total upstream
contributing area for each of thirteen USGS gauge stations in West Virginia. Figure 3
displays the location of each gauge and the defined upstream drainage area for that gauge.
Fig. 3.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 13
For every drainage area, the following criteria were calculated; total area, 30 year average
annual precipitation, 30 year average annual maximum temperature, 30 year average annual
minimum temperature, average drainage area slope, and stream slope. These variables were
explanatory variables, which would be regressed against the dependent variable, the 30year
average annual flow recorded at the gauge stations. The figures 4 to 7 show the distribution
of 30-year precipitation, maximum temperature, minimum temperature, and elevation
across the different areas. By using GIS techniques, it was possible to find the average value
in the drainage areas along with drainage area slope and stream slope for each of the
variables. The data for each gauge area and assembled variables is summarized in table 1.
id# USGS
Gauge
name
Upstream
drainage
area
30yr
annual
30yr
annual
30yr
annual
30yr
annual
stream
elevation
drop
Watershed
Slope
average
(acres) precip ave
(inches)
ave temp
max(F)
ave temp
min(F)
Stream
flow (cfs)
max-min
in (meters)
(degree)
g1 1595200 31296 52 54 35 99.68 418 5
g5 3050000 120352 50 57 37 379.37 607 15
g7 3053500 176708 46 60 38 613.56 643 11
g10 3061000 484507 43 62 39 1158.14 26 13
g11 3061500 74501 42 61 39 168.99 130 13
g12 3062400 7146 43 60 37 16.54 189 9
g13 3066000 55068 53 54 36 210.40 289 6
g17 3114500 289609 42 62 40 665.40 57 15
g19 3180500 85166 53 55 36 273.49 435 14
g21 3189100 338131 53 57 37 1445.61 743 13
g22 3190400 232990 50 59 38 750.36 830 11
g24 3195500 346231 48 59 38 1176.66 933 17
g26 3202400 196645 47 63 39 421.78 583 18
Table 1. Data used in study
The first step in analyzing the data in table 1 was to perform some basic statistics. The
values across the different gauging station locations were investigated. The summarized
statistical data is shown in table 2.
Variable N Mean Median TrMean StDev SE Mean Minimum Maximum Q1 Q3
area 13 187565 176708 176972 144629 40113 7146 484507 64784 313870
precip 13 47.85 48 47.91 4.34 1.2 42 53 43 52.5
maxtemp 13 58.692 59 58.727 3.066 0.85 54 63 56 61.5
mintemp 13 37.615 38 37.636 1.446 0.401 35 40 36.5 39
flow 13 568 422 538 455 126 17 1446 190 954
strmslop 13 452.5 435 447.6 299.3 83 26 933 159.5 693
wsslope 13 12.31 13 12.45 3.88 1.08 5 18 10 15
Table 2. Basic statistics
Water Resources Management and Modeling
14
Fig. 4.
Fig. 5.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 15
Fig. 6.
Fig. 7.
Water Resources Management and Modeling
16
From table 2, it was noted which variables were closely grouped and which varied
significantly among all the 13 different gauges. The area and flow variables have the highest
standard deviation while the precipitation, maximum and minimum temperatures, and
watershed slope have the lowest standard deviation. Other simple statistical graphs, which
were used to gain insights into the data distribution and spreads, are shown in figures 8 to
14. The figures provided a graphical display of the distribution of values across the 13
gauges. Data exploration is important to determine trends and outliers in data that may bias
results (Johnson, et al 2001). In addition, regression results may be impacted from large
variations in data values. A common technique is to normalize data with a simple equation
such as the value of interest minus the minimum value for that variable divided by the
maximum minus minimum within the data range (Kachigan, 1986). However, in this study
the values were not normalized due to the spatial nature of the information source. It was
necessary to identify and incorporate the spatial variability across the entire study area at
the statewide level. The end use of our regression relationship is the ability to query any
raster stream cell and report all the unique information from the spatial analysis. Stream
flow and water quality decisions for permitting may occur in high elevation cold headwater
segments as well as large river systems with much accumulated drainage. Because the study
area had unique topographic features that were to be regressed against representative
stream flow information, the gauge driven delineated watersheds were chosen to represent
this differentiation as best as possible as shown in Figure 3.
Fig. 8.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 17
Fig. 9.
Fig. 10.
Water Resources Management and Modeling
18
Fig. 11.
Fig. 12.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 19
Fig. 13.
Fig. 14.
Water Resources Management and Modeling
20
The next step in analyzing the data was to use generate a best-fit line plot for each of the
independent variables in table 1 regressed against the dependent variable stream flow.
These plots are shown in figures 15 to 20. From these best-fit line plots, the area, stream
slope, and watershed slope variables had the best R squared values and positive linear
relationship. The maximum and minimum temperature variables along with precipitation
had the worst linear fit with stream flow. Their R squared values were very low with the
precipitation variable looking very random in describing stream flow. At this point in the
analysis it appeared that the area, stream slope and watershed slope will be the better
variables to predict stream flow.
While the linear regression plots provided some idea of the extent of the relationship
between two variables, the correlation coefficient gives a summary measure that
communicates the extent of correlation between two variables in a single number (Kachigan,
1986). The higher the correlation coefficient, the more closely grouped are the data points
representing each objects score on the respective variables. Some important assumptions of
the correlation coefficient are that the data line in groupings that are linear in form. The
other important assumptions include that the variables are random and measured on either
an interval or a ratio scale. In addition, the last assumption for the use of the correlation
coefficient is that the two variables have a bivariate normal distribution. The correlation
matrix for the data used in this study is shown in table 3.
area precip maxtemp mintemp flow strslope wsslope
area 1 -0.212 0.470 0.571 0.922 0.138 0.516
precip -0.212 1 -0.850 -0.781 0.039 0.560 -0.279
maxtemp 0.470 -0.850 1 0.930 0.245 -0.226 0.590
mintemp 0.571 -0.781 0.930 1 0.356 -0.217 0.647
flow 0.922 0.039 0.245 0.356 1 0.392 0.435
strslope 0.138 0.560 -0.226 -0.217 0.392 1 0.245
wsslope 0.516 -0.279 0.590 0.647 0.435 0.245 1
Table 3. Correlation matrix
The variables with significant correlations (R > .7) are shaded in table 3. The variables listed
in order of highest correlation to lowest significance are mintemp and maxtemp, flow and
area, precip and maxtemp, and precip and mintemp. The correlations between the weather
data were expected. In areas of higher precipitation, the temperature will be cooler (the
annual averages for maximum temperature will be lower and the annual average for
minimum temperatures will be lower) hence the high negative correlation. The other high
positive correlated variables indicate that the variation in one variable will lead to variation
in the other variable. For regression analysis the variables should be independent.
Collinearity refers to linear relationships within the variables. The amount of
multicollinearity across variables can be examined with principal component analysis of a
sample correlation matrix (Sundberg, 2002) among other methods to remove dependence.
This study examined the smallest eigenvalue and eliminated variables with values less than
0.05 as an indication of substantial collinearity (Hocking, 1996). As expected the
precipitation variables were not independent to the elevation data and therefore removed.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 21
Fig. 15.
Fig. 16.
Water Resources Management and Modeling
22
Fig. 17.
Fig. 18.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 23
Fig. 19.
Fig. 20.
Water Resources Management and Modeling
24
Performing regression analysis on the data was the next step in formulating a relationship
and model to predict and estimate stream flow. Using the technique by Garren (1992) a
regression equation with all the remaining variables was created, evaluate the P values of
each variable, and eliminate variables until the highest adjusted R square is found. The first
run with the regression analysis indicates that the variables area, strmslop and maxtemp
will have the most influence on flow because of their low P values. Table 4 shows the
regression analysis including all the variables.
The regression equation is
flow = 2325 + 0.00310 area - 12.0 precip - 37.3 maxtemp + 8 mintemp
+ 0.423 strmslope - 4.8 wsslope
Predictor Coef StDev T P
Constant 2325 4370 0.53 0.614
area 0.0030987 0.0004281 7.24 0.000
precip -12.02 30.84 -0.39 0.710
maxtemp -37.31 53.50 -0.70 0.512
mintemp 7.8 100.7 0.08 0.941
strmslop 0.4235 0.2281 1.86 0.113
wsslope -4.83 18.54 -0.26 0.803
S = 156.3 R-Sq = 94.1% R-Sq(adj) = 88.2%
Table 4. Regression analysis including all variables
By systematically removing the variables with a high P value and noting the R squared
adjusted value, it was possible to arrive at a final set of variables to use in a regression
equation to estimate stream flow. Table 5 shows the regression analysis results after
removing the variable with the highest P value (mintemp).
The regression equation is
flow = 2492 + 0.00311 area - 12.3 precip - 35.0 maxtemp + 0.421
strmslope
- 4.3 wsslope
Predictor Coef StDev T P
Constant 2492 3522 0.71 0.502
area 0.0031121 0.0003628 8.58 0.000
precip -12.35 28.30 -0.44 0.676
maxtemp -35.00 41.23 -0.85 0.424
strmslop 0.4208 0.2089 2.01 0.084
wsslope -4.33 16.11 -0.27 0.796
S = 144.7 R-Sq = 94.1% R-Sq(adj) = 89.9%
Table 5. Regression analysis with mintemp removed
The R squared adjusted improved slightly to 89.9% with mintemp removed. This process
of removing the current highest P value variable and re-running of the model was
repeated six times. The associated R squared values were noted and table 6 was created
from the results.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 25
As table 6 indicates, the combination of variables that provided the highest R squared
adjusted value were area, maxtemp, and strslope. The associated regression equation with
the optimal set of variables is:
flow = 1232 + 0.00304 area - 23.6 maxtemp + 0.338 strmslope
Variables included in the regression R squared adjusted
Area, mintemp, maxtemp, strslope, wsslope 88.2
Area, maxtemp, strslope, wsslope 89.9
Area, maxtemp, strslope 91.1
Area, maxtemp, strslope 91.8
Area, strslope 90.5
Area 83.6
Table 6. Multiple regression results
The next procedure used in the analysis was discriminant analysis. This technique was used
to identify relationships between qualitative criterion variables and the quantitative
predictor variables in the dataset. The objective was to identify boundaries between the
groups of watersheds that the gauges were associated. The boundaries between the groups
are the characteristics that distinguish or discriminate the objects in the respective groups.
Discriminant analysis allows the user to classify the given objects into groups – or
equivalently, to assign them a qualitative label – based on information on various predictor
or classification variables (Kachigan, 1992).
The gauge station dataset was assigned a qualitative variable based on which major
drainage basin in West Virginia the area was located. The major basins used were the
Monongahela (m), Gauley (g) and Other (x). The class “other” was assigned to gauges that
did not fall in the Monongahela or Gauley drainage basins. Running the discriminant
analysis in Minitab produced the results shown in table 7.
Only gauge one and gauge five were reclassified from the discriminant analysis results. It
should be noted however that the discriminant function should be validated by testing its
efficacy with a fresh sample of analytical objects. Kachigan (1986) notes that the observed
accuracy of prediction on the sample upon which the function was developed will always be
spuriously high, because we will have capitalized on chance relationships. The true
discriminatory power of the function will be found when tested with a completely separate
sample.
By using discriminant analysis, it enabled the investigation of how the given groups differ.
In the next analysis step, cluster analysis, the goal is to find whether a given group can be
partitioned into subgroups that differ. The advantage of the approach is in providing a
better feel of how the clusters are formed and which particular objects are most similar to
one another.
The cluster analysis was performed with distance measures of Pearson and Average and
link methods of single and Euclidean. The Average and Euclidean choices worked the best
in identifying clusters. Figure 21 shows the dendrogram results and table 8 lists the
computation results.
Water Resources Management and Modeling
26
Linear Method for Response: class
Predictors: area precip maxtemp mintemp flow strslope wsslope
Group g m x
Count 2 5 6
Summary of Classification
Put into ....True Group....
Group g m x
g 2 0 0
m 0 4 1
x 0 1 5
Total N 2 5 6
N Correct 2 4 5
Proportion 1.000 0.800 0.833
N = 13 N Correct = 11 Proportion Correct = 0.846
Squared Distance Between Groups
g m x
g 0.0000 14.5434 17.0393
m 14.5434 0.0000 4.5539
x 17.0393 4.5539 0.0000
Linear Discriminant Function for Group
g m x
Constant -7379.7 -7053.9 -7003.2
area -0.0 -0.0 -0.0
precip 85.6 83.5 83.6
maxtemp 86.5 84.2 84.3
mintemp 157.5 155.0 153.0
flow 0.8 0.7 0.7
strslope -0.3 -0.3 -0.3
wsslope -38.0 -37.0 -36.6
Summary of Misclassified Observations
Observation True Pred Group Squared Probability
Group Group Distance
1 ** x m g 20.956 0.000
m 5.163 0.578
x 5.796 0.421
2 ** m x g 23.906 0.000
m 5.223 0.229
x 2.790 0.771
gauge id majshed class FITS1
g1 NorthBranch x m
g5 Tygart m x
g7 Tygart m m
g10 WestFork x x
g11 MonRiver m m
g12 MonRiver m m
g13 Cheat m m
g17 MiddleOhio x x
g19 Greenbrier x x
g21 Gauley g g
g22 Gauley g g
g24 Elk x x
g26 UpGuyandotte x x
Table 7. Discriminant analysis
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 27
Standardized Variables, Euclidean Distance, Average Linkage
Amalgamation Steps
Step Number of Similarity Distance Clusters New Number of obs.
clusters level level joined cluster in new cluster
1 12 85.24 0.933 1 7 1 2
2 11 80.06 1.261 3 11 3 2
3 10 78.24 1.376 2 9 2 2
4 9 70.39 1.872 5 6 5 2
5 8 69.18 1.948 4 8 4 2
6 7 68.16 2.013 10 12 10 2
7 6 61.42 2.439 3 10 3 4
8 5 56.79 2.732 1 2 1 4
9 4 54.13 2.900 3 13 3 5
10 3 45.06 3.473 4 5 4 4
11 2 41.04 3.727 3 4 3 9
12 1 35.49 4.078 1 3 1 13
Final Partition
Number of clusters: 2
Number of Within cluster Average distance Maximum distance
observations sum of squares from centroid from centroid
Cluster1 4 8.340 1.414 1.816
Cluster2 9 46.098 2.188 2.934
Cluster Centroids
Variable Cluster1 Cluster2 Grand centrd
area -0.7923 0.3522 -0.0000
precip 0.9578 -0.4257 -0.0000
maxtemp -1.2045 0.5353 -0.0000
mintemp -1.1175 0.4966 0.0000
flow -0.7181 0.3191 -0.0000
strslope -0.0511 0.0227 -0.0000
wsslope -0.5946 0.2643 -0.0000
Distances Between Cluster Centroids
Cluster1 Cluster2
Cluster1 0.0000 3.2672
Cluster2 3.2672 0.0000
Table 8. Hierarchical cluster analysis of observations
From the clustered results, gauges 1 and 7 (g1 and g13) are the most alike and merge into a
cluster at around 85 on the similarity scale. Gauges 3 and 11 (g7 and g22) are the next most
similar at the 78 level. However, these objects do not form the same cluster until a lower
level of similarity around the 35 level. By clustering the objects, we were able to identify
groups that are alike and because of the small dataset, it was easy to examine the data table
and discover values that make the objects similar.
After cluster analysis, the choice was made to perform a factor analysis as an aid in data
reduction. Although there were only seven variables, the possibility existed to gain insight
into removing the duplicated information from among the set of variables. The results were
assembled as a loading plot – figure 22, a score plot – figure 23, and a scree plot – figure 24.
The output session data is listed in table 9.
Water Resources Management and Modeling
28
Fig. 21.
Fig. 22.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 29
Fig. 23.
Fig. 24.
Water Resources Management and Modeling
30
Principal Component Factor Analysis of the Correlation Matrix
Unrotated Factor Loadings and Communalities
Variable Factor1 Factor2 Communality
area -0.748 -0.512 0.822
precip 0.720 -0.639 0.927
maxtemp -0.913 0.291 0.919
mintemp -0.949 0.192 0.937
flow -0.559 -0.737 0.855
strslope 0.117 -0.821 0.687
wsslope -0.731 -0.286 0.616
Variance 3.6732 2.0898 5.7630
% Var 0.525 0.299 0.823
Rotated Factor Loadings and Communalities
Varimax Rotation
Variable Factor1 Factor2 Communality
area 0.243 0.874 0.822
precip -0.962 0.025 0.927
maxtemp 0.886 0.365 0.919
mintemp 0.849 0.465 0.937
flow -0.047 0.924 0.855
strslope -0.618 0.552 0.687
wsslope 0.375 0.689 0.616
Variance 3.0164 2.7466 5.7630
% Var 0.431 0.392 0.823
Factor Score Coefficients
Variable Factor1 Factor2
area -0.002 0.319
precip -0.347 0.108
maxtemp 0.280 0.054
mintemp 0.257 0.096
flow -0.111 0.368
strslope -0.277 0.280
wsslope 0.064 0.233
Table 9. Factor analysis
From these results, the variables high in loadings on a particular factor would be those
which are highly correlated with one another, but which have little or no correlation with
the variables loading highly on the other factors. The negative loading variable has a
meaning opposite to that of the factor. The size of the loading is an indication of the extent
to which the variable correlates with the factor.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 31
Limitations, and Discussion of Results
The limitations with this study can be attributed to the number of gauges used and the
variables used to predict stream flow. With more complete data over the state, it would have
been able to assemble more gauges for this component of the project. Also, if possible it would
have been good to include variables used to describe interception, evapotranspiration,
infiltration, interflow, saturated overland flow, and baseflow from groundwater. The rate and
areal distribution of rainfall input would have been helpful in establishing the catchment
retention.
Other issues with the data collection make the estimation of stream flow difficult. First, there
is very high variability in recording stream flow data. The stream flow variable exhibited the
highest standard deviation and variation across the year. Second, taking yearly annual
averages was a crude method in which to characterize the varying conditions that occur
across seasons, months, weeks, and even days. Third, the precipitation and temperature
data used in the study needed to be better allocated to the gauge drainage areas (as
compared to using the drainage area average for the variable) because of the amount of
variability that is present in the entire watershed for the precipitation and temperature data.
Overall, the choice of variables to analyze were appropriate based on the success other
studies found. In the study the results of the multivariate regression indicated that stream
flow could best be estimated using area, stream slope and 30 year annual average maximum
temperature. Other data analysis techniques revealed the correlation present between the
two temperature variables, flow and area, and precipitation to the two temperature
variables.
The last important summary from the tests came from the cluster analysis that grouped the
gauge station objects based on similarity. The grouped gauges shared the same ecoregions.
Ecoregions are defined as "regions of relative homogeneity in ecological systems or in
relationships between organisms and their environments" (Omernik 1987). Omernik (1987)
mapped the ecoregions of the conterminous United States, based on regional patterns in
individual maps of land use, land surface form, potential natural vegetation, and soils. A
discriminant analysis using the ecoregion of each gauge station catchment area would have
been a better choice than the using the major river basins used in this study. The similar
gauge station catchment areas identified by the cluster analysis and the associated ecoregion
borders in West Virginia are displayed in figure 25.
Component 6. The environmental database
An environmental database of point data was included within SWPS. These points are
found in the shapefiles directory of SWPS and are loaded for viewing when a user defines a
study area location in the state. A brief listing of some of the files in the environmental
database follows:
 National pollution discharge elimination system sites
 Landfills
 Superfund sites (CERCLIS)
 Hazardous and solid waste sites (RCRIS)
 Toxic release inventory sites
 Coal dams
Water Resources Management and Modeling
32
 Abandoned mine land locations
 Animal feed lots
 Major highways
 Railroads
Fig. 25.
Component 7. UTM latitude/longitude conversion utility
This capability of SWPS allows the user to map coordinates in degrees, minutes and seconds
by using an input dialog screen. The user’s points are then mapped in the UTM zone 17
projection. Points may be added to an existing point feature theme or a new point theme can
be created. The ability to type coordinates and have the points reprojected saves the user
many extra steps. In addition to mapping points from user input, a point can be queried for
its x and y locations in UTM, stateplane, or latitude and longitude coordinates. The user can
identify locations quickly by clicking anywhere in the display to report this information.
Component 8. Statewide map/GIS data layers
All GIS data is organized in the shapefiles and grids directories of SWPS. This data is listed
below. These datasets are provided in addition to the data listed in the environmental
database discussed in component 6.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 33
 Coded hydrology  Roads (1:100K scale)  EPA MRLC land use/cover
 Watersheds  Cities  Digital Raster Graphics
 Lakes/impoundments  USGS gauging stations  SPOT imagery
 Counties  Public wells  Digital Elevation Model
 1:24K quads  Runoff grid  Hillshaded relief
 Major river basins  Flow direction  Elevation TIN
 Abandoned mine lands  Flow accumulation  303-D listed streams
 Watersheds with fish
collection data
 Stream orders
 Major Rivers
 WV GAP land use/ cover
 Bond forfeiture sites
 Expected mean
concentration grid
 Public wells
 Groundwater wells
 Surface water wells
 Cumulative runoff
 Coal Geology
 Override 7Q10 streams
 Landfills
 NW Wetlands
 Springs over 500gpm
 1950 land use/cover
 Shreve stream orders
 Stream length from
mouth
 DRAFT 14 digit HUCS
 Wet weather streams
 Surface mine inventory sites
 Public wells
 Strahler stream orders
 Stream slope
 Max and min stream
elevations
 Surface water zones of
critical concern
 Streamflow
Component 9. Water quality modeling capability
The water quality modeling capabilities of SWPS are built using a landscape driven
approach that uses a predefined runoff and cumulative runoff grid to drive the analysis. It is
essentially a weighted mass balance approach that will show changing concentrations and
loadings based on changing flow conditions only. The runoff grid is based on a relationship
between rainfall and stream flow. It is the main factor that directs flow directions to the
stream or steepest path direction and estimates the stream flow.
The assumptions/limitations of this water quality modeling approach are the following:
1. Streams have the same hydrogeometric properties (stream slope, roughness,
width, and depth).
2. Also assumed are that the streams have the same ecological rate constants
(reareation rates, pollution decay rates and sediment oxygen demand rate).
3. Transport of pollutants is considered to be conservative (values get averaged over
changing flow conditions only) -> no loss or decay of pollutants is considered Does
not consider infiltration, or ground water flow additions
5. Does not include atmospheric conditions such as evapotranspiration
The water quality model in SWPS can be used in two different ways. The first is when the
user has collected point locations of water quality data and wants to associate the sampled
data to instream concentrations and loadings downstream of the sampling points. This is
essentially a weighted mass balance approach using the stream flow and sampled locations
to associate the point location information to stream condition. The input data using this
method needs to be in Mg/L. The resultant modeled levels are reported back as stream
values in Mg/L for concentration and Kg/Yr. for loading. The advantage of this first
method of using sampled data is that it allows the user to see how the data location
information can be used to estimate downstream conditions away from the sampling site.
Water Resources Management and Modeling
34
The second way the water quality model in SWPS can be used is in estimating total
nitrogen, phosphorous and total suspended solids as concentrations and loadings in the
stream based on expected mean concentrations from land use/cover classes. This method
does not require any sampled water quality but uses the cover classes from a land use/cover
grid (30meter-cell size). The thirteen classes for West Virginia from this data set were
aggregated to six general classes because loading values for nitrogen, phosphorous and total
suspended solids were only available for those six classes. The aggregated classes and the
corresponding classes included:
- Urban (low intensity developed, high intensity developed, residential)
- Open/Brush (hay, pasture grass, mixed pasture, other grasses)
- Agriculture (row crops)
- Woodland (conifer forest, mixed forest, deciduous forest)
- Barren (quarry areas, barren transitional areas)
- Wetland (emergent and woody wetlands)
The classes are associated with expected loadings based on the acreage size of the class. The
loadings are annual averages and when used with the modeled stream flow can give
concentration and loading results for the stream. The cover classes and associated expected
mean concentrations levels used in the model are shown below.
Total Nitrogen Total Phosphorous Total Suspended Solids
Urban 1.89 0.009 166
Open/Brush 2.19 0.13 70
Agriculture 3.41 0.24 201
Woodland 0.79 0.006 39
Barren 3.90 0.10 2200
Wetland 0.79 0.006 39
The nutrient export coefficients above are multiplied by the amount (area) of a given land
cover type. It is used as a simulation to estimate the probability of increased nutrient loads
from land cover composition. It should be noted that there are factors other than land cover
that contribute to nutrient export and these are rarely known with certainty. Some of the
factors that may vary across watersheds and may change the expected mean concentration
results include:
 year to year changes in precipitation
 soil type
 slope and slope morphology (convex, concave)
 geology
 cropping practices
 timing of fertilizer application relative to precipitation events
 density of impervious surface
The loading and concentration results in consideration of these assumptions however can
still give insight in comparing expected pollutant values for watersheds. The results should
be thought of in most cases as the worst case scenarios for stream water quality levels.
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 35
Component 10. Groundwater and surface water susceptibility model
The susceptibility ranking model within SWPS was constructed using the steps defined
within the West Virginia Surface Water Assessment Program (SWAP) document. The
susceptibility ranking for ground water systems was based on the physical integrity of the
well and spring infrastructure; hydrologic setting; inventory of potential contaminant
sources and land uses, and water quality. The susceptibility ranking for surface water
systems was based on water quality and the inventory of potential contaminant sources.
A more detailed explanation of the ground and surface water susceptibility models
follows.
Groundwater susceptibility model
To determine the groundwater susceptibility for a site, the physical barrier effectiveness is
first calculated. Physical barrier effectiveness is the Tier 1 assessment. It is used to note if
there is a known impact on water quality, evaluate the source integrity as low or high, and
to find the aquifer vulnerability. Based on these results, the physical barrier effectiveness can
be determined as having high, moderate, or low potential susceptibility. If there is a known
impact on water quality, then the model automatically goes to the Tier Two assessment and
sets the groundwater susceptibility as being high. If there is no known impact on water
quality then the source integrity and aquifer vulnerability set the physical barrier
effectiveness for the Tier Two assessment. The aquifer vulnerability is determined from the
different scenarios listed below;
If Then
All springs High Aquifer Sensitivity
Alluvial Valleys
Unconfined High Aquifer Sensitivity
Confined Moderate Aquifer Sensitivity
Appalachian Plateau Province (fracture) Moderate Aquifer Sensitivity
Folded Plateau Area (fracture) Moderate Aquifer Sensitivity
Karst Areas High Aquifer Sensitivity
Valley and Ridge Province (fracture) Moderate Aquifer Sensitivity
Coal Mine Areas High Sensitivity
From the above scenarios, an aquifer vulnerability was determined. Using this with the
source integrity rating can provide physical barrier effectiveness. Physical barrier
effectiveness is
 High if there is low source integrity and high aquifer sensitivity.
 Low if there is high source integrity and moderate aquifer sensitivity
 Moderate if there is high source integrity and high aquifer sensitivity or low source
integrity and moderate aquifer sensitivity
Again, if there is no known impact on water quality, this method will determine the
physical barrier effectiveness as being high, moderate or low susceptibility. If there is a
known water quality impact then the final groundwater susceptibility is high.
Water Resources Management and Modeling
36
Using the physical barrier effectiveness with the land use concern level determines the
final groundwater susceptibility. The land use concern level is determined from the
percentage of land use in the buffered groundwater site. The percentage of land use was
found for every buffered location. In cases where the groundwater site had a pumping
rate over 25,000 gpd, no buffer was created. For these groundwater sites no land use
percentage was calculated.
The final groundwater susceptibility is rated as:
 High if the physical barrier effectiveness is high
 High if the land use concern is high and the physical barrier effectiveness is moderate
 Moderate if the land use concern is medium or low and the physical barrier
effectiveness is moderate
 Moderate if the land use concern is high or medium and the physical barrier
effectiveness is low
 Low if the land use concern is low and the physical barrier effectiveness is low
The percentage of land use is reported to the user before the tier one assessment appears. He
or she needs to know the associated concern levels with the percentage of land uses that are
reported for each groundwater site. By not hard coding in the land use concern levels for
each buffer, the user has the ability to perform “what if” type scenarios if existing land use
changes or is different than what currently exists.
Surface water susceptibility model
The surface water susceptibility model is slightly less complicated than the groundwater
model. For the surface water susceptibility determination, the percent of land use was
calculated for each of the zone of critical concern. The land use concern level, and if there is
a known water quality impact, are the two factors which are used to determine the surface
water susceptibility. As in the groundwater model, the percent land use is presented to the
user before the model is run so the user can make a determination and perform “what if”
type scenarios with differing land use within the zone of critical concern. If there is a known
water quality impact, then the surface water susceptibility is automatically high. If there is
no known water quality impact, then the final surface water susceptibility is:
 High if land use concern level is high
 High if land use concern level is medium
 Low if land use concern level is moderate
Summary and Conclusion
Drinking water is a critical resource that continues to need protection and management to
assure safe supplies for the public. Since agencies to protect water resources operate at
mostly state jurisdictions, it is important to implement a system at a statewide level. This
chapter discussed watershed tools that integrate spatially explicit data and decision support
to assist managers with both surface and ground water resources. It has three major
components which include; an ability to delineate source water protection areas upstream of
supply water, an inventory of potential contamination sources within various zones of
critical concern, and the determination of the public drinking water supply systems
Tools for Watershed Planning –
Development of a Statewide Source Water Protection System (SWPS) 37
susceptibility to contamination. The current system provides the ability to assess, preserve,
and protect the states source waters for public drinking.
2. References
Anderson, L. and A. Lepisto. 1998. “Links Between Runoff Generation, Climate and Nitrate-
N Leaching From Forested Catchments.” Water, Air, and Soil Pollution. 105: pp227-
237.
Astatkie, T. and W. E. Watt. 1998. “Multiple-Input Transfer Function Modeling of Daily
Stream flow Series Using Nonlinear Inputs.” Water Resources Research. Vol. 34,
No.10, pp2217-2725.
Environmental Systems Research Institute (ESRI). 2010. ArcInfo ArcMap. Redlands, CA.
Gabriele, H. M., and F. E. Perkins. 1997. “Watershed-Specific Model for Stream flow,
Sediment, and Metal Transport.” Journal of Environmental Engineering. pp61-69.
Garren, D. C. 1992. “Improved Techniques in Regression-Based Stream flow Volume
Forecasting.” Journal of Water Resources Planning and Management. Vol. 118, No.
6. pp654-669.
Hocking, R. R. 1996. Methods and Applications of Linear Models. Wiley and Sons, New
York, NY.
Johnston, K., J. M. Ver Hoef, K. Krivoruchko, and N. Lucas. 2001. Using ArcGis
Geostatistical Analyst. Environmental System Research Institute, Redlands, CA.
Kachigan, S. K. 1986. Statistical Analysis. Radius Press, New York, NY.
Moore, R. D. 1997. “Storage-Outflow Modeling of Stream flow Recessions, with
Application to Shallow-Soil Forested Catchments.” Journal of Hydrology.
198(1997) pp260-270.
Omernik, J. M. 1987. “Ecoregions of the conterminous United States.” Annals
of the Association of American Geographers Vol. 77, pp118-125.
Querner, E. P., Tallaksen, L. M., Kasparek, L., and H. A. J. Van Lanen. 1997. “Impact of
Land-Use, Climate Change and Groundwater Abstraction on Stream flow
Droughts Using Physically-Based Models.” Regional Hydrology: Concepts and
Models for Sustainable Water Resource Management. IAHS Publ. No. 246, pp171-
179.
Sharma, A., D. G. Tarboton, and U. Lall. 1997. “Stream flow Simulation: A Nonparametric
Approach.” Water Resources Research. Vol. 33, No. 2, pp291-308.
Smith, J. A. 1991. “Long-Range Stream flow Forecasting Using Nonparametric Regression.”
Water Resources Bulletin. Vol. 27, No. 1, pp39-46.
Sundberg, R. 2002. Collinearity. Encyclopedia of Environmetrics, Vol. 1. John Wiley and
Sons, Ltd, Chichester, West Sussex UK.
Timofeyeva, M. and R. Craig. 1998. “Using Monte Carlo Technique for Modelling the
Natural Variability of Stream flow in Headwaters of the Sierra Nevada, USA.”
Hydrology, Water Resources and Ecology in Headwaters. 1998. No. 248, pp59-
65.
U. S. Geological Survey. 1996. “Prediction of Travel Time and Longitudinal Dispersion in
Rivers and Streams.” Water-Resources Investigations Report 96-4013.
Water Resources Management and Modeling
38
West Virginia Department of Health and Human Resources. August 1, 1999. “State of West
Virginia Source Water Assessment and Protection Program.” Office of
Environmental Health Services, Charleston, WV.
Another Random Scribd Document
with Unrelated Content
CHAPTER VIII
BAD SHEPHERDS AND FALSE PROPHETS
xxiii., xxiv.
"Woe unto the shepherds that destroy and scatter the sheep of
My pasture!"—Jer. xxiii. 1.
"Of what avail is straw instead of grain?... Is not My word like
fire, ... like a hammer that shattereth the rocks?"—Jer. xxiii. 28,
29.
The captivity of Jehoiachin and the deportation of the flower of the
people marked the opening of the last scene in the tragedy of Judah
and of a new period in the ministry of Jeremiah. These events,
together with the accession of Zedekiah as Nebuchadnezzar's
nominee, very largely altered the state of affairs in Jerusalem. And
yet the two main features of the situation were unchanged—the
people and the government persistently disregarded Jeremiah's
exhortations. "Neither Zedekiah, nor his servants, nor the people of
the land, did hearken unto the words of Jehovah which He spake by
the prophet Jeremiah."[101] They would not obey the will of Jehovah
as to their life and worship, and they would not submit to
Nebuchadnezzar. "Zedekiah ... did evil in the sight of Jehovah,
according to all that Jehoiakim had done; ... and Zedekiah rebelled
against the king of Babylon."[102]
It is remarkable that though Jeremiah consistently urged submission
to Babylon, the various arrangements made by Nebuchadnezzar did
very little to improve the prophet's position or increase his influence.
The Chaldean king may have seemed ungrateful only because he
was ignorant of the services rendered to him—Jeremiah would not
enter into direct and personal co-operation with the enemy of his
country, even with him whom Jehovah had appointed to be the
scourge of His disobedient people—but the Chaldean policy served
Nebuchadnezzar as little as it profited Jeremiah. Jehoiakim, in spite
of his forced submission, remained the able and determined foe of
his suzerain, and Zedekiah, to the best of his very limited ability,
followed his predecessor's example.
Zedekiah was uncle of Jehoiachin, half-brother of Jehoiakim, and
own brother to Jehoahaz.[103] Possibly the two brothers owed their
bias against Jeremiah and his teaching to their mother, Josiah's wife
Hamutal, the daughter of another Jeremiah, the Libnite. Ezekiel thus
describes the appointment of the new king: "The king of Babylon ...
took one of the seed royal, and made a covenant with him; he also
put him under an oath, and took away the mighty of the land: that
the kingdom might be base, that it might not lift itself up, but that
by keeping of his covenant it might stand."[104] Apparently
Nebuchadnezzar was careful to choose a feeble prince for his "base
kingdom"; all that we read of Zedekiah suggests that he was weak
and incapable. Henceforth the sovereign counted for little in the
internal struggles of the tottering state. Josiah had firmly maintained
the religious policy of Jeremiah, and Jehoiakim, as firmly, the
opposite policy; but Zedekiah had neither the strength nor the
firmness to enforce a consistent policy and to make one party
permanently dominant. Jeremiah and his enemies were left to fight
it out amongst themselves, so that now their antagonism grew more
bitter and pronounced than during any other reign.
But whatever advantage the prophet might derive from the
weakness of the sovereign was more than counterbalanced by the
recent deportation. In selecting the captives Nebuchadnezzar had
sought merely to weaken Judah by carrying away every one who
would have been an element of strength to the "base kingdom."
Perhaps he rightly believed that neither the prudence of the wise nor
the honour of the virtuous would overcome their patriotic hatred of
subjection; weakness alone would guarantee the obedience of
Judah. He forgot that even weakness is apt to be foolhardy—when
there is no immediate prospect of penalty.
One result of his policy was that the enemies and friends of
Jeremiah were carried away indiscriminately; there was no attempt
to leave behind those who might have counselled submission to
Babylon as the acceptance of a Divine judgment, and thus have
helped to keep Judah loyal to its foreign master. On the contrary
Jeremiah's disciples were chiefly thoughtful and honourable men,
and Nebuchadnezzar's policy in taking away "the mighty of the land"
bereft the prophet of many friends and supporters, amongst them
his disciple Ezekiel and doubtless a large class of whom Daniel and
his three friends might be taken as types. When Jeremiah
characterises the captives as "good figs" and those left behind as
"bad figs,"[105] and the judgment is confirmed and amplified by
Ezekiel,[106] we may be sure that most of the prophet's adherents
were in exile.
We have already had occasion to compare the changes in the
religious policy of the Jewish government to the alternations of
Protestant and Romanist sovereigns among the Tudors; but no Tudor
was as feeble as Zedekiah. He may rather be compared to Charles
IX. of France, helpless between the Huguenots and the League. Only
the Jewish factions were less numerous, less evenly balanced; and
by the speedy advance of Nebuchadnezzar civil dissensions were
merged in national ruin.
The opening years of the new reign passed in nominal allegiance to
Babylon. Jeremiah's influence would be used to induce the vassal
king to observe the covenant he had entered into and to be faithful
to his oath to Nebuchadnezzar. On the other hand a crowd of
"patriotic" prophets urged Zedekiah to set up once more the
standard of national independence, to "come to the help of the Lord
against the mighty." Let us then briefly consider Jeremiah's polemic
against the princes, prophets, and priests of his people. While
Ezekiel in a celebrated chapter[107] denounces the idolatry of the
princes, priests, and women of Judah, their worship of creeping
things and abominable beasts, their weeping for Tammuz, their
adoration of the sun, Jeremiah is chiefly concerned with the perverse
policy of the government and the support it receives from priests
and prophets, who profess to speak in the name of Jehovah.
Jeremiah does not utter against Zedekiah any formal judgment like
those on his three predecessors. Perhaps the prophet did not regard
this impotent sovereign as the responsible representative of the
state, and when the long-expected catastrophe at last befell the
doomed people, neither Zedekiah nor his doings distracted men's
attention from their own personal sufferings and patriotic regrets. At
the point where a paragraph on Zedekiah would naturally have
followed that on Jehoiachin, we have by way of summary and
conclusion to the previous sections a brief denunciation of the
shepherds of Israel.
"Woe unto the shepherds that destroy and scatter the sheep of My
pasture!... Ye have scattered My flock, and driven them away, and
have not cared for them; behold, I will visit upon you the evil of your
doings."
These "shepherds" are primarily the kings, Jehoahaz, Jehoiakim, and
Jehoiachin, who have been condemned by name in the previous
chapter, together with the unhappy Zedekiah, who is too insignificant
to be mentioned. But the term shepherds will also include the ruling
and influential classes of which the king was the leading
representative.
The image is a familiar one in the Old Testament and is found in the
oldest literature of Israel,[108] but the denunciation of the rulers of
Judah as unfaithful shepherds is characteristic of Jeremiah, Ezekiel,
and one of the prophecies appended to the Book of Zechariah.[109]
Ezekiel xxxiv. expands this figure and enforces its lessons:—
"Woe unto the shepherds of Israel that do feed
themselves!
Should not the shepherds feed the sheep? Ye eat the
fat, and ye clothe you with the wool.
Ye kill the fatlings; but ye feed not the sheep.
The diseased have ye not strengthened,
Neither have ye healed the sick,
Neither have ye bound up the bruised,
Neither have ye brought back again that which was
driven away,
Neither have ye sought for that which was lost,
But your rule over them has been harsh and violent.
And for want of a shepherd, they were scattered,
And became food for every beast of the field."[110]
So in Zechariah ix., etc., Jehovah's anger is kindled against the
shepherds, because they do not pity His flock.[111] Elsewhere[112]
Jeremiah speaks of the kings of all nations as shepherds, and
pronounces against them also a like doom. All these passages
illustrate the concern of the prophets for good government. They
were neither Pharisees nor formalists; their religious ideals were
broad and wholesome. Doubtless the elect remnant will endure
through all conditions of society; but the Kingdom of God was not
meant to be a pure Church in a rotten state. This present evil world
is no manure heap to fatten the growth of holiness: it is rather a
mass for the saints to leaven.
Both Jeremiah and Ezekiel turn from the unfaithful shepherds whose
"hungry sheep look up and are not fed" to the true King of Israel,
the "Shepherd of Israel that led Joseph like a flock, and dwelt
between the Cherubim." In the days of the Restoration He will raise
up faithful shepherds, and over them a righteous Branch, the real
Jehovah Zidqenu, instead of the sapless twig who disgraced the
name "Zedekiah." Similarly Ezekiel promises that God will set up one
shepherd over His people, "even My servant David." The pastoral
care of Jehovah for His people is most tenderly and beautifully set
forth in the twenty-third Psalm. Our Lord, the root and the offspring
of David, claims to be the fulfilment of ancient prophecy when He
calls Himself "the Good Shepherd." The words of Christ and of the
Psalmist receive new force and fuller meaning when we contrast
their pictures of the true Shepherd with the portraits of the Jewish
kings drawn by the prophets. Moreover the history of this metaphor
warns us against ignoring the organic life of the Christian society, the
Church, in our concern for the spiritual life of the individual. As Sir
Thomas More said, in applying this figure to Henry VIII., "Of the
multitude of sheep cometh the name of a shepherd."[113] A
shepherd implies not merely a sheep, but a flock; His relation to
each member is tender and personal, but He bestows blessings and
requires service in fellowship with the Family of God.
By a natural sequence the denunciation of the unfaithful shepherds
is followed by a similar utterance "concerning the prophets." It is
true that the prophets are not spoken of as shepherds; and Milton's
use of the figure in Lycidas suggests the New Testament rather than
the Old. Yet the prophets had a large share in guiding the destinies
of Israel in politics as well as in religion, and having passed sentence
on the shepherds—the kings and princes—Jeremiah turns to the
ecclesiastics, chiefly, as the heading implies, to the prophets. The
priests indeed do not escape, but Jeremiah seems to feel that they
are adequately dealt with in two or three casual references. We use
the term "ecclesiastics" advisedly; the prophets were now a large
professional class, more important and even more clerical than the
priests. The prophets and priests together were the clergy of Israel.
They claimed to be devoted servants of Jehovah, and for the most
part the claim was made in all sincerity; but they misunderstood His
character, and mistook for Divine inspiration the suggestions of their
own prejudice and self-will.
Jeremiah's indictment against them has various counts. He accuses
them of speaking without authority, and also of time-serving,
plagiarism, and cant.
First, then, as to their unauthorised utterances: Jeremiah finds them
guilty of an unholy licence in prophesying, a distorted caricature of
that "liberty of prophesying" which is the prerogative of God's
accredited ambassadors.
"Hearken not unto the words of the prophets that
prophesy unto you.
They make fools of you:
The visions which they declare are from their own
hearts,
And not from the mouth of Jehovah.
Who hath stood in the council of Jehovah,
To perceive and hear His word?
Who hath marked His word and heard it?
I sent not the prophets—yet they ran;
I spake not unto them—yet they prophesied."
The evils which Jeremiah describes are such as will always be found
in any large professional class. To use modern terms—in the Church,
as in every profession, there will be men who are not qualified for
the vocation which they follow. They are indeed not called to their
vocation; they "follow," but do not overtake it. They are not sent of
God, yet they run; they have no Divine message, yet they preach.
They have never stood in the council of Jehovah; they might
perhaps have gathered up scraps of the King's purposes from His
true councillors; but when they had opportunity they neither
"marked nor heard"; and yet they discourse concerning heavenly
things with much importance and assurance. But their inspiration, at
its best, has no deeper or richer source than their own shallow
selves; their visions are the mere product of their own imaginations.
Strangers to the true fellowship, their spirit is not "a well of water
springing up unto eternal life," but a stagnant pool. And, unless the
judgment and mercy of God intervene, that pool will in the end be
fed from a fountain whose bitter waters are earthly, sensual,
devilish.
We are always reluctant to speak of ancient prophecy or modern
preaching as a "profession." We may gladly dispense with the word,
if we do not thereby ignore the truth which it inaccurately expresses.
Men lived by prophecy, as, with Apostolic sanction, men live by "the
gospel." They were expected, as ministers are now, though in a less
degree, to justify their claims to an income and an official status, by
discharging religious functions so as to secure the approval of the
people or the authorities. Then, as now, the prophet's reputation,
influence, and social standing, probably even his income, depended
upon the amount of visible success that he could achieve.
In view of such facts, it is futile to ask men of the world not to speak
of the clerical life as a profession. They discern no ethical difference
between a curate's dreams of a bishopric and the aspirations of a
junior barrister to the woolsack. Probably a refusal to recognise the
element common to the ministry with law, medicine, and other
professions, injures both the Church and its servants. One peculiar
difficulty and most insidious temptation of the Christian ministry
consists in its mingled resemblances to and differences from the
other professions. The minister has to work under similar worldly
conditions, and yet to control those conditions by the indwelling
power of the Spirit. He has to "run," it may be twice or even three
times a week, whether he be sent or no: how can he always preach
only that which God has taught him? He is consciously dependent
upon the exercise of his memory, his intellect, his fancy: how can he
avoid speaking "the visions of his own heart"? The Church can never
allow its ministers to regard themselves as mere professional
teachers and lecturers, and yet if they claim to be more, must they
not often fall under Jeremiah's condemnation?
It is one of those practical dilemmas which delight casuists and
distress honest and earnest servants of God. In the early Christian
centuries similar difficulties peopled the Egyptian and Syrian deserts
with ascetics, who had given up the world as a hopeless riddle. A full
discussion of the problem would lead us too far away from the
exposition of Jeremiah, and we will only venture to make two
suggestions.
The necessity, which most ministers are under, of "living by the
gospel," may promote their own spiritual life and add to their
usefulness. It corrects and reduces spiritual pride, and helps them to
understand and sympathise with their lay brethren, most of whom
are subject to a similar trial.
Secondly, as a minister feels the ceaseless pressure of strong
temptation to speak from and live for himself—his lower, egotistic
self—he will be correspondingly driven to a more entire and
persistent surrender to God. The infinite fulness and variety of
Revelation is expressed by the manifold gifts and experience of the
prophets. If only the prophet be surrendered to the Spirit, then what
is most characteristic of himself may become the most forcible
expression of his message. His constant prayer will be that he may
have the child's heart and may never resist the Holy Ghost, that no
personal interest or prejudice, no bias of training or tradition or
current opinion, may dull his hearing when he stands in the council
of the Lord, or betray him into uttering for Christ's gospel the
suggestions of his own self-will or the mere watchwords of his
ecclesiastical faction.
But to return to the ecclesiastics who had stirred Jeremiah's wrath.
The professional prophets naturally adapted their words to the
itching ears of their clients. They were not only officious, but also
time-serving. Had they been true prophets, they would have dealt
faithfully with Judah; they would have sought to convince the people
of sin, and to lead them to repentance; they would thus have given
them yet another opportunity of salvation.
"If they had stood in My council,
They would have caused My people to hear My
words;
They would have turned them from their evil way,
And from the evil of their doings."
But now:—
"They walk in lies and strengthen the hands of
evildoers,
That no one may turn away from his sin.
They say continually unto them that despise the word
of Jehovah,[114]
Ye shall have peace;
And unto every one that walketh in the stubbornness
of his heart they say,
No evil shall come upon you."
Unfortunately, when prophecy becomes professional in the lowest
sense of the word, it is governed by commercial principles. A
sufficiently imperious demand calls forth an abundant supply. A
sovereign can "tune the pulpits"; and a ruling race can obtain from
its clergy formal ecclesiastical sanction for such "domestic
institutions" as slavery. When evildoers grow numerous and
powerful, there will always be prophets to strengthen their hands
and encourage them not to turn away from their sin. But to give the
lie to these false prophets God sends Jeremiahs, who are often
branded as heretics and schismatics, turbulent fellows who turn the
world upside-down.
The self-important, self-seeking spirit leads further to the sin of
plagiarism:—
"Therefore I am against the prophets, is the utterance
of Jehovah,
Who steal My word from one another."
The sin of plagiarism is impossible to the true prophet, partly
because there are no rights of private property in the word of
Jehovah. The Old Testament writers make free use of the works of
their predecessors. For instance, Isaiah ii. 2-4 is almost identical with
Micah iv. 1-3; yet neither author acknowledges his indebtedness to
the other or to any third prophet.[115] Uriah ben Shemaiah
prophesied according to all the words of Jeremiah,[116] who himself
owes much to Hosea, whom he never mentions. Yet he was not
conscious of stealing from his predecessor, and he would have
brought no such charge against Isaiah or Micah or Uriah. In the New
Testament 2 Peter and Jude have so much in common that one must
have used the other without acknowledgment. Yet the Church has
not, on that ground, excluded either Epistle from the Canon. In the
goodly fellowship of the prophets and the glorious company of the
apostles no man says that the things which he utters are his own.
But the mere hireling has no part in the spiritual communism
wherein each may possess all things because he claims nothing.
When a prophet ceases to be the messenger of God, and sinks into
the mercenary purveyor of his own clever sayings and brilliant
fancies, then he is tempted to become a clerical Autolycus, "a
snapper-up of unconsidered trifles." Modern ideas furnish a curious
parallel to Jeremiah's indifference to the borrowings of the true
prophet, and his scorn of the literary pilferings of the false. We hear
only too often of stolen sermons, but no one complains of plagiarism
in prayers. Doubtless among these false prophets charges of
plagiarism were bandied to and fro with much personal acrimony.
But it is interesting to notice that Jeremiah is not denouncing an
injury done to himself; he does not accuse them of thieving from
him, but from one another. Probably assurance and lust of praise
and power would have overcome any awe they felt for Jeremiah. He
was only free from their depredations, because—from their point of
view—his words were not worth stealing. There was nothing to be
gained by repeating his stern denunciations, and even his promises
were not exactly suited to the popular taste.
These prophets were prepared to cater for the average religious
appetite in the most approved fashion—in other words, they were
masters of cant. Their office had been consecrated by the work of
true men of God like Elijah and Isaiah. They themselves claimed to
stand in the genuine prophetic succession, and to inherit the
reverence felt for their great predecessors, quoting their inspired
utterances and adopting their weighty phrases. As Jeremiah's
contemporaries listened to one of their favourite orators, they were
soothed by his assurances of Divine favour and protection, and their
confidence in the speaker was confirmed by the frequent sound of
familiar formulæ in his unctuous sentences. These had the true ring;
they were redolent of sound doctrine, of what popular tradition
regarded as orthodox.
The solemn attestation NE'UM YAHWE, "It is the utterance of
Jehovah," is continually appended to prophecies, almost as if it were
the sign-manual of the Almighty. Isaiah and other prophets
frequently use the term MASSA (A.V., R.V., "burden") as a title,
especially for prophecies concerning neighbouring nations. The
ancient records loved to tell how Jehovah revealed Himself to the
patriarchs in dreams. Jeremiah's rivals included dreams in their
clerical apparatus:—
"Behold, I am against them that prophesy lying
dreams—
Ne'um Yahwe—
And tell them, and lead astray My people
By their lies and their rodomontade;
It was not I who sent or commanded them,
Neither shall they profit this people at all,
Ne'um Yahwe"
These prophets "thought to cause the Lord's people to forget His
name, as their fathers forgot His name for Baal, by their dreams
which they told one another."
Moreover they could glibly repeat the sacred phrases as part of their
professional jargon:—
"Behold, I am against the prophets,
It is the utterance of Jehovah (Ne'um Yahwe),
That use their tongues
To utter utterances (Wayyin'amu Ne'um)."
"To utter utterances"—the prophets uttered them, not Jehovah.
These sham oracles were due to no Diviner source than the
imagination of foolish hearts. But for Jeremiah's grim earnestness,
the last clause would be almost blasphemous. It is virtually a
caricature of the most solemn formula of ancient Hebrew religion.
But this was really degraded when it was used to obtain credence
for the lies which men prophesied out of the deceit of their own
heart. Jeremiah's seeming irreverence was the most forcible way of
bringing this home to his hearers. There are profanations of the
most sacred things which can scarcely be spoken of without an
apparent breach of the Third Commandment. The most awful taking
in vain of the name of the Lord God is not heard among the
publicans and sinners, but in pulpits and on the platforms of
religious meetings.
But these prophets and their clients had a special fondness for the
phrase "The burden of Jehovah," and their unctuous use of it most
especially provoked Jeremiah's indignation:—
"When this people, priest, or prophet shall ask thee,
What is the burden of Jehovah?
Then say unto them, Ye are the burden.[117]
But I will cast you off, Ne'um Yahwe.
If priest or prophet or people shall say, The burden of
Jehovah,
I will punish that man and his house.
And ye shall say to one another,
What hath Jehovah answered? and, What hath
Jehovah spoken?
And ye shall no more make mention of the burden of
Jehovah:
For (if ye do) men's words shall become a burden to
themselves.
Thus shall ye inquire of a prophet,
What hath Jehovah answered thee?
What hath Jehovah spoken unto thee?
But if ye say, The burden of Jehovah,
Thus saith Jehovah: Because ye say this word, The
burden of Jehovah,
When I have sent unto you the command,
Ye shall not say, The burden of Jehovah,
Therefore I will assuredly take you up,
And will cast away from before Me both you and the
city which I gave to you and to your fathers.
I will bring upon you everlasting reproach
And everlasting shame, that shall not be forgotten."
Jeremiah's insistence and vehemence speak for themselves. Their
moral is obvious, though for the most part unheeded. The most
solemn formulæ, hallowed by ancient and sacred associations, used
by inspired teachers as the vehicle of revealed truths, may be
debased till they become the very legend of Antichrist, blazoned on
the Vexilla Regis Inferni. They are like a motto of one of Charles's
Paladins flaunted by his unworthy descendants to give distinction to
cruelty and vice. The Church's line of march is strewn with such
dishonoured relics of her noblest champions. Even our Lord's own
words have not escaped. There is a fashion of discoursing upon "the
gospel" which almost tempts reverent Christians to wish they might
never hear that word again. Neither is this debasing of the moral
currency confined to religious phrases; almost every political and
social watchword has been similarly abused. One of the vilest
tyrannies the world has ever seen—the Reign of Terror—claimed to
be an incarnation of "Liberty, Equality, and Fraternity."
Yet the Bible, with that marvellous catholicity which lifts it so high
above the level of all other religious literature, not only records
Jeremiah's prohibition to use the term "Burden," but also tells us
that centuries later Malachi could still speak of "the burden of the
word of Jehovah." A great phrase that has been discredited by
misuse may yet recover itself; the tarnished and dishonoured sword
of faith may be baptised and burnished anew, and flame in the
forefront of the holy war.
Jeremiah does not stand alone in his unfavourable estimate of the
professional prophets of Judah; a similar depreciation seems to be
implied by the words of Amos: "I am neither a prophet nor of the
sons of the prophets."[118] One of the unknown authors whose
writings have been included in the Book of Zechariah takes up the
teaching of Amos and Jeremiah and carries it a stage further:—
"In that day (it is the utterance of Jehovah Sabaoth) I
will cut off the names of the idols from the land,
They shall not be remembered any more;
Also the prophets and the spirit of uncleanness
Will I expel from the land.
When any shall yet prophesy,
His father and mother that begat him shall say unto
him,
Thou shalt not live, for thou speakest lies in the name
of Jehovah:
And his father and mother that begat him shall thrust
him through when he prophesieth.
In that day every prophet when he prophesieth shall
be ashamed of his vision;
Neither shall any wear a hairy mantle to deceive:
He shall say, I am no prophet;
I am a tiller of the ground,
I was sold for a slave in my youth."[119]
No man with any self-respect would allow his fellows to dub him
prophet; slave was a less humiliating name. No family would endure
the disgrace of having a member who belonged to this despised
caste; parents would rather put their son to death than see him a
prophet. To such extremities may the spirit of time-serving and cant
reduce a national clergy. We are reminded of Latimer's words in his
famous sermon to Convocation in 1536: "All good men in all places
accuse your avarice, your exactions, your tyranny. I commanded you
that ye should feed my sheep, and ye earnestly feed yourselves from
day to day, wallowing in delights and idleness. I commanded you to
teach my law; you teach your own traditions, and seek your own
glory."[120]
Over against their fluent and unctuous cant Jeremiah sets the
terrible reality of his Divine message. Compared to this, their sayings
are like chaff to the wheat; nay, this is too tame a figure—Jehovah's
word is like fire, like a hammer that shatters rocks. He says of
himself:—
"My heart within me is broken; all my bones shake:
I am like a drunken man, like a man whom wine hath
overcome,
Because of Jehovah and His holy words."
Thus we have in chapter xxiii. a full and formal statement of the
controversy between Jeremiah and his brother-prophets. On the one
hand, self-seeking and self-assurance winning popularity by orthodox
phrases, traditional doctrine, and the prophesying of smooth things;
on the other hand, a man to whom the word of the Lord was like a
fire in his bones, who had surrendered prejudice and predilection
that he might himself become a hammer to shatter the Lord's
enemies, a man through whom God wrought so mightily that he
himself reeled and staggered with the blows of which he was the
instrument.
The relation of the two parties was not unlike that of St. Paul and his
Corinthian adversaries: the prophet, like the Apostle, spoke "in
demonstration of the Spirit and of power"; he considered "not the
word of them which are puffed up, but the power. For the kingdom
of God is not in word, but in power." In our next chapter we shall
see the practical working of this antagonism which we have here set
forth.
CHAPTER IX
HANANIAH
xxvii., xxviii.
"Hear now, Hananiah; Jehovah hath not sent thee, but thou
makest this people to trust in a lie."—Jer. xxviii. 15.
The most conspicuous point at issue between Jeremiah and his
opponents was political rather than ecclesiastical. Jeremiah was
anxious that Zedekiah should keep faith with Nebuchadnezzar, and
not involve Judah in useless misery by another hopeless revolt. The
prophets preached the popular doctrine of an imminent Divine
intervention to deliver Judah from her oppressors. They devoted
themselves to the easy task of fanning patriotic enthusiasm, till the
Jews were ready for any enterprise, however reckless.
During the opening years of the new reign, Nebuchadnezzar's recent
capture of Jerusalem and the consequent wholesale deportation
were fresh in men's minds; fear of the Chaldeans together with the
influence of Jeremiah kept the government from any overt act of
rebellion. According to li. 59, the king even paid a visit to Babylon, to
do homage to his suzerain.
It was probably in the fourth year of his reign[121] that the tributary
Syrian states began to prepare for a united revolt against Babylon.
The Assyrian and Chaldean annals constantly mention such
combinations, which were formed and broken up and reformed with
as much ease and variety as patterns in a kaleidoscope. On the
present occasion the kings of Edom, Moab, Ammon, Tyre, and Zidon
sent their ambassadors to Jerusalem to arrange with Zedekiah for
concerted action. But there were more important persons to deal
with in that city than Zedekiah. Doubtless the princes of Judah
welcomed the opportunity for a new revolt. But before the
negotiations were very far advanced, Jeremiah heard what was
going on. By Divine command, he made "bands and bars," i.e.
yokes, for himself and for the ambassadors of the allies, or possibly
for them to carry home to their masters. They received their answer,
not from Zedekiah, but from the true King of Israel, Jehovah Himself.
They had come to solicit armed assistance to deliver them from
Babylon; they were sent back with yokes to wear as a symbol of
their entire and helpless subjection to Nebuchadnezzar. This was the
word of Jehovah:—
"The nation and the kingdom that will not put its neck
beneath the yoke of the king of Babylon,
That nation will I visit with sword and famine and
pestilence until I consume them by his hand."
The allied kings had been encouraged to revolt by oracles similar to
those uttered by the Jewish prophets in the name of Jehovah; but:—
"As for you, hearken not to your prophets, diviners,
dreams, soothsayers and sorcerers,
When they speak unto you, saying, Ye shall not serve
the king of Babylon.
They prophesy a lie unto you, to remove you far from
your land;
That I should drive you out, and that you should
perish.
But the nation that shall bring their neck under the
yoke of the king of Babylon, and serve him,
That nation will I maintain in their own land (it is the
utterance of Jehovah), and they shall till it and
dwell in it."
When he had sent his message to the foreign envoys, Jeremiah
addressed an almost identical admonition to his own king. He bids
him submit to the Chaldean yoke, under the same penalties for
disobedience—sword, pestilence, and famine for himself and his
people. He warns him also against delusive promises of the
prophets, especially in the matter of the sacred vessels.
The popular doctrine of the inviolable sanctity of the Temple had
sustained a severe shock when Nebuchadnezzar carried off the
sacred vessels to Babylon. It was inconceivable that Jehovah would
patiently submit to so gross an indignity. In ancient days the Ark had
plagued its Philistine captors till they were only too thankful to be rid
of it. Later on a graphic narrative in the Book of Daniel told with
what swift vengeance God punished Belshazzar for his profane use
of these very vessels. So now patriotic prophets were convinced that
the golden candlestick, the bowls and chargers of gold and silver,
would soon return in triumph, like the Ark of old; and their return
would be the symbol of the final deliverance of Judah from Babylon.
Naturally the priests above all others would welcome such a
prophecy, and would industriously disseminate it. But Jeremiah
"spake to the priests and all this people, saying, Thus saith Jehovah:
—
"Hearken not unto the words of your prophets, which
prophesy unto you,
Behold, the vessels of the house of Jehovah shall be
brought back from Babylon now speedily:
For they prophesy a lie unto you."
How could Jehovah grant triumphant deliverance to a carnally
minded people who would not understand His Revelation, and did
not discern any essential difference between Him and Moloch and
Baal?
"Hearken not unto them; serve the king of Babylon
and live.
Why should this city become a desolation?"
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com

Water Resources Management And Modeling Purna Nayak

  • 1.
    Water Resources ManagementAnd Modeling Purna Nayak download https://ebookbell.com/product/water-resources-management-and- modeling-purna-nayak-4109700 Explore and download more ebooks at ebookbell.com
  • 2.
    Here are somerecommended products that we believe you will be interested in. You can click the link to download. Geospatial Information Handbook For Water Resources And Watershed Management Volume 2 Methods And Modelling John G Lyon https://ebookbell.com/product/geospatial-information-handbook-for- water-resources-and-watershed-management-volume-2-methods-and- modelling-john-g-lyon-46538182 Water Resources Management And Sustainability Solutions For Arid Regions Mohsen Sherif https://ebookbell.com/product/water-resources-management-and- sustainability-solutions-for-arid-regions-mohsen-sherif-49435410 Public Participation In The Governance Of International Freshwater Resources Water Resources Management And Policy Carl Bruch https://ebookbell.com/product/public-participation-in-the-governance- of-international-freshwater-resources-water-resources-management-and- policy-carl-bruch-2181242 Water Resources Management Innovative And Green Solutions De Gruyter Stem 2nd Edition Brears https://ebookbell.com/product/water-resources-management-innovative- and-green-solutions-de-gruyter-stem-2nd-edition-brears-56327240
  • 3.
    Water Resources ManagementInnovative And Green Solutions Robert C Brears https://ebookbell.com/product/water-resources-management-innovative- and-green-solutions-robert-c-brears-50336436 Enhancing Participation And Governance In Water Resources Management Conventional Approaches And Information Technology Libor Jansky https://ebookbell.com/product/enhancing-participation-and-governance- in-water-resources-management-conventional-approaches-and-information- technology-libor-jansky-1384502 Delta Waters Research To Support Integrated Water And Environmental Management In The Lower Mississippi River 1st Edition National Research Council Division On Earth And Life Studies Water Science And Technology Board Committee On Strategic Research For Integrated Water Resources Management https://ebookbell.com/product/delta-waters-research-to-support- integrated-water-and-environmental-management-in-the-lower- mississippi-river-1st-edition-national-research-council-division-on- earth-and-life-studies-water-science-and-technology-board-committee- on-strategic-research-for-integrated-water-resources- management-51874184 Water Resources Planning And Management R Quentin Grafton Karen Hussey https://ebookbell.com/product/water-resources-planning-and-management- r-quentin-grafton-karen-hussey-2324186 Water Resources Planning And Management Draft R Quentin Grafton Karen Hussey Eds https://ebookbell.com/product/water-resources-planning-and-management- draft-r-quentin-grafton-karen-hussey-eds-4157192
  • 5.
  • 6.
    Water Resources Managementand Modeling Edited by Purna Nayak Published by InTech Janeza Trdine 9, 51000 Rijeka, Croatia Copyright © 2012 InTech All chapters are Open Access distributed under the Creative Commons Attribution 3.0 license, which allows users to download, copy and build upon published articles even for commercial purposes, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications. After this work has been published by InTech, authors have the right to republish it, in whole or part, in any publication of which they are the author, and to make other personal use of the work. Any republication, referencing or personal use of the work must explicitly identify the original source. As for readers, this license allows users to download, copy and build upon published chapters even for commercial purposes, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications. Notice Statements and opinions expressed in the chapters are these of the individual contributors and not necessarily those of the editors or publisher. No responsibility is accepted for the accuracy of information contained in the published chapters. The publisher assumes no responsibility for any damage or injury to persons or property arising out of the use of any materials, instructions, methods or ideas contained in the book. Publishing Process Manager Marina Jozipovic Technical Editor Teodora Smiljanic Cover Designer InTech Design Team First published March, 2012 Printed in Croatia A free online edition of this book is available at www.intechopen.com Additional hard copies can be obtained from [email protected] Water Resources Management and Modeling, Edited by Purna Nayak p. cm. ISBN 978-953-51-0246-5
  • 9.
    Contents Preface IX Part 1Surface Water Modeling 1 Chapter 1 Tools for Watershed Planning – Development of a Statewide Source Water Protection System (SWPS) 3 Michael P. Strager Chapter 2 Strengths, Weaknesses, Opportunities and Threats of Catchment Modelling with Soil and Water Assessment Tool (SWAT) Model 39 Matjaž Glavan and Marina Pintar Chapter 3 Modelling in the Semi Arid Volta Basin of West Africa 65 Raymond Abudu Kasei Chapter 4 Consequences of Land Use Changes on Hydrological Functioning 87 Luc Descroix and Okechukwu Amogu Chapter 5 Fuzzy Nonlinear Function Approximation (FNLLA) Model for River Flow Forecasting 109 P.C. Nayak, K.P. Sudheer and S.K. Jain Chapter 6 San Quintin Lagoon Hydrodynamics Case Study 127 Oscar Delgado-González, Fernando Marván-Gargollo, Adán Mejía-Trejo and Eduardo Gil-Silva Chapter 7 Unsteady 1D Flow Model of Natural Rivers with Vegetated Floodplain – An Application to Analysis of Influence of Land Use on Flood Wave Propagation in the Lower Biebrza Basin 145 Dorota Miroslaw-Swiatek
  • 10.
    VI Contents Chapter 8Hydrology and Methylmercury Availability in Coastal Plain Streams 169 Paul Bradley and Celeste Journey Chapter 9 Contribution of GRACE Satellite Gravimetry in Global and Regional Hydrology, and in Ice Sheets Mass Balance 191 Frappart Frédéric and Ramillien Guillaume Part 2 Groundwater Modeling 215 Chapter 10 Simplified Conceptual Structures and Analytical Solutions for Groundwater Discharge Using Reservoir Equations 217 Alon Rimmer and Andreas Hartmann Chapter 11 Integration of Groundwater Flow Modeling and GIS 239 Arshad Ashraf and Zulfiqar Ahmad Chapter 12 Percolation Approach in Underground Reservoir Modeling 263 Mohsen Masihi and Peter R. King Chapter 13 Quantity and Quality Modeling of Groundwater by Conjugation of ANN and Co-Kriging Approaches 287 Vahid Nourani and Reza Goli Ejlali
  • 13.
    Preface Water Resources Managementintends to optimize the available water resources, which consists of the optimal utilization of surface water and groundwater, to satisfy the requirements of domestic, agricultural and industrial needs. In some parts of the world, there is abundance of water, while in other parts of the world the resources are scanty particularly in developing and under-developed countries. Floods and droughts continue to threaten and affect the livelihoods of most of the population in these countries. Therefore, there is an urgent need for optimal utilization of water resources. With ever increasing population, particularly in developing and under- developed countries, there is an urgent need to cater the needs of the population, where water is the basic requirement. Groundwater and surface water play a pivotal role in agriculture, and an increasing portion of extracted groundwater is used for irrigating agriculture fields. It is estimated that at least 40% of the world's food grains are produced using groundwater, by irrigated farming, both in countries with low GDP as well as in high GDP countries. In arid and semi-arid areas, the dependency on groundwater for water supply is much higher in comparison to other areas. This book is designed to address some of the real issues concerning water resource management, with some illustrative and good case studies relevant to the topic and up-to-date. We hope that the chapters in the book will be of great use to postgraduate and research scholars, providing them with current research trends and applications of water resources for better management. This book consists of two sections: surface water and groundwater. Surface water section covers watershed planning, impact of climate change on Volta Basin, rainfall-runoff and sediment modeling using SWAT model, flood forecasting using fuzzy logic approach, effect of land use changes on hydrology, hydrodynamics and unsteady flow modeling, water quality modeling, information on GRACE satellite and on wetland hydrology. Analytical solutions to groundwater discharge are discussed in groundwater section followed by groundwater flow modeling using MODFLOW, percolation approach, quality and quantity modeling using ANN and Kriging approach. Different modeling approaches are described followed by examples of case studies. The materials presented in this book should help a wide range of readers to apply different simulation techniques to resolve real life problems and issues concerned with water resource management.
  • 14.
    X Preface I amhighly grateful to Intech, Open Access Publisher, for giving me the opportunity to contribute as the Book Editor to this valuable book. In Particular, I would like to thank Ms. Marina Jozipovic, the Publishing Process Manager, for her constant support and cooperation during the preparation of this book. I also acknowledge the support of my colleague Mr. D. Mohana Rangan for his assistance in reviewing the chapters and helping in improving the quality of the contents. Purna Nayak National Institute of Hydrology, Kakinada, AP, India
  • 17.
  • 19.
    1 Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) Michael P. Strager Division of Resource Management, West Virginia University, USA 1. Introduction A Surface Water Protection System (SWPS) was developed to bring spatial data and surface water modeling to the desktop of West Virginia Bureau of Public Health (WVBPH), Office of Environmental Health Services (OEHS), Environmental Engineering Division (EED). The SWPS integrates spatial data and associated information with the overall goal of helping to protect public drinking water supply systems. The SWPS is a specialized GIS project interface, incorporating relevant data layers with customized Geographic Information Systems (GIS) functions. Data layers have been assembled for the entire state of West Virginia. Capabilities of the system include map display and query, zone of critical concern delineation, stream flow modeling, coordinate conversion, water quality modeling, and susceptibility ranking. The system was designed to help meet the goals of the Surface Water Assessment and Protection (SWAP) Program. The goal of the SWAP program is to assess, preserve, and protect West Virginia’s source waters that supply water for the state’s public drinking water supply systems. Additionally, the program seeks to provide for long term availability of abundant, safe water in sufficient quality for present and future citizens of West Virginia. The SWPS was designed to help meet this goal by addressing the three major components of the SWAP program: delineating the source water protection area for surface and groundwater intakes, cataloging all potential contamination sources, and determining the public drinking water supply system’s susceptibility to contamination. This chapter outlines the functions and capabilities of the SWPS and discusses how it addresses the needs of the SWAP program. The following sections discuss the application components. The components consist of: 1. A customized interface for study area selection 2. Integration of the EPA WHAEM and MODFLOW models 3. Delineation of groundwater public supply systems 4. Watershed delineation and zone of critical concern delineation for surface water sites 5. Stream flow model from multivariate regression 6. The environmental database
  • 20.
    Water Resources Managementand Modeling 4 7. UTM latitude/longitude conversion utility 8. Statewide map/GIS data layers 9. Water quality modeling capability 10. Groundwater and surface water susceptibility model Component 1. A customized interface for study area selection Using customized programming we were able to create a GIS interface to allow users to quickly find locations or define study areas for further analysis in the state. The locations may be selected in three ways: by geographical extent (e.g. county, watershed, 1:24,000 quad map, major river basin), by area name or code (e.g. abandoned mine land problem area description number, stream or river name, WV Division of Natural Resource (WVDNR) stream code, public water identification number or name), or by typing in the latitude and longitude coordinates. Once the study area is defined, the system zooms automatically to the extent of the selected feature and all available spatial data layers are then displayed. A discussion of the spatial data layers included is discussed in Component 8 of this document. Component 2. Integrating EPA WHAEM and MODFLOW models The SWPS application has the ability to read output from either EPA WHAEM or MODFLOW models. It does this by importing dxf file formats directly into SWPS from a pull down menu choice. Data can also be converted to shapefile format from SWPS to be read directly into WHAEM and MODFLOW. The data being read into SWPS needs to be in the UTM zone 17, NAD27 projection (with map units meters) for the new data to overlay on the current data existing within SWPS. Consequently, any data exported from SWPS will automatically be in the UTM zone 17 NAD27 coordinate system. Component 3. Delineation of groundwater public supply systems A fixed radius buffer zone was created around each groundwater supply site based on the pumping rate. If the pumping rate was less than or equal to 2,500 gpd, a radius of 500 feet was used. If the pumping rate was greater than 2,500 gpd but less than or equal to 5,000 gpd, a radius of 750 feet was used. If the pumping rate was greater than 5,000 gpd and less than or equal to 10,000 gpd, a radius of 1,000 feet was used. If the pumping rate was greater than 10,000 gpd and less than or equal to 25,000 gpd, then a radius of 1,500 feet was used. There were two exceptions to this fixed radius buffer procedure. The first was for any groundwater site less than or equal to 25,000 gpd that was in a Karst or mine area. These locations regardless of their pumping rate less than 25,000 gpd were buffered 2,000 feet. The second exception was for sites over 25,000 gpd. For these sites, hydro geologic and/or analytical mapping delineations will be done by personnel at the Bureau of Public Health. These were only identified in SWPS as being a well location and are left to more sophisticated groundwater modeling software. To perform buffers automatically, the user can use the GIS to create buffers dialog within SWPS susceptibility ranking menu option. The automatic fixed radius buffering requires knowledge about the pumping rate and fixed radius distance. This information is provided in a pulldown text information box within the susceptibility ranking menu option.
  • 21.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 5 Component 4. Watershed delineation and zone of critical concern delineation for surface water sites The ability to interactively delineate watersheds and zones of critical concern is built into SWPS. In this section, the watershed delineation tool is discussed, followed by the zone of critical concern delineation tool. Watershed Delineation SWPS allows the user to delineate a watershed for any mapped stream location in the state. The watershed is delineated based on the user-clicked point and it is added to the current view’s table of contents as a new theme or map layer labeled “Subwatershed.” The drainage area is reported back to the user as well. If only drainage area is requested, a separate tool allows for quick query of stream drainage area in acres and square miles, without waiting for the watershed boundary to be calculated. The watershed delineation is driven by a hydrologically correct digital elevation model (DEM). The DEM is corrected using stream centerlines for all 1:24,000 streams. The stream centerlines are converted to raster cells and DEM values are calculated for each cell. All off- stream DEM cells are raised by a value of 20 meters to assure the DEM stream locations are the lowest cells in the DEM. This step is necessary to assure of more accurate watershed delineations especially at the mouth of the watersheds. After the DEM is filled of all spurious sinks, flow direction and flow accumulation grids are calculated. These grids help determine the direction of flow and the accumulated area for each cell in the landscape. These grids were necessary for watershed delineation to occur and are important inputs for finding the zones of critical concern for surface water intakes. Surface Water Zones of Critical Concern Stream velocity is the driving factor for determining a five-hour upstream delineation for each surface water intake in WV. Only with stream velocity calculated was it possible to include factors such as high bank-full flow, average flow, stream slope, and drainage area all at once. The velocity equation used in this study came from a report titled “Prediction of Travel Time and Longitudinal Dispersion in Rivers and Streams” (US Geological Survey, Water-Resources Investigations Report 96-4013, 1996). In this report, data were analyzed for over 980 subreaches or about 90 different rivers in the United States representing a wide range of river sizes, slopes, and geomorphic types. The authors found that four variables were available in sufficient quantities for a regression analysis. The variables included the drainage area (Da), the reach slope (S), the mean annual river discharge (Qa), and the discharge at the section at time of the measurement (Q). The report defines peak velocity as: V’ p = VpDa/Q The dimensionless drainage area as: D’ a = Da 1.25 * sqrt(g) / Qa Where g is the acceleration of gravity. The dimensionless relative discharge is defined as: Q’a = Q/Qa
  • 22.
    Water Resources Managementand Modeling 6 The equations are homogeneous, so any consistent system of units can be used in the dimensionless groups. The regression equation that follows has a constant term that has specific units, meters per second. The most convenient set of units for use with the equation are: velocity in meters per second, discharge in cubic meters per second, drainage area in square meters, acceleration of gravity in m/s2, and slope in meters per meter. The equation derived in the report and the equation used in this study for peak velocity in meters per second was the following: Vp = 0.094 + 0.0143 * (D’ a)0.919 * (Q’ a)-0.469 * S 0.159 * Q/Da The standard error estimates of the constant and slope are 0.026 m/s and 0.0003, respectively. This prediction equation had an R2 of 0.70 and a RMS error of 0.157 m/s. Once a velocity grid was calculated as described above, it was used as an inverse weight grid in the flowlength ArcGIS (ESRI, 2010) command. The flowlength command calculates a stream length in meters. If velocity is in meters per second, the inverse velocity as a weight grid will return seconds in our output grid. This calculation of seconds would track how long water takes to move from every cell in the state where a stream is located to where it leaves the state. The higher values will exist in the headwater sections of a watershed. By querying the grid, it is possible to add the appropriate travel time to the cell value and this will the time of travel for an intake. All cells above an intake by 18,000 seconds (5 hours) will be the locations in which water would take to reach the intake. To use this methodology, GIS data layers had to be calculated for drainage area, stream slope, annual average flow, and bank-full flow for all of WV. The sections below describe how each of these grids was created. Drainage area To obtain a drainage area calculation for every stream cell in the state required a hydrogically correct DEM. The process of creating a hydrologically correct DEM was covered in the watershed delineation component described earlier. Essentially, from the DEM the flow direction and flow accumulation values for each stream cell are derived. The output of the flow direction request is an integer grid whose values range from 1 to 255. The values for each direction from the center are: 32 64 128 16 X 1 8 4 2 For example, if the direction of steepest drop were to the left of the current processing cell, its flow direction would be coded as 16. If a cell is lower than its 8 neighbors, that cell is given the value of its lowest neighbor and flow is defined towards this cell (ESRI, 2010). The accumulated flow is based upon the number of cells flowing into each cell in the output grid. The current processing cell is not considered in this accumulation. Output cells with a high flow accumulation are areas of concentrated flow and may be used to identify stream channels. Output cells with a flow accumulation of zero are local topographic highs and may be used to identify ridges. The equation to calculate drainage area from a 20-meter cell sized flow accumulation grid was:
  • 23.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 7 (cell value of flow accumulation grid + 1) * 400 = drainage area in meters squared Stream slope Stream slope was calculated for each stream reach in the state. A stream reach is not necessarily an entire stream but only the section of a stream between junctions. The GIS command streamlink was first used to find all unique streams between stream intersections or junctions. For each of these reaches, the length was calculated from the flowlength GIS command. Having the original DEM allowed us to find the maximum and minimum values for each of the stream reaches. The difference in the maximum and minimum elevations for the stream reach divided by the total reach length gave us our stream reach slope in meters per meter. Annual average flow Annual average flow for each stream cell location was found based on a relationship between drainage area and gauged stream flow. For 88 gauging stations in WV, covering many different rainfall, geological, and elevation regions, we assembled a table of drainage area for the gauges versus the historic annual stream flow for the gauge. After fitting a linear regression line for this data set, we found the following equation for annual stream flow setting the y intercept to zero. Annual stream flow in cfs = 2.05 * drainage area in square miles This equation had a corrected R2 of .9729. The XY plot and equation are shown in Figure 1. Fig. 1. Annual stream flow from gauged stations and drainage area at the gauges Since drainage area is already calculated for each stream cell location, this equation incorporated the drainage area grid to compute a separate grid layer of annual stream flow. This would be another input for the velocity calculation. Bank-full flow The last input for the velocity equation was the bank-full flow measure. Just as with annual average flow, this required a modeled value for every raster stream cell in WV. Using the drainage area vs. mean flow y = 2.05x R2 = 0.9729 -500 0 500 1000 1500 2000 2500 3000 3500 0.00 200.00 400.00 600.00 800.00 1000.00 1200.00 1400.00 1600.00 drainage area sq miles mean flow in cfs
  • 24.
    Water Resources Managementand Modeling 8 same approach to regressing drainage area to gauged stream flow as performed to find an annual average flow equation, this equation was used to find bank-full flow. Bank-full flow as defined by the Bureau of Public Health, is 90% of the annual high flow. To find the 90% of high flow for each gauging station, all historic daily stream flow data was downloaded for each of the 88 gauging stations. This data was then sorted lowest to highest and then numbered lowest to highest after removing repeating values. The value of flow at the 90% of the data became the bank-full flow value for that gauge. These values were then regressed against drainage area at the gauge. The linear regression equation for bank-full stream flow setting the y intercept to zero is listed below. Bank-full stream flow in cfs = 4.357 * drainage area in square miles This equation had a corrected R2 of .9265. The XY plot and equation are shown in Figure 2. Fig. 2. Bank-full stream flow from gauged stations and drainage area at the gauges This equation could be applied to the drainage area grid to calculate the bank-full flow for any stream cell in the state. It was the final input needed in the velocity calculation. The interactive zone of critical concern ability of SWPS delineates the upstream contributing area for a surface water intake in the following way. First, the user locates the surface water intake and makes sure the intake is on the raster stream cell. A button on the interface then initiates the model. The model will query the time of travel value for the intake and then add 18,000 seconds (5 hours) to the queried value upper range. All cells which fit this range are identified and the stream order attribute retrieved for those cells. All cells that are on the main stem stream where the intake existed are buffered 1000 feet on each side of the stream. All tributaries to the main stem are buffered 500 feet on each side of the stream. Next, a watershed boundary for the location of the intake is delineated and used to clip any areas of the buffer that may extend beyond ridgelines. And lastly, the surface water intake is buffered 1000 feet and combined with the clipped buffer to include areas 1000 feet downstream of the intake. drainage area vs. 90% of high flow y = 4.357x R 2 = 0.9265 -1000 0 1000 2000 3000 4000 5000 6000 7000 0 500 1000 1500 2000 drainage area in sq miles 90% of high flow in cfs
  • 25.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 9 This interactive ability allows zones of critical concern to be delineated for any river or stream in WV. Only large rivers which border WV, such as the Ohio, Tug, and Potomac can not be interactively delineated using this method. This is due to unknown drainage areas for these bordering rivers and unknown tributaries to these major rivers coming from the bordering states. This is the major limitation of this modeling approach for WV at this time and in the next version of this watershed tool will account for all outer drainage influences. The Ohio River Sanitation Commission (ORSANCO) is responsible for delineating zones of critical concern for the intakes along the Ohio River. ORSANCO uses uniform 25-mile upstream distances for zones of critical concern for intakes along the Ohio River. This same approach could be applied to other rivers such as the Tug and Potomac in WV. For reservoirs and lakes within the watershed delineation area, a set of standards was set by the Bureau of Public Health and was used in this study. For a reservoir, a buffer of 1000 feet on each bank and 500 feet on each bank of the tributaries that drain into the lake or reservoir was used. When a lake or reservoir is encountered within the five-hour time of travel, a specific delineation was used. If the length of the lake/reservoir was less than or equal to the five hour calculated time of travel distance from the intake, then the entire water body was included. If the length of the lake/reservoir was greater than the calculated five hour time of travel distance from the intake, then the section of water body within the five hour time of travel distance was used to establish the zone of critical concern. Component 5. Stream flow model from multivariate regression Overview This project component for SWPS used multivariate techniques to evaluate stream flow estimation variables in West Virginia. The techniques included correlation analysis, multiple regression, cluster analysis, discriminant analysis and factor analysis. The major goal was to define watershed scale factors to estimate the stream flow at recorded USGS gauges. To do this, the contributing area upstream of each gauge was first delineated. Next, annual averages of precipitation and temperature and landscape based variables for the contributing upstream area were calculated and regressed against 30-year average annual flow at the USGS gauge. Results from the statistical analysis techniques found the most important variables to be upstream drainage area, 30-year annual maximum temperature, and stream slope. While this analysis was limited by the availability of data and assumptions to predict stream flow, the results indicate that stream flow can be modeled with reasonably good results. The following sections include a review of the literature on stream flow estimation techniques, a description of the variables used in this study to predict stream flow, the multivariate statistical methods, and a discussion of results and limitations of the study. Literature Review The intent of this literature review was to determine variables that were used to estimate stream flow in other studies, identify different statistical procedures, and to find limitations in this study based on other papers. The impact of land-use, climate change and groundwater abstraction on stream flow was examined by Qerner et al. (1997). They analyzed the effects of these factors using physical
  • 26.
    Water Resources Managementand Modeling 10 models BILAN, HBVOR, MODFLOW and MODGROW. The models were used to simulate the impact of afforstation, climate warming by 2 and 4 degrees Celsius in combination with an adoption of the precipitation changes in groundwater recharge and groundwater abstractions on stream flow droughts. The authors found that all the physical models can be used to assess the impacts of human activities on stream flow. They also concluded that based on some climate change scenarios they followed out, that the deficit volume of water is very sensitive to both an increase in temperature and a change in precipitation. Even in basins with abundant precipitation, the warming of 2 degrees Celsius would result in a rise in the deficit volume of water by 20 percent. Their findings also acknowledge the importance of using precipitation, temperature, groundwater recharge and groundwater abstractions along with water storage holding capacity of watersheds. Timofeyeva and Craig (1998) used Monte Carlo techniques to estimate month by month variability of temperature and precipitation for drainage basins delineated by a digital elevation model. They also used a runoff grid from the digital elevation model to estimate discharge at selected points and compared this to known gauge station data. The variance of temperature was modeled as the standard error of the regression from the canonical regression equation. For precipitation, they modeled the variance as the standard error of the prediction. This was done to achieve unbiased estimators. When comparing the climate and resulting runoff and stream flow estimators calculated by Monte Carlo estimation, to the observed flow, the simulated results were within the natural variability of the record (Timofeyeva and Craig, 1998). Long-range stream flow forecasting using nonparametric regression procedures was developed by Smith, (1991). The forecasting procedures, which were based solely on daily stream flow data, utilized nonparametric regression to relate a forecast variable to a covariate variable. The techniques were adopted to develop long-term forecasts of minimum daily flow of the Potomac River at Washington, D.C. Smith’s key finding was that to implement nonparametric regression requires the successful specification of “bandwidth parameters.” The bandwidth parameters are chosen to minimize the integrated mean square error of forecasts. Basically, his stream flow technique focussed on examining past history of stream flow and making nonparametric regression forecasts based on what is likely to occur in the future. No additional variables besides historic flow were used to model future conditions. Another nonparameteric approach to stream flow simulation was done by Sharma et al. (1997). They used kernal estimates of the joint and conditional probability density functions to generate synthetic stream flow sequences. Kernal density estimation includes a weighted moving average of the empirical frequency distribution of the data (Sharma, et al. 1997). The reason for this method is to estimate a multivariate density function. This is a nonparametric method for the synthesis of stream flow that is data driven and avoids prior assumptions as to the form of dependence (linear or non linear) and the form of the probability density function. The authors main finding was that the nonparametric method was more flexible for their study than the conventional models used in stochastic hydrology and is capable of reproducing both linear and nonlinear dependence. In addition, their results when applied to a river basin indicated that the nonparametric approach was a feasible alternative to parametric approaches used to model stream flow.
  • 27.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 11 Garren (1992) noted that although multiple regression has been used to predict seasonal stream flow volumes, typical practice has not realized the maximum accuracy obtainable from regression. The forecasting methods he mentions which can help provide superior forecasting include: (1) Using only data known at forecast time; (2) principal components regression; (3) cross validation; and (4) systematically searching for optimal or near-optimal combinations of variables. Some of the variables he used included snow water equivalent, monthly precipitation, and stream flow. The testing of selection sites for a stream flow forecasting study, he feels should be based on data quality, correlation analyses, conceptual appropriateness, professional judgement, and trial and error. The use of principal components regression provides the most satisfactory and statistically rigorous way to deal with intercorrelation of variables. He concluded that the maximum forecast accuracy gain is obtained by proper selection of variables followed by the use of principal components regression and using only known data (no future variables). The results of a multiple-input transfer function modeling for daily stream flow using nonlinear inputs was studied by Astatkie and Watt (1998). They argue that since the relationship between stream flow and its major inputs, precipitation and temperature, are nonlinear, the next best alternative is to use a multiple input transfer function model identification procedure. The transfer function model they use includes variables such as type of terrain, drainage area, watercourse, the rate of areal distribution of rainfall input, catchment retention, loss through evapotranspiration and infiltration into the groundwater, catchment storage, and melting snow. When comparing their modeling technique for stream flow to that of a nonlinear time series model, they found their transfer function model to be direct and relatively easy for modeling multiple inputs. They also found it more accurate in head to head tests against the nonlinear time series model. Since stream flow modeling is an outcome of many runoff estimation models, the literature for deriving runoff grids is applicable to stream flow studies. Anderson and Lepisto (1998) examined the links between runoff generation, climate, and nitrate leaching from forested catchments. One of the things they sought out to prove in their study was that climate will influence the amount of nitrate that can be leached from the soil and the water flow that will transport it to the streams. They found that a negative correlation existed between stream flow and temperature. Significant positive correlation between modeled surface runoff and concentrations of nitrate was found when they considered periods of flow increases during cold periods. Their study identified the importance for identifying and calculating the surface runoff fraction, daily dynamics of soil moisture, groundwater levels, and extensions of saturated areas when doing a contaminant transport or flow estimation study. In another study, Moore (1997) sought to provide an alternative to the matching strip, correlation, and parameter-averaging methods for deriving master recession characteristics from a set of recession segments. The author then choose to apply the method to stream flow recession segments for a small forested catchment in which baseflow is provided by drainage of the saturated zone in the shallow permeable soil. The plots indicated the recessions were non-linear and that the recessions did not follow a common single valued storage outflow relation. The final decision was a model with two linear reservoirs that provided substantially better fit than three single reservoir models, indicating that the form of the recession curve probably depends on not just the volume of subsurface storage, but also on its initial distribution among reservoirs.
  • 28.
    Water Resources Managementand Modeling 12 Gabriele et al. (1997) developed a watershed specific model to quantify stream flow, suspended sediment, and metal transport. The model, which estimated stream flow, included the sum of three major components: quick storm flow, slow storm flow, and long- term base flow. Channel components were included to account for timing effects associated with waters, sediments, and metals coming from different areas. Because of relatively good results from the modeling process, the conceptualizations supported that the study area river was strongly influenced by three major components of flow: quick storm flow, slow storm flow, and long-term base flow. Therefore, sediment inputs can be associated with each of those stream flow components and assign metal pollution concentrations to each flow and sediment input. From this review of other studies, variables were determined that have been used successfully in stream flow estimation. Examining the limitations of other studies has also provided insight into data layers that may not be able to include. Of the statistical techniques used, the multivariate approach, in which components are added or subtracted to achieve the best fit possible, is a sound statistical procedure. In addition to this approach, testing the correlations between variables is another way of finding a model for estimating stream flow in WV. Methodology The first step in assembling data for this study was to delineate the total upstream contributing area for each of thirteen USGS gauge stations in West Virginia. Figure 3 displays the location of each gauge and the defined upstream drainage area for that gauge. Fig. 3.
  • 29.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 13 For every drainage area, the following criteria were calculated; total area, 30 year average annual precipitation, 30 year average annual maximum temperature, 30 year average annual minimum temperature, average drainage area slope, and stream slope. These variables were explanatory variables, which would be regressed against the dependent variable, the 30year average annual flow recorded at the gauge stations. The figures 4 to 7 show the distribution of 30-year precipitation, maximum temperature, minimum temperature, and elevation across the different areas. By using GIS techniques, it was possible to find the average value in the drainage areas along with drainage area slope and stream slope for each of the variables. The data for each gauge area and assembled variables is summarized in table 1. id# USGS Gauge name Upstream drainage area 30yr annual 30yr annual 30yr annual 30yr annual stream elevation drop Watershed Slope average (acres) precip ave (inches) ave temp max(F) ave temp min(F) Stream flow (cfs) max-min in (meters) (degree) g1 1595200 31296 52 54 35 99.68 418 5 g5 3050000 120352 50 57 37 379.37 607 15 g7 3053500 176708 46 60 38 613.56 643 11 g10 3061000 484507 43 62 39 1158.14 26 13 g11 3061500 74501 42 61 39 168.99 130 13 g12 3062400 7146 43 60 37 16.54 189 9 g13 3066000 55068 53 54 36 210.40 289 6 g17 3114500 289609 42 62 40 665.40 57 15 g19 3180500 85166 53 55 36 273.49 435 14 g21 3189100 338131 53 57 37 1445.61 743 13 g22 3190400 232990 50 59 38 750.36 830 11 g24 3195500 346231 48 59 38 1176.66 933 17 g26 3202400 196645 47 63 39 421.78 583 18 Table 1. Data used in study The first step in analyzing the data in table 1 was to perform some basic statistics. The values across the different gauging station locations were investigated. The summarized statistical data is shown in table 2. Variable N Mean Median TrMean StDev SE Mean Minimum Maximum Q1 Q3 area 13 187565 176708 176972 144629 40113 7146 484507 64784 313870 precip 13 47.85 48 47.91 4.34 1.2 42 53 43 52.5 maxtemp 13 58.692 59 58.727 3.066 0.85 54 63 56 61.5 mintemp 13 37.615 38 37.636 1.446 0.401 35 40 36.5 39 flow 13 568 422 538 455 126 17 1446 190 954 strmslop 13 452.5 435 447.6 299.3 83 26 933 159.5 693 wsslope 13 12.31 13 12.45 3.88 1.08 5 18 10 15 Table 2. Basic statistics
  • 30.
    Water Resources Managementand Modeling 14 Fig. 4. Fig. 5.
  • 31.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 15 Fig. 6. Fig. 7.
  • 32.
    Water Resources Managementand Modeling 16 From table 2, it was noted which variables were closely grouped and which varied significantly among all the 13 different gauges. The area and flow variables have the highest standard deviation while the precipitation, maximum and minimum temperatures, and watershed slope have the lowest standard deviation. Other simple statistical graphs, which were used to gain insights into the data distribution and spreads, are shown in figures 8 to 14. The figures provided a graphical display of the distribution of values across the 13 gauges. Data exploration is important to determine trends and outliers in data that may bias results (Johnson, et al 2001). In addition, regression results may be impacted from large variations in data values. A common technique is to normalize data with a simple equation such as the value of interest minus the minimum value for that variable divided by the maximum minus minimum within the data range (Kachigan, 1986). However, in this study the values were not normalized due to the spatial nature of the information source. It was necessary to identify and incorporate the spatial variability across the entire study area at the statewide level. The end use of our regression relationship is the ability to query any raster stream cell and report all the unique information from the spatial analysis. Stream flow and water quality decisions for permitting may occur in high elevation cold headwater segments as well as large river systems with much accumulated drainage. Because the study area had unique topographic features that were to be regressed against representative stream flow information, the gauge driven delineated watersheds were chosen to represent this differentiation as best as possible as shown in Figure 3. Fig. 8.
  • 33.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 17 Fig. 9. Fig. 10.
  • 34.
    Water Resources Managementand Modeling 18 Fig. 11. Fig. 12.
  • 35.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 19 Fig. 13. Fig. 14.
  • 36.
    Water Resources Managementand Modeling 20 The next step in analyzing the data was to use generate a best-fit line plot for each of the independent variables in table 1 regressed against the dependent variable stream flow. These plots are shown in figures 15 to 20. From these best-fit line plots, the area, stream slope, and watershed slope variables had the best R squared values and positive linear relationship. The maximum and minimum temperature variables along with precipitation had the worst linear fit with stream flow. Their R squared values were very low with the precipitation variable looking very random in describing stream flow. At this point in the analysis it appeared that the area, stream slope and watershed slope will be the better variables to predict stream flow. While the linear regression plots provided some idea of the extent of the relationship between two variables, the correlation coefficient gives a summary measure that communicates the extent of correlation between two variables in a single number (Kachigan, 1986). The higher the correlation coefficient, the more closely grouped are the data points representing each objects score on the respective variables. Some important assumptions of the correlation coefficient are that the data line in groupings that are linear in form. The other important assumptions include that the variables are random and measured on either an interval or a ratio scale. In addition, the last assumption for the use of the correlation coefficient is that the two variables have a bivariate normal distribution. The correlation matrix for the data used in this study is shown in table 3. area precip maxtemp mintemp flow strslope wsslope area 1 -0.212 0.470 0.571 0.922 0.138 0.516 precip -0.212 1 -0.850 -0.781 0.039 0.560 -0.279 maxtemp 0.470 -0.850 1 0.930 0.245 -0.226 0.590 mintemp 0.571 -0.781 0.930 1 0.356 -0.217 0.647 flow 0.922 0.039 0.245 0.356 1 0.392 0.435 strslope 0.138 0.560 -0.226 -0.217 0.392 1 0.245 wsslope 0.516 -0.279 0.590 0.647 0.435 0.245 1 Table 3. Correlation matrix The variables with significant correlations (R > .7) are shaded in table 3. The variables listed in order of highest correlation to lowest significance are mintemp and maxtemp, flow and area, precip and maxtemp, and precip and mintemp. The correlations between the weather data were expected. In areas of higher precipitation, the temperature will be cooler (the annual averages for maximum temperature will be lower and the annual average for minimum temperatures will be lower) hence the high negative correlation. The other high positive correlated variables indicate that the variation in one variable will lead to variation in the other variable. For regression analysis the variables should be independent. Collinearity refers to linear relationships within the variables. The amount of multicollinearity across variables can be examined with principal component analysis of a sample correlation matrix (Sundberg, 2002) among other methods to remove dependence. This study examined the smallest eigenvalue and eliminated variables with values less than 0.05 as an indication of substantial collinearity (Hocking, 1996). As expected the precipitation variables were not independent to the elevation data and therefore removed.
  • 37.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 21 Fig. 15. Fig. 16.
  • 38.
    Water Resources Managementand Modeling 22 Fig. 17. Fig. 18.
  • 39.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 23 Fig. 19. Fig. 20.
  • 40.
    Water Resources Managementand Modeling 24 Performing regression analysis on the data was the next step in formulating a relationship and model to predict and estimate stream flow. Using the technique by Garren (1992) a regression equation with all the remaining variables was created, evaluate the P values of each variable, and eliminate variables until the highest adjusted R square is found. The first run with the regression analysis indicates that the variables area, strmslop and maxtemp will have the most influence on flow because of their low P values. Table 4 shows the regression analysis including all the variables. The regression equation is flow = 2325 + 0.00310 area - 12.0 precip - 37.3 maxtemp + 8 mintemp + 0.423 strmslope - 4.8 wsslope Predictor Coef StDev T P Constant 2325 4370 0.53 0.614 area 0.0030987 0.0004281 7.24 0.000 precip -12.02 30.84 -0.39 0.710 maxtemp -37.31 53.50 -0.70 0.512 mintemp 7.8 100.7 0.08 0.941 strmslop 0.4235 0.2281 1.86 0.113 wsslope -4.83 18.54 -0.26 0.803 S = 156.3 R-Sq = 94.1% R-Sq(adj) = 88.2% Table 4. Regression analysis including all variables By systematically removing the variables with a high P value and noting the R squared adjusted value, it was possible to arrive at a final set of variables to use in a regression equation to estimate stream flow. Table 5 shows the regression analysis results after removing the variable with the highest P value (mintemp). The regression equation is flow = 2492 + 0.00311 area - 12.3 precip - 35.0 maxtemp + 0.421 strmslope - 4.3 wsslope Predictor Coef StDev T P Constant 2492 3522 0.71 0.502 area 0.0031121 0.0003628 8.58 0.000 precip -12.35 28.30 -0.44 0.676 maxtemp -35.00 41.23 -0.85 0.424 strmslop 0.4208 0.2089 2.01 0.084 wsslope -4.33 16.11 -0.27 0.796 S = 144.7 R-Sq = 94.1% R-Sq(adj) = 89.9% Table 5. Regression analysis with mintemp removed The R squared adjusted improved slightly to 89.9% with mintemp removed. This process of removing the current highest P value variable and re-running of the model was repeated six times. The associated R squared values were noted and table 6 was created from the results.
  • 41.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 25 As table 6 indicates, the combination of variables that provided the highest R squared adjusted value were area, maxtemp, and strslope. The associated regression equation with the optimal set of variables is: flow = 1232 + 0.00304 area - 23.6 maxtemp + 0.338 strmslope Variables included in the regression R squared adjusted Area, mintemp, maxtemp, strslope, wsslope 88.2 Area, maxtemp, strslope, wsslope 89.9 Area, maxtemp, strslope 91.1 Area, maxtemp, strslope 91.8 Area, strslope 90.5 Area 83.6 Table 6. Multiple regression results The next procedure used in the analysis was discriminant analysis. This technique was used to identify relationships between qualitative criterion variables and the quantitative predictor variables in the dataset. The objective was to identify boundaries between the groups of watersheds that the gauges were associated. The boundaries between the groups are the characteristics that distinguish or discriminate the objects in the respective groups. Discriminant analysis allows the user to classify the given objects into groups – or equivalently, to assign them a qualitative label – based on information on various predictor or classification variables (Kachigan, 1992). The gauge station dataset was assigned a qualitative variable based on which major drainage basin in West Virginia the area was located. The major basins used were the Monongahela (m), Gauley (g) and Other (x). The class “other” was assigned to gauges that did not fall in the Monongahela or Gauley drainage basins. Running the discriminant analysis in Minitab produced the results shown in table 7. Only gauge one and gauge five were reclassified from the discriminant analysis results. It should be noted however that the discriminant function should be validated by testing its efficacy with a fresh sample of analytical objects. Kachigan (1986) notes that the observed accuracy of prediction on the sample upon which the function was developed will always be spuriously high, because we will have capitalized on chance relationships. The true discriminatory power of the function will be found when tested with a completely separate sample. By using discriminant analysis, it enabled the investigation of how the given groups differ. In the next analysis step, cluster analysis, the goal is to find whether a given group can be partitioned into subgroups that differ. The advantage of the approach is in providing a better feel of how the clusters are formed and which particular objects are most similar to one another. The cluster analysis was performed with distance measures of Pearson and Average and link methods of single and Euclidean. The Average and Euclidean choices worked the best in identifying clusters. Figure 21 shows the dendrogram results and table 8 lists the computation results.
  • 42.
    Water Resources Managementand Modeling 26 Linear Method for Response: class Predictors: area precip maxtemp mintemp flow strslope wsslope Group g m x Count 2 5 6 Summary of Classification Put into ....True Group.... Group g m x g 2 0 0 m 0 4 1 x 0 1 5 Total N 2 5 6 N Correct 2 4 5 Proportion 1.000 0.800 0.833 N = 13 N Correct = 11 Proportion Correct = 0.846 Squared Distance Between Groups g m x g 0.0000 14.5434 17.0393 m 14.5434 0.0000 4.5539 x 17.0393 4.5539 0.0000 Linear Discriminant Function for Group g m x Constant -7379.7 -7053.9 -7003.2 area -0.0 -0.0 -0.0 precip 85.6 83.5 83.6 maxtemp 86.5 84.2 84.3 mintemp 157.5 155.0 153.0 flow 0.8 0.7 0.7 strslope -0.3 -0.3 -0.3 wsslope -38.0 -37.0 -36.6 Summary of Misclassified Observations Observation True Pred Group Squared Probability Group Group Distance 1 ** x m g 20.956 0.000 m 5.163 0.578 x 5.796 0.421 2 ** m x g 23.906 0.000 m 5.223 0.229 x 2.790 0.771 gauge id majshed class FITS1 g1 NorthBranch x m g5 Tygart m x g7 Tygart m m g10 WestFork x x g11 MonRiver m m g12 MonRiver m m g13 Cheat m m g17 MiddleOhio x x g19 Greenbrier x x g21 Gauley g g g22 Gauley g g g24 Elk x x g26 UpGuyandotte x x Table 7. Discriminant analysis
  • 43.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 27 Standardized Variables, Euclidean Distance, Average Linkage Amalgamation Steps Step Number of Similarity Distance Clusters New Number of obs. clusters level level joined cluster in new cluster 1 12 85.24 0.933 1 7 1 2 2 11 80.06 1.261 3 11 3 2 3 10 78.24 1.376 2 9 2 2 4 9 70.39 1.872 5 6 5 2 5 8 69.18 1.948 4 8 4 2 6 7 68.16 2.013 10 12 10 2 7 6 61.42 2.439 3 10 3 4 8 5 56.79 2.732 1 2 1 4 9 4 54.13 2.900 3 13 3 5 10 3 45.06 3.473 4 5 4 4 11 2 41.04 3.727 3 4 3 9 12 1 35.49 4.078 1 3 1 13 Final Partition Number of clusters: 2 Number of Within cluster Average distance Maximum distance observations sum of squares from centroid from centroid Cluster1 4 8.340 1.414 1.816 Cluster2 9 46.098 2.188 2.934 Cluster Centroids Variable Cluster1 Cluster2 Grand centrd area -0.7923 0.3522 -0.0000 precip 0.9578 -0.4257 -0.0000 maxtemp -1.2045 0.5353 -0.0000 mintemp -1.1175 0.4966 0.0000 flow -0.7181 0.3191 -0.0000 strslope -0.0511 0.0227 -0.0000 wsslope -0.5946 0.2643 -0.0000 Distances Between Cluster Centroids Cluster1 Cluster2 Cluster1 0.0000 3.2672 Cluster2 3.2672 0.0000 Table 8. Hierarchical cluster analysis of observations From the clustered results, gauges 1 and 7 (g1 and g13) are the most alike and merge into a cluster at around 85 on the similarity scale. Gauges 3 and 11 (g7 and g22) are the next most similar at the 78 level. However, these objects do not form the same cluster until a lower level of similarity around the 35 level. By clustering the objects, we were able to identify groups that are alike and because of the small dataset, it was easy to examine the data table and discover values that make the objects similar. After cluster analysis, the choice was made to perform a factor analysis as an aid in data reduction. Although there were only seven variables, the possibility existed to gain insight into removing the duplicated information from among the set of variables. The results were assembled as a loading plot – figure 22, a score plot – figure 23, and a scree plot – figure 24. The output session data is listed in table 9.
  • 44.
    Water Resources Managementand Modeling 28 Fig. 21. Fig. 22.
  • 45.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 29 Fig. 23. Fig. 24.
  • 46.
    Water Resources Managementand Modeling 30 Principal Component Factor Analysis of the Correlation Matrix Unrotated Factor Loadings and Communalities Variable Factor1 Factor2 Communality area -0.748 -0.512 0.822 precip 0.720 -0.639 0.927 maxtemp -0.913 0.291 0.919 mintemp -0.949 0.192 0.937 flow -0.559 -0.737 0.855 strslope 0.117 -0.821 0.687 wsslope -0.731 -0.286 0.616 Variance 3.6732 2.0898 5.7630 % Var 0.525 0.299 0.823 Rotated Factor Loadings and Communalities Varimax Rotation Variable Factor1 Factor2 Communality area 0.243 0.874 0.822 precip -0.962 0.025 0.927 maxtemp 0.886 0.365 0.919 mintemp 0.849 0.465 0.937 flow -0.047 0.924 0.855 strslope -0.618 0.552 0.687 wsslope 0.375 0.689 0.616 Variance 3.0164 2.7466 5.7630 % Var 0.431 0.392 0.823 Factor Score Coefficients Variable Factor1 Factor2 area -0.002 0.319 precip -0.347 0.108 maxtemp 0.280 0.054 mintemp 0.257 0.096 flow -0.111 0.368 strslope -0.277 0.280 wsslope 0.064 0.233 Table 9. Factor analysis From these results, the variables high in loadings on a particular factor would be those which are highly correlated with one another, but which have little or no correlation with the variables loading highly on the other factors. The negative loading variable has a meaning opposite to that of the factor. The size of the loading is an indication of the extent to which the variable correlates with the factor.
  • 47.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 31 Limitations, and Discussion of Results The limitations with this study can be attributed to the number of gauges used and the variables used to predict stream flow. With more complete data over the state, it would have been able to assemble more gauges for this component of the project. Also, if possible it would have been good to include variables used to describe interception, evapotranspiration, infiltration, interflow, saturated overland flow, and baseflow from groundwater. The rate and areal distribution of rainfall input would have been helpful in establishing the catchment retention. Other issues with the data collection make the estimation of stream flow difficult. First, there is very high variability in recording stream flow data. The stream flow variable exhibited the highest standard deviation and variation across the year. Second, taking yearly annual averages was a crude method in which to characterize the varying conditions that occur across seasons, months, weeks, and even days. Third, the precipitation and temperature data used in the study needed to be better allocated to the gauge drainage areas (as compared to using the drainage area average for the variable) because of the amount of variability that is present in the entire watershed for the precipitation and temperature data. Overall, the choice of variables to analyze were appropriate based on the success other studies found. In the study the results of the multivariate regression indicated that stream flow could best be estimated using area, stream slope and 30 year annual average maximum temperature. Other data analysis techniques revealed the correlation present between the two temperature variables, flow and area, and precipitation to the two temperature variables. The last important summary from the tests came from the cluster analysis that grouped the gauge station objects based on similarity. The grouped gauges shared the same ecoregions. Ecoregions are defined as "regions of relative homogeneity in ecological systems or in relationships between organisms and their environments" (Omernik 1987). Omernik (1987) mapped the ecoregions of the conterminous United States, based on regional patterns in individual maps of land use, land surface form, potential natural vegetation, and soils. A discriminant analysis using the ecoregion of each gauge station catchment area would have been a better choice than the using the major river basins used in this study. The similar gauge station catchment areas identified by the cluster analysis and the associated ecoregion borders in West Virginia are displayed in figure 25. Component 6. The environmental database An environmental database of point data was included within SWPS. These points are found in the shapefiles directory of SWPS and are loaded for viewing when a user defines a study area location in the state. A brief listing of some of the files in the environmental database follows:  National pollution discharge elimination system sites  Landfills  Superfund sites (CERCLIS)  Hazardous and solid waste sites (RCRIS)  Toxic release inventory sites  Coal dams
  • 48.
    Water Resources Managementand Modeling 32  Abandoned mine land locations  Animal feed lots  Major highways  Railroads Fig. 25. Component 7. UTM latitude/longitude conversion utility This capability of SWPS allows the user to map coordinates in degrees, minutes and seconds by using an input dialog screen. The user’s points are then mapped in the UTM zone 17 projection. Points may be added to an existing point feature theme or a new point theme can be created. The ability to type coordinates and have the points reprojected saves the user many extra steps. In addition to mapping points from user input, a point can be queried for its x and y locations in UTM, stateplane, or latitude and longitude coordinates. The user can identify locations quickly by clicking anywhere in the display to report this information. Component 8. Statewide map/GIS data layers All GIS data is organized in the shapefiles and grids directories of SWPS. This data is listed below. These datasets are provided in addition to the data listed in the environmental database discussed in component 6.
  • 49.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 33  Coded hydrology  Roads (1:100K scale)  EPA MRLC land use/cover  Watersheds  Cities  Digital Raster Graphics  Lakes/impoundments  USGS gauging stations  SPOT imagery  Counties  Public wells  Digital Elevation Model  1:24K quads  Runoff grid  Hillshaded relief  Major river basins  Flow direction  Elevation TIN  Abandoned mine lands  Flow accumulation  303-D listed streams  Watersheds with fish collection data  Stream orders  Major Rivers  WV GAP land use/ cover  Bond forfeiture sites  Expected mean concentration grid  Public wells  Groundwater wells  Surface water wells  Cumulative runoff  Coal Geology  Override 7Q10 streams  Landfills  NW Wetlands  Springs over 500gpm  1950 land use/cover  Shreve stream orders  Stream length from mouth  DRAFT 14 digit HUCS  Wet weather streams  Surface mine inventory sites  Public wells  Strahler stream orders  Stream slope  Max and min stream elevations  Surface water zones of critical concern  Streamflow Component 9. Water quality modeling capability The water quality modeling capabilities of SWPS are built using a landscape driven approach that uses a predefined runoff and cumulative runoff grid to drive the analysis. It is essentially a weighted mass balance approach that will show changing concentrations and loadings based on changing flow conditions only. The runoff grid is based on a relationship between rainfall and stream flow. It is the main factor that directs flow directions to the stream or steepest path direction and estimates the stream flow. The assumptions/limitations of this water quality modeling approach are the following: 1. Streams have the same hydrogeometric properties (stream slope, roughness, width, and depth). 2. Also assumed are that the streams have the same ecological rate constants (reareation rates, pollution decay rates and sediment oxygen demand rate). 3. Transport of pollutants is considered to be conservative (values get averaged over changing flow conditions only) -> no loss or decay of pollutants is considered Does not consider infiltration, or ground water flow additions 5. Does not include atmospheric conditions such as evapotranspiration The water quality model in SWPS can be used in two different ways. The first is when the user has collected point locations of water quality data and wants to associate the sampled data to instream concentrations and loadings downstream of the sampling points. This is essentially a weighted mass balance approach using the stream flow and sampled locations to associate the point location information to stream condition. The input data using this method needs to be in Mg/L. The resultant modeled levels are reported back as stream values in Mg/L for concentration and Kg/Yr. for loading. The advantage of this first method of using sampled data is that it allows the user to see how the data location information can be used to estimate downstream conditions away from the sampling site.
  • 50.
    Water Resources Managementand Modeling 34 The second way the water quality model in SWPS can be used is in estimating total nitrogen, phosphorous and total suspended solids as concentrations and loadings in the stream based on expected mean concentrations from land use/cover classes. This method does not require any sampled water quality but uses the cover classes from a land use/cover grid (30meter-cell size). The thirteen classes for West Virginia from this data set were aggregated to six general classes because loading values for nitrogen, phosphorous and total suspended solids were only available for those six classes. The aggregated classes and the corresponding classes included: - Urban (low intensity developed, high intensity developed, residential) - Open/Brush (hay, pasture grass, mixed pasture, other grasses) - Agriculture (row crops) - Woodland (conifer forest, mixed forest, deciduous forest) - Barren (quarry areas, barren transitional areas) - Wetland (emergent and woody wetlands) The classes are associated with expected loadings based on the acreage size of the class. The loadings are annual averages and when used with the modeled stream flow can give concentration and loading results for the stream. The cover classes and associated expected mean concentrations levels used in the model are shown below. Total Nitrogen Total Phosphorous Total Suspended Solids Urban 1.89 0.009 166 Open/Brush 2.19 0.13 70 Agriculture 3.41 0.24 201 Woodland 0.79 0.006 39 Barren 3.90 0.10 2200 Wetland 0.79 0.006 39 The nutrient export coefficients above are multiplied by the amount (area) of a given land cover type. It is used as a simulation to estimate the probability of increased nutrient loads from land cover composition. It should be noted that there are factors other than land cover that contribute to nutrient export and these are rarely known with certainty. Some of the factors that may vary across watersheds and may change the expected mean concentration results include:  year to year changes in precipitation  soil type  slope and slope morphology (convex, concave)  geology  cropping practices  timing of fertilizer application relative to precipitation events  density of impervious surface The loading and concentration results in consideration of these assumptions however can still give insight in comparing expected pollutant values for watersheds. The results should be thought of in most cases as the worst case scenarios for stream water quality levels.
  • 51.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 35 Component 10. Groundwater and surface water susceptibility model The susceptibility ranking model within SWPS was constructed using the steps defined within the West Virginia Surface Water Assessment Program (SWAP) document. The susceptibility ranking for ground water systems was based on the physical integrity of the well and spring infrastructure; hydrologic setting; inventory of potential contaminant sources and land uses, and water quality. The susceptibility ranking for surface water systems was based on water quality and the inventory of potential contaminant sources. A more detailed explanation of the ground and surface water susceptibility models follows. Groundwater susceptibility model To determine the groundwater susceptibility for a site, the physical barrier effectiveness is first calculated. Physical barrier effectiveness is the Tier 1 assessment. It is used to note if there is a known impact on water quality, evaluate the source integrity as low or high, and to find the aquifer vulnerability. Based on these results, the physical barrier effectiveness can be determined as having high, moderate, or low potential susceptibility. If there is a known impact on water quality, then the model automatically goes to the Tier Two assessment and sets the groundwater susceptibility as being high. If there is no known impact on water quality then the source integrity and aquifer vulnerability set the physical barrier effectiveness for the Tier Two assessment. The aquifer vulnerability is determined from the different scenarios listed below; If Then All springs High Aquifer Sensitivity Alluvial Valleys Unconfined High Aquifer Sensitivity Confined Moderate Aquifer Sensitivity Appalachian Plateau Province (fracture) Moderate Aquifer Sensitivity Folded Plateau Area (fracture) Moderate Aquifer Sensitivity Karst Areas High Aquifer Sensitivity Valley and Ridge Province (fracture) Moderate Aquifer Sensitivity Coal Mine Areas High Sensitivity From the above scenarios, an aquifer vulnerability was determined. Using this with the source integrity rating can provide physical barrier effectiveness. Physical barrier effectiveness is  High if there is low source integrity and high aquifer sensitivity.  Low if there is high source integrity and moderate aquifer sensitivity  Moderate if there is high source integrity and high aquifer sensitivity or low source integrity and moderate aquifer sensitivity Again, if there is no known impact on water quality, this method will determine the physical barrier effectiveness as being high, moderate or low susceptibility. If there is a known water quality impact then the final groundwater susceptibility is high.
  • 52.
    Water Resources Managementand Modeling 36 Using the physical barrier effectiveness with the land use concern level determines the final groundwater susceptibility. The land use concern level is determined from the percentage of land use in the buffered groundwater site. The percentage of land use was found for every buffered location. In cases where the groundwater site had a pumping rate over 25,000 gpd, no buffer was created. For these groundwater sites no land use percentage was calculated. The final groundwater susceptibility is rated as:  High if the physical barrier effectiveness is high  High if the land use concern is high and the physical barrier effectiveness is moderate  Moderate if the land use concern is medium or low and the physical barrier effectiveness is moderate  Moderate if the land use concern is high or medium and the physical barrier effectiveness is low  Low if the land use concern is low and the physical barrier effectiveness is low The percentage of land use is reported to the user before the tier one assessment appears. He or she needs to know the associated concern levels with the percentage of land uses that are reported for each groundwater site. By not hard coding in the land use concern levels for each buffer, the user has the ability to perform “what if” type scenarios if existing land use changes or is different than what currently exists. Surface water susceptibility model The surface water susceptibility model is slightly less complicated than the groundwater model. For the surface water susceptibility determination, the percent of land use was calculated for each of the zone of critical concern. The land use concern level, and if there is a known water quality impact, are the two factors which are used to determine the surface water susceptibility. As in the groundwater model, the percent land use is presented to the user before the model is run so the user can make a determination and perform “what if” type scenarios with differing land use within the zone of critical concern. If there is a known water quality impact, then the surface water susceptibility is automatically high. If there is no known water quality impact, then the final surface water susceptibility is:  High if land use concern level is high  High if land use concern level is medium  Low if land use concern level is moderate Summary and Conclusion Drinking water is a critical resource that continues to need protection and management to assure safe supplies for the public. Since agencies to protect water resources operate at mostly state jurisdictions, it is important to implement a system at a statewide level. This chapter discussed watershed tools that integrate spatially explicit data and decision support to assist managers with both surface and ground water resources. It has three major components which include; an ability to delineate source water protection areas upstream of supply water, an inventory of potential contamination sources within various zones of critical concern, and the determination of the public drinking water supply systems
  • 53.
    Tools for WatershedPlanning – Development of a Statewide Source Water Protection System (SWPS) 37 susceptibility to contamination. The current system provides the ability to assess, preserve, and protect the states source waters for public drinking. 2. References Anderson, L. and A. Lepisto. 1998. “Links Between Runoff Generation, Climate and Nitrate- N Leaching From Forested Catchments.” Water, Air, and Soil Pollution. 105: pp227- 237. Astatkie, T. and W. E. Watt. 1998. “Multiple-Input Transfer Function Modeling of Daily Stream flow Series Using Nonlinear Inputs.” Water Resources Research. Vol. 34, No.10, pp2217-2725. Environmental Systems Research Institute (ESRI). 2010. ArcInfo ArcMap. Redlands, CA. Gabriele, H. M., and F. E. Perkins. 1997. “Watershed-Specific Model for Stream flow, Sediment, and Metal Transport.” Journal of Environmental Engineering. pp61-69. Garren, D. C. 1992. “Improved Techniques in Regression-Based Stream flow Volume Forecasting.” Journal of Water Resources Planning and Management. Vol. 118, No. 6. pp654-669. Hocking, R. R. 1996. Methods and Applications of Linear Models. Wiley and Sons, New York, NY. Johnston, K., J. M. Ver Hoef, K. Krivoruchko, and N. Lucas. 2001. Using ArcGis Geostatistical Analyst. Environmental System Research Institute, Redlands, CA. Kachigan, S. K. 1986. Statistical Analysis. Radius Press, New York, NY. Moore, R. D. 1997. “Storage-Outflow Modeling of Stream flow Recessions, with Application to Shallow-Soil Forested Catchments.” Journal of Hydrology. 198(1997) pp260-270. Omernik, J. M. 1987. “Ecoregions of the conterminous United States.” Annals of the Association of American Geographers Vol. 77, pp118-125. Querner, E. P., Tallaksen, L. M., Kasparek, L., and H. A. J. Van Lanen. 1997. “Impact of Land-Use, Climate Change and Groundwater Abstraction on Stream flow Droughts Using Physically-Based Models.” Regional Hydrology: Concepts and Models for Sustainable Water Resource Management. IAHS Publ. No. 246, pp171- 179. Sharma, A., D. G. Tarboton, and U. Lall. 1997. “Stream flow Simulation: A Nonparametric Approach.” Water Resources Research. Vol. 33, No. 2, pp291-308. Smith, J. A. 1991. “Long-Range Stream flow Forecasting Using Nonparametric Regression.” Water Resources Bulletin. Vol. 27, No. 1, pp39-46. Sundberg, R. 2002. Collinearity. Encyclopedia of Environmetrics, Vol. 1. John Wiley and Sons, Ltd, Chichester, West Sussex UK. Timofeyeva, M. and R. Craig. 1998. “Using Monte Carlo Technique for Modelling the Natural Variability of Stream flow in Headwaters of the Sierra Nevada, USA.” Hydrology, Water Resources and Ecology in Headwaters. 1998. No. 248, pp59- 65. U. S. Geological Survey. 1996. “Prediction of Travel Time and Longitudinal Dispersion in Rivers and Streams.” Water-Resources Investigations Report 96-4013.
  • 54.
    Water Resources Managementand Modeling 38 West Virginia Department of Health and Human Resources. August 1, 1999. “State of West Virginia Source Water Assessment and Protection Program.” Office of Environmental Health Services, Charleston, WV.
  • 55.
    Another Random ScribdDocument with Unrelated Content
  • 56.
    CHAPTER VIII BAD SHEPHERDSAND FALSE PROPHETS xxiii., xxiv. "Woe unto the shepherds that destroy and scatter the sheep of My pasture!"—Jer. xxiii. 1. "Of what avail is straw instead of grain?... Is not My word like fire, ... like a hammer that shattereth the rocks?"—Jer. xxiii. 28, 29. The captivity of Jehoiachin and the deportation of the flower of the people marked the opening of the last scene in the tragedy of Judah and of a new period in the ministry of Jeremiah. These events, together with the accession of Zedekiah as Nebuchadnezzar's nominee, very largely altered the state of affairs in Jerusalem. And yet the two main features of the situation were unchanged—the people and the government persistently disregarded Jeremiah's exhortations. "Neither Zedekiah, nor his servants, nor the people of the land, did hearken unto the words of Jehovah which He spake by the prophet Jeremiah."[101] They would not obey the will of Jehovah as to their life and worship, and they would not submit to Nebuchadnezzar. "Zedekiah ... did evil in the sight of Jehovah, according to all that Jehoiakim had done; ... and Zedekiah rebelled against the king of Babylon."[102] It is remarkable that though Jeremiah consistently urged submission to Babylon, the various arrangements made by Nebuchadnezzar did very little to improve the prophet's position or increase his influence. The Chaldean king may have seemed ungrateful only because he
  • 57.
    was ignorant ofthe services rendered to him—Jeremiah would not enter into direct and personal co-operation with the enemy of his country, even with him whom Jehovah had appointed to be the scourge of His disobedient people—but the Chaldean policy served Nebuchadnezzar as little as it profited Jeremiah. Jehoiakim, in spite of his forced submission, remained the able and determined foe of his suzerain, and Zedekiah, to the best of his very limited ability, followed his predecessor's example. Zedekiah was uncle of Jehoiachin, half-brother of Jehoiakim, and own brother to Jehoahaz.[103] Possibly the two brothers owed their bias against Jeremiah and his teaching to their mother, Josiah's wife Hamutal, the daughter of another Jeremiah, the Libnite. Ezekiel thus describes the appointment of the new king: "The king of Babylon ... took one of the seed royal, and made a covenant with him; he also put him under an oath, and took away the mighty of the land: that the kingdom might be base, that it might not lift itself up, but that by keeping of his covenant it might stand."[104] Apparently Nebuchadnezzar was careful to choose a feeble prince for his "base kingdom"; all that we read of Zedekiah suggests that he was weak and incapable. Henceforth the sovereign counted for little in the internal struggles of the tottering state. Josiah had firmly maintained the religious policy of Jeremiah, and Jehoiakim, as firmly, the opposite policy; but Zedekiah had neither the strength nor the firmness to enforce a consistent policy and to make one party permanently dominant. Jeremiah and his enemies were left to fight it out amongst themselves, so that now their antagonism grew more bitter and pronounced than during any other reign. But whatever advantage the prophet might derive from the weakness of the sovereign was more than counterbalanced by the recent deportation. In selecting the captives Nebuchadnezzar had sought merely to weaken Judah by carrying away every one who would have been an element of strength to the "base kingdom." Perhaps he rightly believed that neither the prudence of the wise nor the honour of the virtuous would overcome their patriotic hatred of
  • 58.
    subjection; weakness alonewould guarantee the obedience of Judah. He forgot that even weakness is apt to be foolhardy—when there is no immediate prospect of penalty. One result of his policy was that the enemies and friends of Jeremiah were carried away indiscriminately; there was no attempt to leave behind those who might have counselled submission to Babylon as the acceptance of a Divine judgment, and thus have helped to keep Judah loyal to its foreign master. On the contrary Jeremiah's disciples were chiefly thoughtful and honourable men, and Nebuchadnezzar's policy in taking away "the mighty of the land" bereft the prophet of many friends and supporters, amongst them his disciple Ezekiel and doubtless a large class of whom Daniel and his three friends might be taken as types. When Jeremiah characterises the captives as "good figs" and those left behind as "bad figs,"[105] and the judgment is confirmed and amplified by Ezekiel,[106] we may be sure that most of the prophet's adherents were in exile. We have already had occasion to compare the changes in the religious policy of the Jewish government to the alternations of Protestant and Romanist sovereigns among the Tudors; but no Tudor was as feeble as Zedekiah. He may rather be compared to Charles IX. of France, helpless between the Huguenots and the League. Only the Jewish factions were less numerous, less evenly balanced; and by the speedy advance of Nebuchadnezzar civil dissensions were merged in national ruin. The opening years of the new reign passed in nominal allegiance to Babylon. Jeremiah's influence would be used to induce the vassal king to observe the covenant he had entered into and to be faithful to his oath to Nebuchadnezzar. On the other hand a crowd of "patriotic" prophets urged Zedekiah to set up once more the standard of national independence, to "come to the help of the Lord against the mighty." Let us then briefly consider Jeremiah's polemic against the princes, prophets, and priests of his people. While
  • 59.
    Ezekiel in acelebrated chapter[107] denounces the idolatry of the princes, priests, and women of Judah, their worship of creeping things and abominable beasts, their weeping for Tammuz, their adoration of the sun, Jeremiah is chiefly concerned with the perverse policy of the government and the support it receives from priests and prophets, who profess to speak in the name of Jehovah. Jeremiah does not utter against Zedekiah any formal judgment like those on his three predecessors. Perhaps the prophet did not regard this impotent sovereign as the responsible representative of the state, and when the long-expected catastrophe at last befell the doomed people, neither Zedekiah nor his doings distracted men's attention from their own personal sufferings and patriotic regrets. At the point where a paragraph on Zedekiah would naturally have followed that on Jehoiachin, we have by way of summary and conclusion to the previous sections a brief denunciation of the shepherds of Israel. "Woe unto the shepherds that destroy and scatter the sheep of My pasture!... Ye have scattered My flock, and driven them away, and have not cared for them; behold, I will visit upon you the evil of your doings." These "shepherds" are primarily the kings, Jehoahaz, Jehoiakim, and Jehoiachin, who have been condemned by name in the previous chapter, together with the unhappy Zedekiah, who is too insignificant to be mentioned. But the term shepherds will also include the ruling and influential classes of which the king was the leading representative. The image is a familiar one in the Old Testament and is found in the oldest literature of Israel,[108] but the denunciation of the rulers of Judah as unfaithful shepherds is characteristic of Jeremiah, Ezekiel, and one of the prophecies appended to the Book of Zechariah.[109] Ezekiel xxxiv. expands this figure and enforces its lessons:—
  • 60.
    "Woe unto theshepherds of Israel that do feed themselves! Should not the shepherds feed the sheep? Ye eat the fat, and ye clothe you with the wool. Ye kill the fatlings; but ye feed not the sheep. The diseased have ye not strengthened, Neither have ye healed the sick, Neither have ye bound up the bruised, Neither have ye brought back again that which was driven away, Neither have ye sought for that which was lost, But your rule over them has been harsh and violent. And for want of a shepherd, they were scattered, And became food for every beast of the field."[110] So in Zechariah ix., etc., Jehovah's anger is kindled against the shepherds, because they do not pity His flock.[111] Elsewhere[112] Jeremiah speaks of the kings of all nations as shepherds, and pronounces against them also a like doom. All these passages illustrate the concern of the prophets for good government. They were neither Pharisees nor formalists; their religious ideals were broad and wholesome. Doubtless the elect remnant will endure through all conditions of society; but the Kingdom of God was not meant to be a pure Church in a rotten state. This present evil world is no manure heap to fatten the growth of holiness: it is rather a mass for the saints to leaven. Both Jeremiah and Ezekiel turn from the unfaithful shepherds whose "hungry sheep look up and are not fed" to the true King of Israel, the "Shepherd of Israel that led Joseph like a flock, and dwelt between the Cherubim." In the days of the Restoration He will raise up faithful shepherds, and over them a righteous Branch, the real Jehovah Zidqenu, instead of the sapless twig who disgraced the name "Zedekiah." Similarly Ezekiel promises that God will set up one shepherd over His people, "even My servant David." The pastoral care of Jehovah for His people is most tenderly and beautifully set
  • 61.
    forth in thetwenty-third Psalm. Our Lord, the root and the offspring of David, claims to be the fulfilment of ancient prophecy when He calls Himself "the Good Shepherd." The words of Christ and of the Psalmist receive new force and fuller meaning when we contrast their pictures of the true Shepherd with the portraits of the Jewish kings drawn by the prophets. Moreover the history of this metaphor warns us against ignoring the organic life of the Christian society, the Church, in our concern for the spiritual life of the individual. As Sir Thomas More said, in applying this figure to Henry VIII., "Of the multitude of sheep cometh the name of a shepherd."[113] A shepherd implies not merely a sheep, but a flock; His relation to each member is tender and personal, but He bestows blessings and requires service in fellowship with the Family of God. By a natural sequence the denunciation of the unfaithful shepherds is followed by a similar utterance "concerning the prophets." It is true that the prophets are not spoken of as shepherds; and Milton's use of the figure in Lycidas suggests the New Testament rather than the Old. Yet the prophets had a large share in guiding the destinies of Israel in politics as well as in religion, and having passed sentence on the shepherds—the kings and princes—Jeremiah turns to the ecclesiastics, chiefly, as the heading implies, to the prophets. The priests indeed do not escape, but Jeremiah seems to feel that they are adequately dealt with in two or three casual references. We use the term "ecclesiastics" advisedly; the prophets were now a large professional class, more important and even more clerical than the priests. The prophets and priests together were the clergy of Israel. They claimed to be devoted servants of Jehovah, and for the most part the claim was made in all sincerity; but they misunderstood His character, and mistook for Divine inspiration the suggestions of their own prejudice and self-will. Jeremiah's indictment against them has various counts. He accuses them of speaking without authority, and also of time-serving, plagiarism, and cant.
  • 62.
    First, then, asto their unauthorised utterances: Jeremiah finds them guilty of an unholy licence in prophesying, a distorted caricature of that "liberty of prophesying" which is the prerogative of God's accredited ambassadors. "Hearken not unto the words of the prophets that prophesy unto you. They make fools of you: The visions which they declare are from their own hearts, And not from the mouth of Jehovah. Who hath stood in the council of Jehovah, To perceive and hear His word? Who hath marked His word and heard it? I sent not the prophets—yet they ran; I spake not unto them—yet they prophesied." The evils which Jeremiah describes are such as will always be found in any large professional class. To use modern terms—in the Church, as in every profession, there will be men who are not qualified for the vocation which they follow. They are indeed not called to their vocation; they "follow," but do not overtake it. They are not sent of God, yet they run; they have no Divine message, yet they preach. They have never stood in the council of Jehovah; they might perhaps have gathered up scraps of the King's purposes from His true councillors; but when they had opportunity they neither "marked nor heard"; and yet they discourse concerning heavenly things with much importance and assurance. But their inspiration, at its best, has no deeper or richer source than their own shallow selves; their visions are the mere product of their own imaginations. Strangers to the true fellowship, their spirit is not "a well of water
  • 63.
    springing up untoeternal life," but a stagnant pool. And, unless the judgment and mercy of God intervene, that pool will in the end be fed from a fountain whose bitter waters are earthly, sensual, devilish. We are always reluctant to speak of ancient prophecy or modern preaching as a "profession." We may gladly dispense with the word, if we do not thereby ignore the truth which it inaccurately expresses. Men lived by prophecy, as, with Apostolic sanction, men live by "the gospel." They were expected, as ministers are now, though in a less degree, to justify their claims to an income and an official status, by discharging religious functions so as to secure the approval of the people or the authorities. Then, as now, the prophet's reputation, influence, and social standing, probably even his income, depended upon the amount of visible success that he could achieve. In view of such facts, it is futile to ask men of the world not to speak of the clerical life as a profession. They discern no ethical difference between a curate's dreams of a bishopric and the aspirations of a junior barrister to the woolsack. Probably a refusal to recognise the element common to the ministry with law, medicine, and other professions, injures both the Church and its servants. One peculiar difficulty and most insidious temptation of the Christian ministry consists in its mingled resemblances to and differences from the other professions. The minister has to work under similar worldly conditions, and yet to control those conditions by the indwelling power of the Spirit. He has to "run," it may be twice or even three times a week, whether he be sent or no: how can he always preach only that which God has taught him? He is consciously dependent upon the exercise of his memory, his intellect, his fancy: how can he avoid speaking "the visions of his own heart"? The Church can never allow its ministers to regard themselves as mere professional teachers and lecturers, and yet if they claim to be more, must they not often fall under Jeremiah's condemnation? It is one of those practical dilemmas which delight casuists and distress honest and earnest servants of God. In the early Christian
  • 64.
    centuries similar difficultiespeopled the Egyptian and Syrian deserts with ascetics, who had given up the world as a hopeless riddle. A full discussion of the problem would lead us too far away from the exposition of Jeremiah, and we will only venture to make two suggestions. The necessity, which most ministers are under, of "living by the gospel," may promote their own spiritual life and add to their usefulness. It corrects and reduces spiritual pride, and helps them to understand and sympathise with their lay brethren, most of whom are subject to a similar trial. Secondly, as a minister feels the ceaseless pressure of strong temptation to speak from and live for himself—his lower, egotistic self—he will be correspondingly driven to a more entire and persistent surrender to God. The infinite fulness and variety of Revelation is expressed by the manifold gifts and experience of the prophets. If only the prophet be surrendered to the Spirit, then what is most characteristic of himself may become the most forcible expression of his message. His constant prayer will be that he may have the child's heart and may never resist the Holy Ghost, that no personal interest or prejudice, no bias of training or tradition or current opinion, may dull his hearing when he stands in the council of the Lord, or betray him into uttering for Christ's gospel the suggestions of his own self-will or the mere watchwords of his ecclesiastical faction. But to return to the ecclesiastics who had stirred Jeremiah's wrath. The professional prophets naturally adapted their words to the itching ears of their clients. They were not only officious, but also time-serving. Had they been true prophets, they would have dealt faithfully with Judah; they would have sought to convince the people of sin, and to lead them to repentance; they would thus have given them yet another opportunity of salvation. "If they had stood in My council,
  • 65.
    They would havecaused My people to hear My words; They would have turned them from their evil way, And from the evil of their doings." But now:— "They walk in lies and strengthen the hands of evildoers, That no one may turn away from his sin. They say continually unto them that despise the word of Jehovah,[114] Ye shall have peace; And unto every one that walketh in the stubbornness of his heart they say, No evil shall come upon you." Unfortunately, when prophecy becomes professional in the lowest sense of the word, it is governed by commercial principles. A sufficiently imperious demand calls forth an abundant supply. A sovereign can "tune the pulpits"; and a ruling race can obtain from its clergy formal ecclesiastical sanction for such "domestic institutions" as slavery. When evildoers grow numerous and powerful, there will always be prophets to strengthen their hands and encourage them not to turn away from their sin. But to give the lie to these false prophets God sends Jeremiahs, who are often branded as heretics and schismatics, turbulent fellows who turn the world upside-down. The self-important, self-seeking spirit leads further to the sin of plagiarism:—
  • 66.
    "Therefore I amagainst the prophets, is the utterance of Jehovah, Who steal My word from one another." The sin of plagiarism is impossible to the true prophet, partly because there are no rights of private property in the word of Jehovah. The Old Testament writers make free use of the works of their predecessors. For instance, Isaiah ii. 2-4 is almost identical with Micah iv. 1-3; yet neither author acknowledges his indebtedness to the other or to any third prophet.[115] Uriah ben Shemaiah prophesied according to all the words of Jeremiah,[116] who himself owes much to Hosea, whom he never mentions. Yet he was not conscious of stealing from his predecessor, and he would have brought no such charge against Isaiah or Micah or Uriah. In the New Testament 2 Peter and Jude have so much in common that one must have used the other without acknowledgment. Yet the Church has not, on that ground, excluded either Epistle from the Canon. In the goodly fellowship of the prophets and the glorious company of the apostles no man says that the things which he utters are his own. But the mere hireling has no part in the spiritual communism wherein each may possess all things because he claims nothing. When a prophet ceases to be the messenger of God, and sinks into the mercenary purveyor of his own clever sayings and brilliant fancies, then he is tempted to become a clerical Autolycus, "a snapper-up of unconsidered trifles." Modern ideas furnish a curious parallel to Jeremiah's indifference to the borrowings of the true prophet, and his scorn of the literary pilferings of the false. We hear only too often of stolen sermons, but no one complains of plagiarism in prayers. Doubtless among these false prophets charges of plagiarism were bandied to and fro with much personal acrimony. But it is interesting to notice that Jeremiah is not denouncing an injury done to himself; he does not accuse them of thieving from him, but from one another. Probably assurance and lust of praise and power would have overcome any awe they felt for Jeremiah. He was only free from their depredations, because—from their point of
  • 67.
    view—his words werenot worth stealing. There was nothing to be gained by repeating his stern denunciations, and even his promises were not exactly suited to the popular taste. These prophets were prepared to cater for the average religious appetite in the most approved fashion—in other words, they were masters of cant. Their office had been consecrated by the work of true men of God like Elijah and Isaiah. They themselves claimed to stand in the genuine prophetic succession, and to inherit the reverence felt for their great predecessors, quoting their inspired utterances and adopting their weighty phrases. As Jeremiah's contemporaries listened to one of their favourite orators, they were soothed by his assurances of Divine favour and protection, and their confidence in the speaker was confirmed by the frequent sound of familiar formulæ in his unctuous sentences. These had the true ring; they were redolent of sound doctrine, of what popular tradition regarded as orthodox. The solemn attestation NE'UM YAHWE, "It is the utterance of Jehovah," is continually appended to prophecies, almost as if it were the sign-manual of the Almighty. Isaiah and other prophets frequently use the term MASSA (A.V., R.V., "burden") as a title, especially for prophecies concerning neighbouring nations. The ancient records loved to tell how Jehovah revealed Himself to the patriarchs in dreams. Jeremiah's rivals included dreams in their clerical apparatus:— "Behold, I am against them that prophesy lying dreams— Ne'um Yahwe— And tell them, and lead astray My people By their lies and their rodomontade; It was not I who sent or commanded them, Neither shall they profit this people at all, Ne'um Yahwe"
  • 68.
    These prophets "thoughtto cause the Lord's people to forget His name, as their fathers forgot His name for Baal, by their dreams which they told one another." Moreover they could glibly repeat the sacred phrases as part of their professional jargon:— "Behold, I am against the prophets, It is the utterance of Jehovah (Ne'um Yahwe), That use their tongues To utter utterances (Wayyin'amu Ne'um)." "To utter utterances"—the prophets uttered them, not Jehovah. These sham oracles were due to no Diviner source than the imagination of foolish hearts. But for Jeremiah's grim earnestness, the last clause would be almost blasphemous. It is virtually a caricature of the most solemn formula of ancient Hebrew religion. But this was really degraded when it was used to obtain credence for the lies which men prophesied out of the deceit of their own heart. Jeremiah's seeming irreverence was the most forcible way of bringing this home to his hearers. There are profanations of the most sacred things which can scarcely be spoken of without an apparent breach of the Third Commandment. The most awful taking in vain of the name of the Lord God is not heard among the publicans and sinners, but in pulpits and on the platforms of religious meetings. But these prophets and their clients had a special fondness for the phrase "The burden of Jehovah," and their unctuous use of it most especially provoked Jeremiah's indignation:— "When this people, priest, or prophet shall ask thee, What is the burden of Jehovah? Then say unto them, Ye are the burden.[117] But I will cast you off, Ne'um Yahwe. If priest or prophet or people shall say, The burden of Jehovah,
  • 69.
    I will punishthat man and his house. And ye shall say to one another, What hath Jehovah answered? and, What hath Jehovah spoken? And ye shall no more make mention of the burden of Jehovah: For (if ye do) men's words shall become a burden to themselves. Thus shall ye inquire of a prophet, What hath Jehovah answered thee? What hath Jehovah spoken unto thee? But if ye say, The burden of Jehovah, Thus saith Jehovah: Because ye say this word, The burden of Jehovah, When I have sent unto you the command, Ye shall not say, The burden of Jehovah, Therefore I will assuredly take you up, And will cast away from before Me both you and the city which I gave to you and to your fathers. I will bring upon you everlasting reproach And everlasting shame, that shall not be forgotten." Jeremiah's insistence and vehemence speak for themselves. Their moral is obvious, though for the most part unheeded. The most solemn formulæ, hallowed by ancient and sacred associations, used by inspired teachers as the vehicle of revealed truths, may be debased till they become the very legend of Antichrist, blazoned on the Vexilla Regis Inferni. They are like a motto of one of Charles's Paladins flaunted by his unworthy descendants to give distinction to cruelty and vice. The Church's line of march is strewn with such dishonoured relics of her noblest champions. Even our Lord's own
  • 70.
    words have notescaped. There is a fashion of discoursing upon "the gospel" which almost tempts reverent Christians to wish they might never hear that word again. Neither is this debasing of the moral currency confined to religious phrases; almost every political and social watchword has been similarly abused. One of the vilest tyrannies the world has ever seen—the Reign of Terror—claimed to be an incarnation of "Liberty, Equality, and Fraternity." Yet the Bible, with that marvellous catholicity which lifts it so high above the level of all other religious literature, not only records Jeremiah's prohibition to use the term "Burden," but also tells us that centuries later Malachi could still speak of "the burden of the word of Jehovah." A great phrase that has been discredited by misuse may yet recover itself; the tarnished and dishonoured sword of faith may be baptised and burnished anew, and flame in the forefront of the holy war. Jeremiah does not stand alone in his unfavourable estimate of the professional prophets of Judah; a similar depreciation seems to be implied by the words of Amos: "I am neither a prophet nor of the sons of the prophets."[118] One of the unknown authors whose writings have been included in the Book of Zechariah takes up the teaching of Amos and Jeremiah and carries it a stage further:— "In that day (it is the utterance of Jehovah Sabaoth) I will cut off the names of the idols from the land, They shall not be remembered any more; Also the prophets and the spirit of uncleanness Will I expel from the land. When any shall yet prophesy, His father and mother that begat him shall say unto him, Thou shalt not live, for thou speakest lies in the name of Jehovah: And his father and mother that begat him shall thrust him through when he prophesieth.
  • 71.
    In that dayevery prophet when he prophesieth shall be ashamed of his vision; Neither shall any wear a hairy mantle to deceive: He shall say, I am no prophet; I am a tiller of the ground, I was sold for a slave in my youth."[119] No man with any self-respect would allow his fellows to dub him prophet; slave was a less humiliating name. No family would endure the disgrace of having a member who belonged to this despised caste; parents would rather put their son to death than see him a prophet. To such extremities may the spirit of time-serving and cant reduce a national clergy. We are reminded of Latimer's words in his famous sermon to Convocation in 1536: "All good men in all places accuse your avarice, your exactions, your tyranny. I commanded you that ye should feed my sheep, and ye earnestly feed yourselves from day to day, wallowing in delights and idleness. I commanded you to teach my law; you teach your own traditions, and seek your own glory."[120] Over against their fluent and unctuous cant Jeremiah sets the terrible reality of his Divine message. Compared to this, their sayings are like chaff to the wheat; nay, this is too tame a figure—Jehovah's word is like fire, like a hammer that shatters rocks. He says of himself:— "My heart within me is broken; all my bones shake: I am like a drunken man, like a man whom wine hath overcome, Because of Jehovah and His holy words." Thus we have in chapter xxiii. a full and formal statement of the controversy between Jeremiah and his brother-prophets. On the one hand, self-seeking and self-assurance winning popularity by orthodox phrases, traditional doctrine, and the prophesying of smooth things; on the other hand, a man to whom the word of the Lord was like a
  • 72.
    fire in hisbones, who had surrendered prejudice and predilection that he might himself become a hammer to shatter the Lord's enemies, a man through whom God wrought so mightily that he himself reeled and staggered with the blows of which he was the instrument. The relation of the two parties was not unlike that of St. Paul and his Corinthian adversaries: the prophet, like the Apostle, spoke "in demonstration of the Spirit and of power"; he considered "not the word of them which are puffed up, but the power. For the kingdom of God is not in word, but in power." In our next chapter we shall see the practical working of this antagonism which we have here set forth.
  • 73.
    CHAPTER IX HANANIAH xxvii., xxviii. "Hearnow, Hananiah; Jehovah hath not sent thee, but thou makest this people to trust in a lie."—Jer. xxviii. 15. The most conspicuous point at issue between Jeremiah and his opponents was political rather than ecclesiastical. Jeremiah was anxious that Zedekiah should keep faith with Nebuchadnezzar, and not involve Judah in useless misery by another hopeless revolt. The prophets preached the popular doctrine of an imminent Divine intervention to deliver Judah from her oppressors. They devoted themselves to the easy task of fanning patriotic enthusiasm, till the Jews were ready for any enterprise, however reckless. During the opening years of the new reign, Nebuchadnezzar's recent capture of Jerusalem and the consequent wholesale deportation were fresh in men's minds; fear of the Chaldeans together with the influence of Jeremiah kept the government from any overt act of rebellion. According to li. 59, the king even paid a visit to Babylon, to do homage to his suzerain. It was probably in the fourth year of his reign[121] that the tributary Syrian states began to prepare for a united revolt against Babylon. The Assyrian and Chaldean annals constantly mention such combinations, which were formed and broken up and reformed with as much ease and variety as patterns in a kaleidoscope. On the present occasion the kings of Edom, Moab, Ammon, Tyre, and Zidon sent their ambassadors to Jerusalem to arrange with Zedekiah for concerted action. But there were more important persons to deal
  • 74.
    with in thatcity than Zedekiah. Doubtless the princes of Judah welcomed the opportunity for a new revolt. But before the negotiations were very far advanced, Jeremiah heard what was going on. By Divine command, he made "bands and bars," i.e. yokes, for himself and for the ambassadors of the allies, or possibly for them to carry home to their masters. They received their answer, not from Zedekiah, but from the true King of Israel, Jehovah Himself. They had come to solicit armed assistance to deliver them from Babylon; they were sent back with yokes to wear as a symbol of their entire and helpless subjection to Nebuchadnezzar. This was the word of Jehovah:— "The nation and the kingdom that will not put its neck beneath the yoke of the king of Babylon, That nation will I visit with sword and famine and pestilence until I consume them by his hand." The allied kings had been encouraged to revolt by oracles similar to those uttered by the Jewish prophets in the name of Jehovah; but:— "As for you, hearken not to your prophets, diviners, dreams, soothsayers and sorcerers, When they speak unto you, saying, Ye shall not serve the king of Babylon. They prophesy a lie unto you, to remove you far from your land; That I should drive you out, and that you should perish. But the nation that shall bring their neck under the yoke of the king of Babylon, and serve him, That nation will I maintain in their own land (it is the utterance of Jehovah), and they shall till it and dwell in it." When he had sent his message to the foreign envoys, Jeremiah addressed an almost identical admonition to his own king. He bids
  • 75.
    him submit tothe Chaldean yoke, under the same penalties for disobedience—sword, pestilence, and famine for himself and his people. He warns him also against delusive promises of the prophets, especially in the matter of the sacred vessels. The popular doctrine of the inviolable sanctity of the Temple had sustained a severe shock when Nebuchadnezzar carried off the sacred vessels to Babylon. It was inconceivable that Jehovah would patiently submit to so gross an indignity. In ancient days the Ark had plagued its Philistine captors till they were only too thankful to be rid of it. Later on a graphic narrative in the Book of Daniel told with what swift vengeance God punished Belshazzar for his profane use of these very vessels. So now patriotic prophets were convinced that the golden candlestick, the bowls and chargers of gold and silver, would soon return in triumph, like the Ark of old; and their return would be the symbol of the final deliverance of Judah from Babylon. Naturally the priests above all others would welcome such a prophecy, and would industriously disseminate it. But Jeremiah "spake to the priests and all this people, saying, Thus saith Jehovah: — "Hearken not unto the words of your prophets, which prophesy unto you, Behold, the vessels of the house of Jehovah shall be brought back from Babylon now speedily: For they prophesy a lie unto you." How could Jehovah grant triumphant deliverance to a carnally minded people who would not understand His Revelation, and did not discern any essential difference between Him and Moloch and Baal? "Hearken not unto them; serve the king of Babylon and live. Why should this city become a desolation?"
  • 76.
    Welcome to ourwebsite – the perfect destination for book lovers and knowledge seekers. We believe that every book holds a new world, offering opportunities for learning, discovery, and personal growth. That’s why we are dedicated to bringing you a diverse collection of books, ranging from classic literature and specialized publications to self-development guides and children's books. More than just a book-buying platform, we strive to be a bridge connecting you with timeless cultural and intellectual values. With an elegant, user-friendly interface and a smart search system, you can quickly find the books that best suit your interests. Additionally, our special promotions and home delivery services help you save time and fully enjoy the joy of reading. Join us on a journey of knowledge exploration, passion nurturing, and personal growth every day! ebookbell.com