water quality assessment in the hawkesbury …

i

School of Computing, Engineering and Mathematics

WATER QUALITY ASSESSMENT IN THE

HAWKESBURY NEPEAN RIVER SYSTEM,

NEW SOUTH WALES

Kuruppu Arachchige Upeka Kanchnamalie Kuruppu

Supervisory panel:

Principal Supervisor : A/Prof. Ataur Rahman

Co-supervisors :

A/Prof. Arumugam Sathasivan

Prof. Basant Maheshwari

A/Prof. Gary Dennis

This thesis is presented for the degree of Master of Engineering (Honours) in

the

Western Sydney University

02 May, 2016

ii

Statement of Authentication

I hereby declare that this thesis is my own work and to the best of my knowledge

it contains no materials previously published or written by another person, nor

material which to a substantial extent has been accepted for the award of any other

degree or diploma at Western Sydney University or any other educational

institution, except where due acknowledgement is made in the thesis.

I also declare that the intellectual content of this thesis is the product of my own

work, except to the extent that assistance from others in the project’s design and

conception or in style, presentation and linguistic expression is acknowledged.

Signature……………………………………………………..

Date…………………………………………………………..

iii

Abstract

Surface waters are the most vulnerable to pollution due to their easy accessibility

for disposal of wastewaters. Both the natural processes as well as the

anthropogenic influences together determine the quality of surface water. The

Hawkesbury Nepean River system (HNRS) is an icon of Australia’s largest city,

Sydney, with important ecological, social and economic values. Since European

settlement, the reliance on this river system has steadily increased to meet the

drinking water requirements of the population, and it now provides 97% of fresh

drinking water to more than 4.8 million people living in and around Sydney.

HNRS has been placed under increasing pressure and the environmental health of

the river system has suffered due to the increasing development and population

growth over time. The river regulation has resulted in large volumes of water

being extracted for drinking water, irrigation and industrial uses. There are a

number of sewage treatment plants (STPs) located in the catchment, and

stormwater runoff from agricultural and urban areas can also carry pollutants into

the river system. Algal and introduced macrophyte blooms have commonly

occurred in the past and are likely to continue to occur in the future unless serious

intervention is made by the NSW Government.

Identifying the deteriorated section of a river and actual sources of pollution along

different parts of the river helps to make suitable pollution prevention activities.

Therefore, this study attempts to investigate the state of the HNRS, using water

quality data from the past 20 years. Therefore, the following objectives are

iv

primarily emphasized in this thesis:

Assess the water quality in the HNRS.

Assess the trend of water quality in the HNRS.

Develop prediction equations to predict water quality from surrogate water

quality parameters.

Assess the impact of land use on the water quality of the HNRS.

Develop a water quality index for the river in order to conduct an overall

evaluation of the water quality of the river.

This thesis consists of a series of experimental and numerical studies. They

include exploratory analysis, trend analysis, principal component analysis, factor

analysis, regression analysis, and application of water quality index method to

make an overall water quality assessment of the HNRS.

This study has found that the concentrations of total phosphorus, nitrogen oxides

and chlorophyll along the HNRS are higher than those recommended by the

Australian and New Zealand Environment and Conservation Council (ANZECC)

guidelines. An increasing trend for turbidity, chlorophyll-a, alkalinity, dissolved

organic carbon, total iron, total aluminium, total manganese and reactive silicate

has also been detected for majority of the monitoring stations. Application of the

Canadian Water Quality Index (WQI) method shows that the water quality at 9

stations fall under either the poor or marginal category. Stations N14 and N35

v

were found to be the most polluted stations in the HNRS among the 9 stations.

There are many sewage treatment plants discharging treated wastewater to

upstream of N35. Also, the dominant land use in this part of the catchment

includes rural, grazing, commercial gardening, intensive agriculture and urban and

industrial activities. These land uses can be attributed to the low WQI at N35.

Water quality at station N14 should be improved due to dilution by high quality

inflows from the Colo River and the undisturbed upstream catchment. The high

pollutant levels at N14 need to be investigated to find the possible reasons and to

devise controlling measures. Although an improvement in water quality can be

seen at some stations downstream of the undisturbed parts of the catchment, there

has been an overall water quality deterioration in the HNRS during the last

decade.

The HNRS is a very important river system of Australia .The findings of this

study would provide an important basis for better land use planning in the

catchment of the HNRS, which would improve the overall state of the river water

quality.

vi

Dedications

To my loving husband Sameera,

for his love, understanding, encouragement and great support

&

to my loving daughter Vinuki and son Imeth.

vii

Acknowledgement

I would like to express my deep and sincere appreciation to my principle

supervisor, Associate Professor Ataur Rahman, for his endless support,

exceptional advice, guidance, supervision and encouragement throughout every

stage of my Masters Research. I would like to thank my co-supervisors,

Associate Professor Arumugam Sathasivan, Professor Basant Maheshwari

and Associate Professor Gary Dennis, for their valuable guidance and support. I

would also like to acknowledge the assistance of Dr Md Mahmudul Haque in

statistical analysis.

I would like to acknowledge Ms. Tracey Schultz and Mr. Ramen Charan at

Sydney Catchment Authority for their great support by providing the water

quality data needed for this study.

I would like to thank my work supervisor Mr. Kiran KC and the University of

Western Sydney for providing the opportunity to undertake a Masters Research

degree.

I am indebted to my parents, Mr. Sisira Kuruppu and Mrs. Manel Hyacinth, for

their love, support and inspiration.

viii

PREFACE

This thesis is submitted in fulfilment of the requirements for the degree of Masters

Honours at The Western Sydney University, NSW, Australia. The work described

herein was performed by the candidate from the School of Computing,

Engineering and Mathematics, Western Sydney University. The candidate was

supervised by Associate Professor Ataur Rahman (as Principal Supervisor) during

the period of March 2013 to October 2015. The thesis has been supported by

papers and book chapters that have been submitted for consideration, accepted or

published in internationally renowned journals and conferences. These papers and

book chapters are listed below:

Book chapters

Kuruppu, U., Haque, M.M., Rahman, A. (2016), Water quality in the

urban rivers: A case study for the Hawkesbury-Nepean River system in

Australia. In Water Resources: Problems and Solutions, Edited by

Jonathan Y.S. Leung, OMICS Group International – eBooks, USA.

(Accepted and in press).

Journal papers

Kuruppu, U., Rahman, A. (2015). Trends in water quality data in the

Hawkesbury-Nepean River System, Australia, Journal of Water and

Climate Change, doi:10.2166/wcc.2015.120. (IF=1.044, 5-Year IF=1.00,

ix

relative ranking 52/81 in water resources category, ISSN: 2040-2244, Q2,

ERA 2010 ranking: B).

Conference papers

Kuruppu, U., Rahman, A. (2013). An Exploratory Analysis of Water

Quality in the Nepean River, Australia, 35th IAHR World Congress.

September 8 to 13, 2013 Chengdu, China, 1-6.

Kuruppu, U., Rahman. A., Haque, M.M., Sathasivan, A. (2013). Water

Quality Investigation in the Hawkesbury- Nepean River in Sydney Using

Principal Component Analysis, 20th

International Congress on Modelling

and Simulation, 1 to 6 December, 2013, Adelaide, Australia, 2646-2652.

x

TABLE OF CONTENTS

CHAPTER 1

INTRODUCTION ................................................................................................. 1

1.1 Overview .................................................................................................. 1

1.2 Background .............................................................................................. 2

1.3 Expected Outcomes, Values and Benefits ................................................ 3

1.3.1 Why is this particular piece of research worth doing? ...................... 3

1.3.2 What special groups stand to benefit? ............................................... 4

1.4 Research Questions .................................................................................. 4

1.5 Methodology ............................................................................................ 5

1.6 Thesis Structure ........................................................................................ 6

CHAPTER 2

LITERATURE REVIEW ....................................................................................... 8

2.1 River Water Quality ................................................................................. 8

2.2 Hawkesbury-Nepean River System ........................................................ 14

CHAPTER 3

DESCRIPTION OF METHODS .......................................................................... 18

xi

3.1 Overview ................................................................................................ 18

3.2 Preliminary data analysis – Boxplots (box-and-whisker plots) .............. 20

3.3 Principal Component Analysis and Factor Analysis .............................. 21

3.4 Mann–Kendall statistical test and Sen’s slope analysis ......................... 23

3.5 Regression analysis ................................................................................ 25

3.6 Water Quality index method .................................................................. 26

3.7 Chapter Summary ................................................................................... 30

CHAPTER 4

THE STUDY AREA AND DATA ....................................................................... 32

4.1 Overview ................................................................................................ 32

4.2 Description of land use in Hawkesbury Nepean River catchment and

information on treated waste water discharge to HNRS ................................... 32

4.3 Data Requirements ................................................................................. 37

4.4 Water sampling and testing .................................................................... 41

4.4.1 Location Selection and Characterisation ......................................... 41

4.5 Sampling locations ................................................................................. 41

xii

CHAPTER 5

RESULTS ON ASSESSMENT OF RIVER WATER QUALITY ....................... 48

5.1 Overview ................................................................................................ 48

5.2 Preliminary water quality data analysis .................................................. 48

5.2.1 pH .................................................................................................... 48

5.2.2 Temperature .................................................................................... 50

5.2.3 Dissolved Oxygen ........................................................................... 52

5.2.4 Conductivity .................................................................................... 53

5.2.5 Turbidity .......................................................................................... 55

5.2.6 Phosphorus ...................................................................................... 58

5.2.7 Nitrogen .......................................................................................... 63

5.2.8 Alkalinity ........................................................................................ 70

5.2.9 Suspended solids ............................................................................. 71

5.2.10 Algae and chlorophyll-a .................................................................. 73

5.3 Results from principal component analysis (PCA) ................................ 78

5.4 Long term trends in water quality data ................................................... 84

xiii

5.5 Results from regression analysis for developing prediction equations for

water quality parameters.................................................................................... 97

5.6 Results of water quality assessment by using Water Quality Index ..... 101

5.7 Comparison of measured water quality data with SCA data ................ 110

5.7.1 pH .................................................................................................. 111

5.7.2 Dissolved Oxygen ......................................................................... 112

5.7.3 Electrical Conductivity.................................................................. 114

5.7.4 Turbidity ........................................................................................ 115

5.7.5 Nitrogen Oxides ............................................................................ 117

5.7.6 Ammonical Nitrogen ..................................................................... 118

5.7.7 Temperature .................................................................................. 120

5.8 Chapter Summary ................................................................................. 121

CHAPTER 6

SUMMARY AND CONCLUSTIONS ............................................................... 124

6.1 Summary .............................................................................................. 124

6.1 Preliminary water quality data analysis ................................................ 124

xiv

6.2 Trend Analysis...................................................................................... 125

6.3 Regression Analysis ............................................................................. 125

6.4 Application of Canadian Water Quality Index method ........................ 126

6.5 Comparison of measured water quality data with SCA data ................ 127

6.6 Conclusion ............................................................................................ 127

6.7 Limitations of the study ........................................................................ 127

6.7 Suggestions for Future Research .......................................................... 128

REFERENCES .................................................................................................. 129

xv

TABLE OF FIGURES

Figure 1.1. Methodology. ........................................................................................ 5

Figure 2.1: Land use in Hawksburn-Nepean catchment (BOM, 2013). ............... 16

Figure 3.1. Components of a default boxplot. ....................................................... 19

Figure 4.1. Schematic diagram of the HNRS with land use details. ..................... 34

Figure 4.2. Locations of the 9 sampling stations adopted in the preliminary

assessment. ............................................................................................................ 39

Figure 4.3. Locations of sampling stations. .......................................................... 42

Figure 4.4. Sampling stations. ............................................................................... 43

Figure 5.1. Box plot of pH values at different measuring stations along the

Hawkesbury Nepean River System. ...................................................................... 50

Figure 5.2. Box plot of measured temperature along the Hawkesbury Nepean

River System. ........................................................................................................ 51

Figure 5.3. Box plot of DO along the Hawkesbury Nepean River System. ........ 52

Figure 5.4. Box plot of conductivity along the Hawkesbury Nepean River System

for all sampling stations (with the scale up to 50 mS/cm). ................................... 54

Figure 5.5. Box plot of conductivity along the Hawkesbury Nepean River System

for all sampling stations (with the scale up to 3 mS/cm). ..................................... 55

Figure 5.6. Box plot of turbidity along the Hawkesbury Nepean River System

(with the scale up to 500 NTU). ............................................................................ 57

Figure 5.7. Box plot of turbidity along the Hawkesbury Nepean River System

xvi

(with the scale up to 100 NTU). ............................................................................ 58

Figure 5.8. Box plot of total phosphorus along the HNRS (with the scale up to 0.4

mg/L). .................................................................................................................... 60

Figure 5.9. Box plot of total phosphorus along the Hawkesbury Nepean River

System (with the scale up to 0.2 mg/L). ............................................................... 61

Figure 5.10. Box plot of filterable phosphorus along the Hawkesbury Nepean

River System (with the scale up to 0.25 mg/L). .................................................... 62

Figure 5.11. Box plot of filterable phosphorus along the Hawkesbury Nepean

River System (with the scale up to 0.05 mg/L). .................................................... 63

Figure 5.12. Box plot of total nitrogen along the Hawkesbury Nepean River

System. .................................................................................................................. 64

Figure 5.13. Box plot of total nitrogen along the Hawkesbury Nepean River

System. .................................................................................................................. 65

Figure 5.14. Box plot of nitrogen oxidised along the Hawkesbury Nepean River

System. .................................................................................................................. 66

Figure 5.15. Box plot of ammoniacal nitrogen along the Hawkesbury Nepean

River System (with the scale up to 1.2 mg/L). ...................................................... 67

Figure 5.16. Box plot of ammoniacal nitrogen along the Hawkesbury Nepean

River System (with the scale up to 0.5 mg/L). ...................................................... 68

Figure 5.17. Box plot of nitrogen TKN along the Hawkesbury Nepean River

System. .................................................................................................................. 69

Figure 5.18. Box plot of alkalinity along the Hawkesbury Nepean River System.

............................................................................................................................... 71

xvii

Figure 5.19. Box plot of suspended solids along the Hawkesbury Nepean River

System (with scale up to 400 mg/L). .................................................................... 72

Figure 5.20. Box plot of suspended solids along the Hawkesbury Nepean River

System (with scale up to 50 mg/L). ...................................................................... 73

Figure 5.21. Box plot of algal total count along the Hawkesbury Nepean River

System (with the scale up to 700,000 cells/mL). .................................................. 74

Figure 5.22. Box plot of algal total count along the Hawkesbury Nepean River

System (with the scale up to 200,000 cells/mL). .................................................. 75

Figure 5.23. Box plot of chlorophyll-a along the Hawkesbury Nepean River

System (with the scale up to 250 ug/L). ................................................................ 76

Figure 5.24. Box plot of chlorophyll-a along the Hawkesbury Nepean River

System (with the scale up to 50 ug/L). .................................................................. 77

Figure 5.25. Median values of pH along the Hawkesbury Nepean River System. 87

Figure 5.26. Decreasing trend of DO at station N35............................................. 88

Figure 5.27. Decreasing trend of EC at station N14. ............................................ 89

Figure 5.28. Median values of chlorophyll-a along the Hawkesbury Nepean River

System. .................................................................................................................. 93

Figure 5.29. Increasing trend of alkalinity at station N92..................................... 95

Figure 5.30. Increasing trends of reactive silicate at station N35. ........................ 96

Figure 5.31. Plot of standardized residuals against estimate for Chlorophyll-a. 100

Figure 5.32. Plot of standardized residuals against estimate for total nitrogen. . 100

xviii

Figure 5.33. Plot of standardized residuals against estimate for total phosphorous.

............................................................................................................................. 101

Figure 5.34. Change in WQI over time for 9 monitoring stations in HNRS. ..... 103

Figure 5.35. Average WQI along the HNRS. ..................................................... 104

Figure 5.36. Scope, frequency and amplitude values at 9 monitoring stations in

HNRS. ................................................................................................................. 105

Figure 5.37. pH at S1 and N67. ........................................................................... 111






Figure 5.43. Electrical conductivity at S1 and N67. ........................................... 114



Figure 5.46. Turbidity at S1 and N67. ................................................................ 115



Figure 5.49. Nitrogen oxides at S1 and N67. ...................................................... 117


xix


Figure 5.52. Ammonical nitrogen at S1 and N67. .............................................. 118



Figure 5.55. Temperature at S1 and N67. ........................................................... 120



xx

LIST OF TABLES

Table 4.1: Sewage treatment plants along the HWNRS ....................................... 35

Table 4.2: Water quality monitoring stations used in the preliminary assessment 38

Table 4.3: Water quality parameters considered in the preliminary assessment .. 40

Table 4.4: Water quality data at Blaxland Crossing ............................................. 45

Table 4.5: Water quality data at M4...................................................................... 46

Table 4.6: Water quality data at Weir Reserve ..................................................... 47

Table 5.1: Principal components with eigenvalues > 1 ........................................ 78

Table 5.2: Component score coefficients for first three PCs (for monitoring

stations) ................................................................................................................. 79

Table 5.3: Varimax rotated factor loadings (for first 5 factors) ............................ 80

Table 5.4: Explained variance and eigenvalues (for water parameters) ............... 81

Table 5.5: Component loadings for first eight PCs (water quality parameters) .... 83

Table 5.6: Median values of water quality parameters and ANZECC (2000)

guidelines .............................................................................................................. 85

Table 5.7: Mann-Kendal test results and yearly Sen’s slope ................................ 86

Table 5.8: Correlations among water quality parameters at station N44 of the

HNRS .................................................................................................................... 98

Table 5.9: Water quality parameters and ANZEC Guidelines for Fresh and Marine

Water Quality ...................................................................................................... 102

Table 5.10: Amplitudes at 9 stations in different years ...................................... 106

xxi

Table 5.11: Water quality results at N14 ............................................................ 107

xxii

SYMBOLS

ALK Alkalinity

B A constant

CHLA Chlorophyll-a

DO Dissolved oxygen

DOC Dissolved organic carbon

EC Conductivity field

ECOCC Enterococci

ECOL E. coli

F1 Scope

F2 Frequency

F3 Amplitude

AF Aluminium filtered

FI Iron filtered

FM Manganese filtered

xxiii

FP Phosphorus filterable

F-A Factor analysis

LOR Lorenzen

MK Mann–Kendall

n Length of the data set

nes Normalised sum of excursions

NH-N Nitrogen ammonical

NO Nitrogen oxidised

PH pH

r Pearson correlation coefficient

PHA Phaeophytin

Q Slop

R2 Coefficient of determination

RS Silicate reactive

SS Suspended solids

ti Number of ties of extent i

xxiv

TA Aluminium total

TCOL True colour

TEMP Temperature

TI Iron total

TKN Nitrogen TKN

TM Manganese total

TN Nitrogen total

TP Phosphorus total

TUR Turbidity

UV UV absorbing constituents

VFs Varifactors

x Sequential data value

Z Standard test statistics

xxv

ABBREVIATIONS

ANZECC New Zealand guidelines for fresh and marine water quality

BOM Burro of metrology

EPA Environmental protection authority

HDPE High density poly ethylene

HNRS Hawkesbury Nepean River system

IUCN International union for conservation of nature and natural

Resources

NSW New South Wales

PCA Principal Component analysis

SCA Sydney catchment authority

UK United Kingdom

US United States

WQA Water quality analyser

CHAPTER 01: Introduction

1

CHAPTER 1

INTRODUCTION

1

1.1 OVERVIEW

Society benefits immeasurably from rivers. They are the main water resource in many

inland areas for drinking, irrigation and industrial purposes. Also, rivers provide a

recreational value to the adjoining community by supporting boating, fishing and

outdoor activities. Although rivers contain only about 0.0001% of the total amount of

water on earth at any given time, they are vital carriers of water and nutrients to areas

all around the earth. They are a critical component of the hydrological cycle, acting as

drainage channels for surface water. The world's rivers drain nearly 75% of the earth's

land surface. They act as habitats, and provide nourishment and means of transport to

countless organisms; their powerful forces create majestic scenery; they provide travel

routes for exploration, commerce and recreation; they leave valuable deposits of

sediments, such as sand and gravel; they form vast floodplains where many of our cities

are built; and their power provides much of the electrical energy (e.g. hydro-electricity)

we use in our everyday lives. Water quality in the urban environment has become

important in recent Australian urban development and water management (e.g. Van der

Sterren et al., 2009; Van der Sterren et al., 2015). This thesis focuses on the study of an

important river in Australia, known as the Hawkesbury Nepean River system. This

chapter presents the background of the research, expected outcomes, values and


2

benefits, special groups stand to benefit, research questions, and methodology and the

thesis structure.

1.2 BACKGROUND

This Masters thesis provides findings of a study examining the pollution level and their

sources along different parts of the Hawkesbury-Nepean River system (HNRS), located

in New South Wales,Australia.

The objectives of this study are to:

Assess the water quality in the Hawkesbury-Nepean River system.

Assess the trend of water quality in the Hawkesbury-Nepean River system.

Develop prediction equations to predict water quality from surrogate water quality

variables.

Develop a water quality index to describe the overall quality of the river.

The selection of the HNRS as the key focus for this study was based on a number of

reasons. Firstly, it is the main source of fresh drinking water supply to more than 4.8

million people living in, and around, Sydney. Secondly, over 70% of the HNRS flows

through extensive peri-urban areas in Western Sydney and, as a result, the river system

clearly indicates a gradual degradation due to peri-urban pressures such as water

extraction for agriculture, discharge of treated sewage and pollutants from humans.


3

Thirdly, the HNRS has a large number of interested stakeholders, making it easy to

study the conflicting social issues towards its sustainable management. Finally, there is

an existing historic and current water quality data set available for this river system,

which can be used for analysis and comparison during the study.

1.3 EXPECTED OUTCOMES, VALUES AND BENEFITS

1.3.1 Why is this particular piece of research worth doing?

Being the main fresh drinking water supply for more than 4.8 million people;

monitoring and assessing the water quality of the Hawkesbury-Nepean River system is

of immense importance. Although many government organizations, researchers and

environmental agencies have spent millions of dollars every year to monitor and collect

water quality data along the river, the full capacity of the water quality data set has not

been used to draw meaningful conclusions describing the state of the river. This is due

to the complexity of analysing these data. Investigating the most important sampling

stations and water quality parameters is needed for developing cost-effective

monitoring programs. Also, this collected complex water quality data should be

summarized in a way that can be easily understood by the public, water distributors,

planners, managers and policy makers. Available data should be effectively used to

understand the current state of the river and develop restoration plans, estimate the

ecological risks associated with land use plans in a watershed, or select among pre-

existing, alternative development options to minimise overall river degradation.


4

1.3.2 What special groups stand to benefit?

Findings of this research support the long term river health management strategies to

achieve sustainable, river health goals, as well as short term advisory information for

frequent river users. River management authorities will also benefit from the outcomes

of this research. Further, it will be a good source of information for different research of

this large river system while providing guidance for the selection of water quality

indicators for efficient monitoring. The general public who have an interest in the

Hawkesbury-Nepean River system will receive important information that will help

them to make informed decisions.

1.4 RESEARCH QUESTIONS

The following research questions were addressed in this study:

Has the water quality in the Hawkesbury-Nepean River system (HNRS) improved

in recent years?

Is it possible to link the water quality in the HNRS with land use changes?

Is it possible to develop surrogate equations to predict water quality from easily

measureable water quality parameters?

Does the water quality in this river meet the national standards (e.g. Sydney Water,

Sydney Catchment Authority and Australian and New Zealand water quality

guidelines)?


5

1.5 METHODOLOGY

Figure 1.1 presents the overall methodology adopted in this study. It consists of data

collation and application of various statistical techniques to address the research

questions. The main statistical techniques adopted in this thesis include box plot

analysis, trend investigation, principal component analysis and regression analysis. A

water quality index method has been used to make an overall assessment of water

quality in the HNRS.

Figure 1.1. Methodology adopted in this study.


6

1.6 THESIS STRUCTURE

The research presented in this study has been organised into six chapters, as outlined

below.

Chapter 1 presents a brief introduction to the proposed research, including the

background and the importance of performing the proposed research. The aims,

objectives and the research questions of the proposed research are also presented in this

chapter.

Chapter 2 presents a literature review on previous and ongoing water quality,

monitoring programs for the Hawkesbury-Nepean River system, and other similar

international studies. The gaps in the research are identified.

Chapter 3 presents the description of the methods in detail and the underlying

assumptions and limitations.

Chapter 4 presents the study area and data collation procedure. A summary of the

measured data is also presented in this chapter.

Chapter 5 presents the preliminary data analysis. This applies Principal Component

Analysis (PCA) and Factor Analysis (FA) to reduce the dimensionality of the data set

and multiple linear regression analysis to developing prediction equations using easily

measurable parameters, and Mann-Kendall (MK) test and Sen’s slope estimator to

assess the trends of water quality parameters.


7

Chapter 6 presents a summary, conclusions and recommendations for further study.

CHAPTER 02: Literature Review

8

CHAPTER 2

LITERATURE REVIEW

2

2.1 RIVER WATER QUALITY

Water quality in urban environments is important in terms of management of stormwater and

receiving water quality of river systems (Van der Sterren et al., 2012). River water quality

depends on various geologic, climatic, catchment and land use characteristics. Among these,

climate and land use are the key drivers of water quality in a river system. But determining

the relative influence of these factors on water quality remains a significant challenge for

aquatic science and management (Interlandi and Crockett, 2003). Various pollutant sources

related to industries, urbanization, agriculture and mining can have a strong impact on a river

system (Kendall et al., 2007; Tian and Fernandez, 2000). In recent years, an increasing

awareness has been noticed in different countries about the impacts of anthropogenic

activities on river water quantity and quality (Dawson and Macklin, 1998; Ma et al., 2009;

Erturk et al., 2010; Tabari et al., 2011). Climate change and urbanisation are key factors

affecting the future of water quality and quantity in urbanised catchments, and are associated

with significant uncertainty (Astarair-Imani et al., 2012). Pollutant build up and wash off in

connection with urban catchments have become a focus of current research in different

countries (Rahman et al., 2002; Egodawatta et al., 2009; Van der Sterren et al., 2013; Van der

Sterren et al., 2014; Haddad et al., 2013). In this regard, the roles of rainwater harvesting

systems and water sensitive urban designs have become relevant to control water quality in


9

urban rivers (Van Der Sterren et al., 2015; Eroksuz and Rahman, 2010; Rahman et al., 2012).

Rivers play a major role in assimilating or carrying off industrial and municipal wastewater,

manure discharges and runoff from agricultural fields, roadways and streets, which are

responsible for river pollution (Stroomberg et al., 1995; Vega et al., 1998). Applying the

concept of health to rivers is a logical outgrowth of scientific principles, legal mandates, and

changing societal values (Karr, 1999). Surface waters can be contaminated by human

activities in two ways: (1) by point sources, such as sewage treatment discharge and

industrial discharge; and (2) by non-point sources such as runoff from urban and agricultural

areas. Non-point sources are especially difficult to detect since they generally encompass

large areas in drainage basins and involve complex biotic and abiotic interactions (Solbe,

1986).

Over the past century, humans have changed many rivers dramatically, threatening river

health. As a result, societal well-being is also threatened because goods and services critical

to human society are being depleted. Having reliable information of water quality is essential

for effective and efficient water management, as it provides information regarding the

condition, or health, of rivers and their adjacent landscapes, and to diagnose causes of

degradation. Based on this information, we can develop restoration plans, estimate the

ecological risks associated with land use plans in a watershed, or select pre-existing,

alternative development options to minimise river degradation.

Watershed management and catchment scale studies have become increasingly more

important in determining the impact of human development on water quality both within the


10

watershed, as well as that of receiving waters. Although these studies have become more

common in the past 20 years, they still leave many questions unanswered. For example, there

is a dispute regarding whether land use of the entire catchment, or that of the riparian zone is

more important in influencing the water quality, while all other factors remain constant

(Osborne and Wiley, 1988).

As water drains from the land surface, it carries residues from the land. Surface runoff,

especially under the first flush phenomena, is an important source of non-point source

pollution. Runoff from different types of land use may be enriched with different kinds of

contaminants. For example, runoff from agricultural lands may be enriched with nutrients and

sediments whereas runoff from highly developed urban areas may be enriched with rubber

fragments, heavy metals, as well as sodium and sulphate from road de-icers at some

locations. Moreover, through evapotranspiration, interception, infiltration, percolation and

absorption, different types and coverage of vegetative surfaces can modify the land surface

characteristics, water balance, hydrologic cycle, and the surface water temperature (LeBlanc

et al., 1997).

As a result, the quantity of water available for runoff, streamflow and groundwater flow, as

well as the physical, chemical and biological processes in the receiving water bodies can be

affected. It is therefore, conceivable that there is a strong relationship between land-use types

and the quantity and quality of water (Gburek and Folmar, 1999). Many peri-urban rivers

draining from extensive urban and agricultural areas in Australia have become highly

degraded over the past few decades and remain a sensitive issue in the agenda of river


11

management authorities (Pinto and Maheshwari, 2011). With the expansion of cities into

peri-urban areas, there has been a rapid increase in the number of sewage treatment facilities

that discharge treated effluent into peri-urban waterways. Similarly, land use patterns can

alter the quality and quantity of nutrients and sediment-rich stormwater runoff during high

rainfall events (Pinto et al., 2012). Algal blooms in Australian freshwaters cost the

community between AUD180 and AUD240 million every year (Atech, 2000). Rivers that are

severely impacted due to anthropogenic influence are said to be suffering from urban stream

syndrome (Walsh et al., 2005). Hence, prediction of water quality for river health

management and issuing short and long-term advisories on the suitability of water quality to a

wide range of river users are important (Pinto et al., 2012).

The river water quantity is also controlled by the climate (e.g. precipitation and wind). Over

the last decade, there has been a rising concern that global warming may be impacting, and

may continue to significantly impact the temperature and precipitation patterns. For example,

this was recognized by the Great Lakes Regional Assessment Team in their study of the

potential impacts of climate change in the Great Lakes region (Sousounis and Bisanz 2000).

Eutrophication can be influenced by climate, including precipitation, temperature and solar

radiation. Precipitation and temperature firstly act on water discharge, which is widely

acknowledged to be a dominant factor influencing eutrophication in river systems (Lack,

1971). The largest algal blooms always occur during periods of low flows and reduced

velocity, when the residence times are longer (Bowes et al., 2012). Furthermore, seasonal

rises in water discharge often coincide with a decline in eutrophication abundance (Lack,

1971). Air temperature also strongly influence water temperature (warmer air means warmer


12

water). Solar radiation is also a key factor for algal blooms (Whitehead and Hornberger,

1984) which is likely to vary in the future due to climate change and anthropogenic factors

(Stanhill and Cohen, 2001).

Long-term surveys and monitoring programs of water quality are an adequate approach to a

better knowledge of river hydrochemistry and pollution, but they produce large sets of data

which are often difficult to interpret (Dixon and Chiswell, 1996). Also, it is quite expensive

to monitor a river for a large number of water quality parameters. Most discussions on trend

detection focus on analysing a single variable, while routine monitoring programs ordinarily

measure several variables. The problem of data reduction and interpretation of multi-

constituent chemical and physical measurements can be approached through the application

of multivariate statistical techniques and exploratory data analysis (Massart et al., 1988;

Wenning and Erickson, 1994). The usefulness of multivariate statistical tools in the treatment

of analytical and environmental data is reflected by the increasing number of papers cited in

Analytical Chemistry Reviews bases on these techniques (Brown et al., 1994).

The identification of trends in water quality can also be used to either confirm the

effectiveness of certain management actions, or to establish a need for new management

interventions. Many water quality monitoring networks have been established in Australia

with the primary objective of detecting temporal trends in water quality to meet ANZECC

guidelines (ANZECC, 2000). Statistical tests for trend analysis provide evidence if a trend is

detected, but not the reason and hence the reason for the change/trend should be investigated

(WQA, 2013). There are many previous research studies on spatial and temporal changes in


13

water quality in river systems, such as the Han River in South Korea (Chang, 2008), the

Struma River in Bulgaria (Astel et al., 2007), the Lake Tahoe basin in the USA (Stubblefield

et al., 2007), the Amu Darya River in Central Asia (Crosa et al., 2006), the water bodies of

New Seine River in France (Meybeck, 2002) and the Frome River in the UK (Hanrahan et al.,

2003).

Traditionally the assessment of river water quality has been based solely on the measurement

of physical, chemical and some biological characteristics. While these measurements may be

efficient for regulating effluent discharges and protecting humans, they are not very useful for

large-scale management of catchments, or for assessing whether river ecosystems are being

protected. Measurements of aquatic biota, to identify structural or functional integrity of

ecosystems, have recently gained acceptance for river assessment. Empirical evidence from

studies of river ecosystems under stress suggests that a small group of biological ecosystem

level indicators can assess the river condition. However, physical and chemical features of

the environment affect these indicators, the structure and function of which may be changed

by human activities. The term ‘river health’, applied to the assessment of river conditions, is

often seen as being analogous with human health, giving many a sense of understanding.

Unfortunately, the meaning of ‘river health’ remains obscure. It is not clear what aspects of

river health sets of ecosystem-level indicators actually identify, nor how physical, chemical

and biological characteristics may be integrated into measures instead of simply being

observations of causes and effects. Increased examination of relationships between

environmental variables that affect aquatic biota, such as habitat structure, flow regime,

energy sources, water quality and biotic interactions and biological conditions, are required in


14

the study of river health (Norris et al, 1999).

To assess the health of freshwater for biotic species and humans, various guidelines have

been developed internationally e.g. IUCN Global Freshwater Initiative (International Union

for Conservation of Nature and Natural Resources), Healthy Watershed Initiative in the US

(Young and Sanzone, 2002), Pressure, State, Response model in Australia (Commonwealth

of Australia, 1996) and EU Water Framework Directive in Europe (Kaika, 2003), and

Australian and New Zealand Guidelines for Fresh and Marine Water Quality (ANZECC)

(ANZECC, 2000). One of the long-term goals of the monitoring program of stream water

quality is to detect changes or trends in pollution levels over time and to identify, describe,

and explain the major factors affecting such trends, and to devise a strategy to improve the

overall water quality of a river system (Yu et al., 1993).

2.2 HAWKESBURY-NEPEAN RIVER SYSTEM

Sydney is the most populous city in Australia with a population of over 4.5 million. The

surroundings of Sydney are highly urbanized as compared to the rest of Australia due to

continued, high residential developments in the region over the past several decades.

Populations have moved away from city centres to the peri-urban surrounding areas at higher

rates, resulting in exponential increases in commercial and residential developments. Even

though the chemical composition of fresh surface water in the Hawkesbury-Nepean River

System (HNRS), located in New South Wales (NSW), Australia, has been extensively

studied over the past 20 years. Majority of the monitored data is strongly biased towards pH,


15

conductivity, turbidity, dissolved oxygen, major ions and nutrients. These data are routinely

monitored by state government authorities to provide information on the quality of Sydney’s

potable water supply and sewage treatment plants. The full potential of data has not been

used to examine the long-term spatial trends in the chemistry of the freshwater reaches of the

HNRS (Markich and Brown, 1998).

The Hawkesbury-Nepean River System (HNRS) is the main source of fresh drinking water

supply to more than 4.8 million people living in, and around, Sydney. The HNRS system is a

combination of two major rivers (Figure 2.1): the Nepean River (155 km) and the

Hawkesbury River (145 km) (Markich and Brown, 1998). The river system is complex in

nature; the upper part contains poorly accessible gorges, the middle part is running through

irrigated farm lands and the lower part has tidal slopes with deposited soil pockets (Diamond,

2004). The middle part of the river is being continuously influenced by increasing population

growth, urbanization, industrialization and other human activities which cause contamination

of the quality of the river water from different sources (e.g. sewage, stormwater, runoff from

disused mines, toxic forms of blue-green algae, and waste from domestic and native animals).

Pinto and Maheshwari (2011) have shown that river health in peri-urban landscapes are prone

to higher degrees of degradation. Within the HNR catchment, vegetation clearance has been

continuously practised over the last 200 years causing increased subsurface and agricultural

runoff and sediment loads into the river system (Thomas et al. 2000). Land use in the HNR

catchment includes regions that are heavily peri-urbanised and industrialised, and which are

important for recreational and agricultural activities and tourism (Baginska et al. 2003, Pinto

and Maheshwari, 2010). Agricultural runoff contributes approximately 40% to 50% of


16

phosphorus loads and 25% of nitrate loads into the HNRS which are believed to have

originated from agricultural and animal farms (Markich and Brown 1998).

Figure 2.1: Land use in Hawksbury-Nepean catchment (BOM, 2013).

This river system has been subjected to multiple disturbances since European settlement,

including extensive clearing of over 37% of the catchment for agriculture, urban and

industrial land use, nutrient enrichment associated with sewage, urban runoff and wastewater

disposal, extractive industries, regulation and diversion of river flows, and mining (Gehrke


17

and Harris, 1996).

Unlike other natural rivers where flow is dominated by rainfall events, the flow of HNRS is

highly regulated by impoundments and treated effluent discharges from sewage treatment

plants. There are about 22 dams and 15 weirs situated along the HNRS. The major dam on

this river is at Warragamba, which holds about 2.057 × 109 km3 of water, captured from a

9000 km2 catchment area (Turner and Erskine, 2005).

There are 18 sewage treatment plants along the HNRS discharging significant volumes of

treated municipal wastewater into the river. The river system has been increasingly regulated

since completion of the first diversion weirs in 1888, with the largest dam, Warragamba

Dam, completed in1960 to provide the main water supply for the Sydney metropolitan area.

Twenty-nine dams of 7m or more in height, and another 52 smaller structures, now regulate

flows in the river system (Marsden and Gehrke, 1996).

Concern about the ability of the river to support the increasing water demands of a growing

urban population has led to the enactment of legislation in New South Wales requiring the

Sydney Water Corporation to protect the aquatic environment by conducting its activities in

an ecologically sustainable manner (WBC, 1994).

The above review highlights the complex nature of the land use and water quality interaction

of the HNRS and it underlines the importance of assessing the water quality of this important

river system, which is the focus of this study.

CHAPTER 03: Description of Methods

18

CHAPTER 3

DESCRIPTION OF METHODS

3

3.1 OVERVIEW

A regular water quality monitoring program generates reliable data which reflects the

state of the water quality of a river. However, generating good data is not enough to

meet the objectives of a water quality monitoring program; data must be processed and

presented in a manner that provides the understanding of the spatial and temporal

patterns in water quality parameters. The intent is to use a collected set of data to

explain the current state of the water more widely and make necessary controls to

overcome future water quality issues. Water quality data usually exhibit the following

characteristics: non-normal distribution, presence of outliers, missing values, values

below detection limits (censored), and serial dependence. It is essential to apply an

appropriate, statistical methodology when analysing water quality data to draw valid

conclusions, and hence it provides useful advices in water management. This chapter

presents a detailed description of statistical methods used in this research to assess the

water quality in the Hawkesbury-Nepean river system.

Different forms of graphs have been used to provide visual summaries of data quickly

and clearly to describe important information contained in the data, and provide insight

into the data. Graphs help to determine if more complicated modelling is necessary.


19

Three particularly useful graphical methods are boxplots, scatter plots, and Q-Q plots.

In this study, box plots have been used for exploratory data analysis as it provides

summaries of a dataset.

Water quality monitoring programs generate complex multidimensional data.

Multivariate statistical techniques need to be used to extract useful information from

this data. In this study, factor analysis and principal component analysis have been

performed to identify the most significant water quality monitoring stations, and water

quality parameters in the HNRS.

The rank based non-parametric Mann–Kendall (MK) statistical test has been used to

assess the trend in the water quality time series data as these tests are more suitable for

non-normally distributed data and censored data which are frequently encountered in

hydro-meteorological time series (Yue et al., 2002). For the MK test, data is not needed

to conform to any particular distribution and moreover, it has less sensitivity to data

gaps (Tabari et al., 2011).

Pearson correlation coefficients have been used to examine the correlations among

various pollutants. Multiple linear regression technique has been used to develop the

prediction equations for water quality parameters such as Chlorophyll-a, total

phosphorous and total nitrogen (which are difficult to measure) as a function of easily

measurable water quality parameters. The plots of standardized residuals are examined

and coefficient of determination (R2) and standard error of estimates are used to assess

the adequacy of the developed prediction equations.


20

Water quality index method has been used to compare water quality parameters with

respective regulatory standards, which gives a single indicator to describe the overall

quality of a water body (Boyacioglu, 2010).

3.2 PRELIMINARY DATA ANALYSIS – BOXPLOTS

A boxplot is a very useful and convenient tool to provide summaries of a dataset and is

often used in exploratory data analysis. A boxplot usually presents a dataset through

five numbers: extreme values (minimum and maximum values), median (50th

percentile), 25th

percentile, and 75th percentile. It also indicates the degree of

dispersion, the degree of skew and unusual values of the data (outliers). Furthermore,

boxplots can display differences between different populations without making any

assumptions of the underlying statistical distribution. Figure 3.1 illustrates the

components of a default boxplot.

Figure 3.1. Components of a default boxplot.


21

3.3 PRINCIPAL COMPONENT ANALYSIS AND FACTOR ANALYSIS

The multidimensionality (i.e. different sampling stations and different parameters over

time) of the data makes analysis more complicated. Principal component analysis

(PCA) and factor analysis (FA) are the two multivariate techniques with the central aim

of reducing as much laity of a multivariate data set as much of possible, while still

retaining their variation/useful information as much as possible. This objective is

achieved by transforming the original variables to a new set of hypothetical variables,

called principal components or factors (PC/F) that are uncorrelated. They are obtained

as a linear combination of the original variables. Principal components or factors

explain the original variance in a monotonically decreasing way (Kovács et al., 2012).

Factor analysis (F-A) is similar to principal component analysis, but the two are not

identical. In F-A, components extracted from PCA are rotated according to a

mathematically established rule (i.e., varimax, equamax and quarimax) yielding easily

interpretable new variables, called varifactors (VFs) (Pinto et al., 2013). F-A uses

regression modelling techniques to test hypotheses producing error terms, while PCA is

a descriptive statistical technique (Bartholomew et al., 2008). The difference between

PCs obtained in PCA and VFs obtained in F-A is that PCs are linear combinations of

observable water quality parameters but VF are unobservable, hypothetical and latent

variables (Alberto et al., 2001).

The differences between PCA and F-A are further illustrated by Suhr (2009) as follows:


22

PCA results in principal components that account for a maximal amount of variance

for observed variables. F-A accounts for common variance in the data. PCA inserts

ones on the diagonals of the correlation matrix. F-A adjusts the diagonals of the

correlation matrix with the unique factors.

PCA minimizes the sum of squared perpendicular distance to the component axis.

F-A estimates factors which influence responses on observed variables.

The component scores in PCA represent a linear combination of the observed

variables weighted by eigenvectors. The observed variables in F-A are linear

combinations of the underlying and unique factors.

In this study, PCA was performed first to identify the most important water quality

monitoring station(s) in the HNRS. For the purpose of this analysis, the median value

of each parameter was used, as the median is better suited for a skewed distribution to

describe the central tendency of the data. In this analysis stations with correlation

coefficient greater than 0.9 were taken as principal water quality monitoring stations.

Equations for principal components were derived by considering the loadings of the

variables (water quality monitoring stations).

An F-A was employed to further identify the monitoring stations that are important in

revealing surface water quality variations. Varimax rotation was selected as the data

rotation method, as it makes an orthogonal rotation of the factor axes to maximize the

variance of the squared loadings of a factor on all the variables in a factor matrix which

http://en.wikipedia.org/wiki/Eigenvalues_and_eigenvectors


23

has the effect of differentiating the original variables by extracted factors. Each factor

has either large or small loadings of any particular variable. A varimax solution was

used to identify each variable with a single factor. This is the most common rotation

option used in PCA and F-A. However, the orthogonality (i.e., independence) of factors

is often an unrealistic assumption (Russell, 2002). In the second step, PCA was

performed on water quality data to identify the principal components that explain most

of the variance in the water quality data set.

3.4 MANN–KENDALL STATISTICAL TEST AND SEN’S SLOPE ANALYSIS

The MK test is based on the test statistics defined as follows (equation 3.1):

3.1

Where sgn(Ө) is taken as equation 3.2:

3.2

Where xi and xj are the sequential data values, n is the length of the data set, and E(S)

and V(S) are as follows (equations 3.3 and 3.4):

3.3

3.4


24

Where it is the number of ties of extent i. The standard test statistics Z is computed by

equation 3.5.

3.5

Positive values of Z indicate increasing trends while negative values indicate decreasing

trends. When testing either increasing or decreasing monotonic trends at an α

significance level, the null hypothesis is rejected for absolute values of Z greater than Z

(1-α/2), obtained from the standard normal cumulative distribution table (Tabari and

Ahmadi, 2011).

Sen’s method uses a liner model to estimate the slope of the trend and variance of the

residuals should remain constant over time (Drapela and Drapelpva, 2011). If a linear

trend is present in a time series, the true slope (change per unit time) can be estimated

by using a simple nonparametric procedure (Sen, 1968; Drapela and Drapelpva, 2011).

This liner model )t(f can be described as follows (equation 3.6):

BQt)t(f 3.6

Where Q is the slope and B is a constant.

Slopes of all data pairs are calculated and the median value is taken as the Sen’s slope

(equation 3.7):


25

kj

xxQ

kj

i

3.7

3.5 REGRESSION ANALYSIS

Regression analysis generates an equation to describe the statistical relationship

between one or more predictors and the response variable, and to predict new

observations. Regression generally uses the ordinary least squares method, which

derives the equation by minimizing the sum of the squared residuals.

Regression results indicate the direction, size, and statistical significance of the

relationship between a predictor and response.

Sign of each coefficient indicates the direction of the relationship.

Coefficients represent the mean change in the response for one unit of change in the

predictor while holding other predictors in the model constant.

P value for each coefficient tests the null hypothesis that the coefficient is equal to

zero (no effect). Therefore, low p-values suggest the predictor is a meaningful

addition to your model.

The equation predicts new observations for given specified predictor values.

The aim of regression analysis is to construct mathematical models which describe or

explain relationships that may exist between variables (Draper and Smith, 1981).

Pearson correlation coefficients are used in this study to examine the correlations


26

among various pollutants. Multiple linear regression technique is used to develop the

prediction equations for Chlorophyll-a, total phosphorous and total nitrogen as a

function of easily measurable water quality parameters. The plots of standardized

residuals are examined and the coefficient of determination (R2) and the standard error

of estimates are used to assess the adequacy of the developed prediction equations.

3.6 WATER QUALITY INDEX (WQI) METHOD

First studies on WQI were done in 1848 in Germany (Sarkar and Abbasi, 2006; Lumb

et al., 2011); Horton (1965) developed the first WQI based on 8 water quality

parameters. Dede et al. (2013), used 5 WQI methods (Oregon WQI, Aquatic toxicity

index, overall index of pollution, universal water quality index and CCME WQI) to

evaluate surface water quality and concluded that CCME WQI is the only method that

allows utilization of all the available parameters in the calculation of overall index

value.

WQI can be used for tracking changes at one site over time, and for comparisons

among sites in a river. It was simply developed to provide a broad overview of

environmental performance (Khan et al., 2004).

Though the WQI provides a meaningful summarization of the quality of a water body,

it is not a substitute for detailed analysis of water quality data and should not be used as

a sole tool for management of water bodies (Al-Janabi et al., 2012).


27

The Canadian Council of Ministers of the Environment (CCME) Water Quality Index

is based on a formula developed by the British Columbia Ministry of Environment,

Lands and Parks and modified by Alberta Environment. The Index incorporates three

elements:

Scope (F1)- the number of variables not meeting water quality objectives;

Frequency (F2) - the number of times these objectives are not met;

Amplitude (F3) - the amount by which the objectives are not met.

Scope (F1) - Scope assesses the extent of water quality guideline non-compliance over

the time period of interest, which means the number of parameters whose objective

limits are not met. It has been adopted directly from the British Columbia Water

Quality Index:

𝐹1 =𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑓𝑎𝑖𝑙𝑒𝑑 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠

𝑇𝑎𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠 × 100 3.8

Where, the variables indicate those water quality parameters whose objective values

(threshold limits) are specified and observed values at the sampling sites are available

for the index calculation.

Frequency (F2) - The frequency (i.e. how many occasions the tested or observed values

were off the acceptable limits) with which the objectives are not met, which represents


28

the percentage of individual tests that do not meet the objectives (“failed tests”):

𝐹2 =𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑓𝑎𝑖𝑙𝑒𝑑 𝑡𝑒𝑠𝑡𝑠

𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒𝑠 × 100 3.9

The formulation of this factor is drawn directly from the British Columbia Water

Quality Index.

Amplitude (F3) - The amount by which the objectives are not met (amplitude)

represents the amount by which the failed test values do not meet their objectives, and

is calculated in three steps. The number of times by which an individual concentration

is greater than (or less than, when the objective is a minimum) the objective is termed

as an “excursion” and is expressed as follows. When the test value must not exceed the

objective:

𝑒𝑥𝑐𝑢𝑟𝑠𝑖𝑜𝑛𝑖 = (𝐹𝑎𝑖𝑙𝑑 𝑡𝑒𝑠𝑡 𝑣𝑎𝑙𝑢𝑒𝑖𝑂𝑏𝑗𝑒𝑐𝑡𝑖𝑣𝑒𝑗

) − 1 3.10

For the cases in which the test value must not fall below the objective:

𝑒𝑥𝑐𝑢𝑟𝑠𝑖𝑜𝑛𝑖 = (𝑂𝑏𝑗𝑒𝑐𝑡𝑖𝑣𝑒𝑗

𝐹𝑎𝑖𝑙𝑑 𝑡𝑒𝑠𝑡 𝑣𝑎𝑙𝑢𝑒𝑖) − 1 3.11

The collective amount, by which the individual tests are out of compliance, is

calculated summing the excursions of individual tests from their objectives and then

dividing the sum by the total number of tests. This variable, referred to as the

normalized sum of excursions (nse) is calculated as:


29

𝑛𝑠𝑒 =∑ 𝑒𝑥𝑐𝑢𝑟𝑠𝑖𝑜𝑛𝑖𝑛𝑖=1

𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑡𝑒𝑠𝑡𝑠 3.12

3F is then calculated by an asymptotic function that scales the normalized sum of the

excursions from objectives ( nse ) to yield a value between 0 and 100.

𝐹3 = (𝑛𝑠𝑒

0.01𝑛𝑠𝑒 + 0.01) 3.13

The CWQI is finally calculated as:

𝐶𝑊𝑄𝐼 = 100 −

(

√𝐹1

2 + 𝐹22 + 𝐹3

2

1.732

)

3.14

The factor of 1.732 has been introduced to scale the index from 0 to 100. Since the

individual index factors can range as high as 100, it means that the vector length can

reach a maximum of 173.2 as shown below:

√1002 + 1002 + 1002 = √30000 = 173.2 3.15

The index produces a number between 0 (worst water quality) and 100 (best water

quality). These numbers are divided into 5 descriptive categories to simplify the

presentation:

Excellent: (CCME WQI Value 95-100) – water quality is protected with a virtual

absence of threat or impairment; conditions very close to natural or pristine levels.

Good: (CCME WQI Value 80-94) – water quality is protected with only a minor

degree of threat or impairment; conditions rarely depart from natural or desirable


30

levels.

Fair: (CCME WQI Value 65-79) – water quality is usually protected but

occasionally threatened or impaired; conditions sometimes depart from natural or

desirable levels.

Marginal: (CCME WQI Value 45-64) – water quality is frequently threatened or

impaired; conditions often depart from natural or desirable levels.

Poor: (CCME WQI Value 0-44) – water quality is almost always threatened or

impaired; conditions usually depart from natural or desirable levels.

3.7 CHAPTER SUMMARY

The discretion of methods used to assess the water quality in the Hawkesbury Nepean

River system (HNRS) have been presented in this chapter. It describes the use of

boxplots as the preliminary data analysis tool to identify the distribution of the water

quality data. Thereafter, the use of principal component analysis (PCA) and factor

analysis (FA) is presented to identify the most significant water quality monitoring

stations in the Hawkesbury-Nepean River System. The mathematical formulation of

rank-based non-parametric Mann–Kendall (MK) statistical test and Sen’s slope analysis

have been presented, which were used to assess the trend in the water quality time

series data. With the aim of developing prediction equations for complex water quality

parameters, regression analysis was performed; the description of the method is


31

presented in this chapter. Finally, it describes the water quality index method, which

was used to identify and assess deteriorated sections in the Hawkesbury-Nepean River

System (HNRS). In addition, it identifies the water quality parameters contributing to

poor water quality and tracks changes in water quality at different sites over time.

CHAPTER 04: Study Area and Data

32

CHAPTER 4

THE STUDY AREA AND DATA

4

4.1 OVERVIEW

The water quality data and land data for this research project have been collected from

the Hawkesbury-Nepean River System (HNRS) and its catchment area. HNRS is the

main source of fresh drinking water supply to more than 4.8 million people living in,

and around Sydney. The HNRS system is a combination of two major rivers: the

Nepean River (155 km) and the Hawkesbury River (145 km). At present, the HNRS is

under increasing pressures from peri-urbanisation and industrialization. This chapter

provides a description of land use in the Hawkesbury Nepean River catchment,

information on treated waste water discharge to HNRS, water quality parameters used

for this analysis and the water sampling and testings.

4.2 DESCRIPTION OF LAND USE IN HAWKESBURY NEPEAN RIVER

CATCHMENT AND INFORMATION ON TREATED WASTE WATER

DISCHARGE TO HNRS

More than 1.3 billion litres of wastewater is collected daily and treated by Sydney

Water, following strict license conditions issued by the NSW Environment Protection

Authority (EPA), before it is re-used or discharged into rivers.

The wastewater transported to water recycling plants goes through many treatment


33

steps including filtration and disinfection to remove nearly all biodegradable organic

material and nutrients. A schematic diagram of the HNRS with the land use details is

presented in Figure 4.1 and Table 4.1.


34

Figure 4.1. Schematic diagram of the HNRS with land use details.


35

Table 4.1: Sewage treatment plants (STP) along the HWNRS

STP Treatment level Completed date

Discharge

(ML/day) Discharge location

Picton Tertiary (includes additional Phosphorus

removal and disinfection)

30/06/2009 1.5 Re-used on-site for agricultural irrigation

Precautionary discharge to Stone Quarry

Creek

West

Camden

Tertiary (includes additional Phosphorus


30/06/2006 10.7 Re-used at Agricultural Institute. Remainder

discharged via Matahill Creek to the Nepean

River

Wallacia Tertiary (includes additional phosphorus

and nitrogen removal and disinfection)

30/06/2009 0.8 Warragamba River

Penrith Tertiary (includes additional Phosphorus

and Nitrogen removal and disinfection)

30/06/2009 22.4 Re-used locally. Remainder transferred to St

Marys Advanced Water Treatment Plant.

Some excess discharged to Boundary Creek

St. Marys Tertiary (includes ultrafiltration, reverse

osmosis, de-carbonation, additional

Phosphorous and Nitrogen removal and

disinfection)

30/06/2009 33.5 Re-used locally and at Dunheved. Remainder

discharged to Nepean River. Some excess

discharged to South Creek

Winmalee Tertiary(includes additional phosphorus 30/06/2009 16.5 Unnamed creek to the Nepean River


36

STP Treatment level Completed date

Discharge

(ML/day) Discharge location

and nitrogen removal and disinfection)

North

Richmond

Tertiary (includes additional phosphorus


30/06/2009 0.9 Redbank Creek to the Hawkesbury River

Riverstone Tertiary(includes additional phosphorus


30/06/2009 1.8 Eastern Creek to South Creek

Quakers Hill Tertiary (includes additional Phosphorus


30/06/1974 31.1 Re-used locally and at Ashlar Golf Course.

Remainder transferred to St Marys Advanced

Water Treatment Plant. Some excess

discharged to Breakfast Creek

Rouse Hill Tertiary (includes additional Phosphorus


also includes ultra-violet irradiation and

super-chlorination for reuse water

30/06/2009 15.3 Recycled back to households for non-drinking

use. Excess discharged to Second Ponds Creek

via wetlands to Cattai Creek

Castle Hill Tertiary (includes additional Phosphorus


30/06/2009 6.5 Cattai Creek


37

4.3 DATA REQUIREMENTS

This research required long term water quality data, rainfall data and land use data from

the catchment. Water quality data has been collected in-house laboratory testing and

from Sydney Catchment Authority, as well as by field and laboratory testing for a

period of one year. Rainfall data has been obtained from the Australian Bureau of

Meteorology. Land use data has been collected from NSW government departments

and other available sources.

The locations of the selected water quality monitoring stations are presented in Table

4.2. Figure 4.2 and Table 4.3 illustrates various water quality parameters examined in

this preliminary assessment.


38

Table 4.2: Water quality monitoring stations used in the preliminary assessment

Site code Site Longitudes Latitudes

N92 Nepean River at Maldon Weir upstream of Stone quarry Creek and Picton Sewage Treatment Plant 150.62 -34.2

N75 Nepean River at Sharpes Weir downstream of Matahil Creek and Camden Sewage Treatment Plant 150.67 -34.03

N67 Nepean River at Wallacia Bridge upstream of Warragamba River 150.63 -33.86

N57 Nepean River at Penrith Weir upstream of Boundary Creek and Penrith Sewage Treatment Plant 150.68 -33.74

N44 Nepean River at Yarramundi Bridge upstream of Grose River 150.69 -33.61

N42 Hawkesbury River at North Richmond upstream of North Richmond Water Treatments Works 150.71 -33.59

N35 Hawkesbury River at Wilberforce upstream of Cattai Creek 150.83 -33.58

N21 Hawkesbury River at Lower Portland upstream of Colo River 150.88 -33.43

N14 Hawkesbury River at Wisemans Ferry downstream of Car Ferry 150.98 -33.38

http://wqdb.sca.nsw.gov.au:8800/perl-bin/wqsys/site_sum_rep.pl?site=N57&sdes=Nepean%20River%20@%20Penrith%20Weir

http://wqdb.sca.nsw.gov.au:8800/perl-bin/wqsys/site_sum_rep.pl?site=N44&sdes=Nepean%20River%20@%20Yarramundi

http://wqdb.sca.nsw.gov.au:8800/perl-bin/wqsys/site_sum_rep.pl?site=N42&sdes=Hawkesbury%20River%20@%20Nth%20Richmond

http://wqdb.sca.nsw.gov.au:8800/perl-bin/wqsys/site_sum_rep.pl?site=N21&sdes=Hawkesbury%20River%20@%20Lower%20Portland

http://wqdb.sca.nsw.gov.au:8800/perl-bin/wqsys/site_sum_rep.pl?site=N14&sdes=Hawkesbury%20River%20@%20Wisemans%20Ferry


39

Figure 4.2. Locations of the 9 sampling stations adopted in the preliminary assessment

(Reproduced from: http://www.lahistoriaconmapas.com/atlas/map-river/Cook-Islands-

river-map.htm).


40

Table 4.3: Water quality parameters considered in the preliminary assessment

Water quality

Parameter

Abbreviation Units Min Max Median

pH PH 5.78 9.94 7.63

Lorenzen LOR ug/L 0.10 539.90 4.40

Iron Total TI mg/L 0.04 5.62 0.29

Phaeophytin PHA ug/L 0.10 25.20 0.80

Nitrogen TKN TKN mg/L 0.02 5.40 0.27

Temperature TEMP Deg C 8.10 30.60 19.50

Chlorophyll-a CHLA ug/L 0.20 253.10 5.10

E. coli ECOL orgs/10

0mL

0.00 6100.00 13.00

Iron Filtered FI mg/L 0.01 3.43 0.09

True Colour TCOL 1.00 93.00 11.00

Nitrogen Total TN mg/L 0.08 5.90 0.45

Turbidity TUR NTU -0.60 380.00 3.85

Alkalinity ALK mgCaC

O3/L

1.00 298.00 40.00

Aluminium Total TA mg/L 0.01 3.97 0.08

Manganese Total TM mg/L 0.00 0.48 0.03

Dissolved Oxygen DO mg/L 1.50 16.20 9.10

Enterococci ECOCC cfu/100

mL

0.00 8400.00 20.00

Phosphorus Total TP mg/L 0.01 0.18 0.01

Suspended Solids SS mg/L 1.00 105.00 3.00

Nitrogen Oxidised NO mg/L 0.00 5.00 0.17

Aluminium Filtered FA mg/L 0.00 0.45 0.01

Manganese Filtered FM mg/L 0.00 0.35 0.01

Conductivity Field EC mS/cm 0.01 48.40 0.30

Nitrogen Ammonical NH-N mg/L 0.01 0.41 0.01

Phosphorus Filterable FP mg/L 0.00 0.11 0.01

Silicate Reactive RS SiO2

mg/L

0.01 14.90 1.71

Dissolved Organic

Carbon

DOC mg/L 0.20 350.00 4.60

UV Absorbing

constituents

UV 0.01 0.93 0.12


41

4.4 WATER SAMPLING AND TESTING

In addition to the water quality data obtained from Sydney Water, as a part of this

study, water samples were also collected from selected sampling stations fortnightly for

a period of one year.

4.4.1 Location Selection and Characterisation

Samples were collected from three locations along the HNRS. The selection of the

locations was governed by three factors. Firstly, these locations are largely exposed to

impacts from extensive agricultural and urban activities. Secondly, pre-established river

management authority monitoring stations were found in close vicinity of these

locations, providing access to existing water quality data sets if required. Thirdly, the

locations were easily accessible by a boat.

4.5 SAMPLING LOCATIONS

River water samples were collected fortnightly from three sampling stations, and

testing for different water quality parameters was done following the standard methods.

Sampling stations are presented in Figure 4.3 and Figure 4.4.


42

Figure 4.3. Locations of sampling stations in the HNRS (Reproduced from: Google maps).

S1

S3

S2


43

S1 - Blaxland Crossing S2 - M4 S3 - Weir Reserve

Figure 4.4. Sampling stations.

The digital water quality multi probes (HACH HQ 40D) were utilised to obtain the

measurements of temperature (measured in degrees Celsius), pH, dissolved oxygen

(DO measured in milligrams per litre) and electrical conductivity (EC measured in

micro Siemens per centimetre at 250C). Turbidity was measured using HACH 2000NT

turbid meter. From each sub-site, 1 L of water sample was collected in an acid rinsed,

high-density polyethylene bottle (HDPE) for laboratory analysis. Ammoniacal nitrogen

(NH3 -N) and Nitrogen Oxides (NOx) were measured in the laboratory. The Gallery

(Thermo Scientific), a high precision, chemistry automated analyser, was adopted for

measuring NH3 -N, nitrite and NOx concentrations. It is a fully automated instrument

that provides analyses on optical multi-cell cuvette which provides a discrete analysis.

NH3 -N included free ammonia, ammonium and ammonia associated with chloramine

determined by using colorimetric method. Available ammonia reacts with hypochlorite

ions generated by the alkaline hydrolysis of sodium dichloroisocyanurate to form


44

mono-chloramine which reacts with salicylate ions in the presence of sodium

nitroprusside, at around pH 12.6, to form a blue compound. The compound is measured

spectrophotometrically at 660 nm. Nitrite is measured by reaction with sulphanilamide

and N-(1-naphthyl)-ethylenediamine dihydorchloride to form a highly colored azo-dye,

thus, the absorbance is measured spectrophotometrically at 540 nm or 520 nm. The

determination of nitrate is done by catalytically reducing the nitrate ions into nitrite ions

(possibly by nitrate reductive enzyme in the presence of reduced nicotinaminde

dinucleotide), the total nitrite ions are then measured by sulphanilamide method as the

NOx, and nitrate is obtained by deduction nitrite from the NOx. The analyser has the

detection limit for NH3 -N, nitrite and NOx of 0.002 mg-N/L. Standard curves for NH3

-N, nitrite and NOx were calibrated for the range 0.0 to 1.0 mg-N/L using stock

solutions of ammonium chloride, sodium nitrite and sodium nitrate, respectively. The

experimental errors were 1.5% for NH3 -N, NOx measurement. The obtained water

quality data are provided in Tables 4.4 to 4.6.


45

Table 4.4: Water quality data at Blaxland Crossing

Date pH DO

(mg/L)

Tem

(deg C)

EC

(us/cm)

Turbidity /

(NTU)

NOx

(mg/L)

NH3 -N

(mg/L)

22/02/13 7.36 8.3 25 253 10.5 0.352 0.005

08/03/13 7.32 7.9 23.6 190 9.2 0.281 0.011

22/03/13 7.27 7.2 22.5 175 9.5 0.256 0.016

05/04/13 7.32 8.2 22.6 202 6.1 0.307 0.013

19/04/13 7.23 8.6 21.4 215 7.0 0.289 0.004

03/05/13 7.08 7.9 19.2 226 6.2 0.264 0.026

17/05/13 7.01 8.8 16.6 233 8.0 0.257 0.016

31/05/13 7.21 9.5 16.5 228 7.0 0.214 0.019

14/06/13 7.34 10.4 15.2 194 12.6 0.236 0.004

28/06/13 7.48 10.9 14.4 150 38 0.256 0.018

12/07/13 7.52 9.9 14.6 162 8.6 0.254 0.012

26/07/13 7.67 10.7 13.6 175 7.2 0.266 0.004

09/08/13 7.85 11.3 15.5 206 5.7 0.307 0.005

23/08/13 7.89 10.4 16.4 216 6.4 0.298 0.012

06/09/13 7.91 9.2 20.5 263 5.2 0.275 0.014

20/09/13 7.94 8.7 21.8 290 6.3 0.268 0.004

04/10/13 7.94 10.5 21.1 338 6.1 0.277 0.01

18/10/13 7.63 8.7 22.1 328 4.2 0.192 0.034

01/11/13 7.44 7.8 23.9 309 6 0.178 0.033

15/11/13 7.46 7.5 24.6 287 6.2 0.167 0.042

24/01/14 7.51 6.8 26.7 278 5.1 0.035 0.056

07/02/14 7.32 6.1 25.4 271 5.4 0.128 0.052

21/02/14 7.21 5.8 24.8 263 5.8 0.218 0.053

07/03/14 7.53 7.9 25.5 248 3.6 0.246 0.056


46

Table 4.5: Water quality data at M4

Date pH DO

(mg/L)

Tem

(deg C)

EC

(us/cm)

Turbidity

(NTU)

NOx

(mg/L)

NH3 -N

(mg/L)

22/02/13 7.53 9.8 24.8 239 6.96 0.386 0.005

08/03/13 7.42 9.6 24.3 220 6.8 0.286 0.006

22/03/13 7.32 9.4 23.5 170 6.1 0.175 0.008

05/04/13 7.21 8.3 22.4 190 5.92 0.110 0.004

19/04/13 7.28 8.5 21.5 212 5.8 0.213 0.007

03/05/13 7.32 8.6 19.5 234 6.2 0.267 0.009

17/05/13 7.34 8.7 15.4 256 6.54 0.314 0.008

31/05/13 7.38 9.2 15.3 246 6.8 0.246 0.005

14/06/13 7.46 10.7 14.8 186 14.2 0.257 0.008

28/06/13 7.6 11.6 14.4 157 25 0.251 0.013

12/07/13 7.59 11.4 13.5 164 4.5 0.213 0.016

26/07/13 7.64 11.1 13.1 169 4.2 0.224 0.004

09/08/13 7.77 11.3 12.9 175 3.5 0.234 0.005

23/08/13 7.67 10.5 13 201 5 0.223 0.004

06/09/13 7.58 10.4 16.7 237 5.6 0.284 0.007

20/09/13 7.68 10.4 18 274 4.4 0.322 0.005

04/10/13 7.78 9.8 20.4 294 4.8 0.116 0.005

18/10/13 7.83 9.1 21.4 303 5.1 0.004 0.007

01/11/13 7.46 8.7 21.6 286 5.9 0.121 0.006

15/11/13 7.32 8.2 21.8 294 6 0.191 0.005

24/01/14 7.19 6.2 25.7 266 3.9 0.007 0.005

07/02/14 7.21 6.6 25.4 268 3.7 0.086 0.004

21/02/14 7.29 6.9 25.5 256 3.1 0.024 0.004

07/03/14 7.22 6.6 25.6 259 3.4 0.064 0.004


47

Table 4.6: Water quality data at Weir Reserve

Date pH DO

(mg/L)

Tem

(deg C)

EC

(us/cm)

Turbidity

(NTU)

NOx

(mg/L)

NH3 -N

(mg/L)

22/02/13 7.4 8.1 24.2 304 13.1 0.536 0.007

08/03/13 7.42 8.8 23.8 258 7.8 0.326 0.004

22/03/13 7.47 9 23.1 209 6.34 0.262 0.005

05/04/13 7.31 8.4 21 261 6.16 0.376 0.013

19/04/13 7.41 9.2 20.8 282 6.3 0.352 0.011

03/05/13 7.56 9.6 16.4 276 5.8 0.325 0.012

17/05/13 7.59 10.3 13.7 288 5.59 0.431 0.013

31/05/13 7.62 10.5 13.2 264 5.4 0.426 0.011

14/06/13 7.71 10.4 13.4 249 4.8 0.482 0.008

28/06/13 7.68 11.2 12.8 267 5.2 0.428 0.004

12/07/13 7.34 10.9 12.6 253 4.7 0.448 0.005

26/07/13 7.68 11.4 12.3 241 4.6 0.472 0.007

09/08/13 7.73 11.4 12 233 4.2 0.481 0.005

23/08/13 7.71 11.5 11.9 263 3.1 0.408 0.005

06/09/13 7.58 10.6 15.8 266 6.2 0.395 0.004

20/09/13 7.53 10.8 16.8 371 7.4 0.531 0.011

04/10/13 7.71 10.2 18.9 267 5.4 0.321 0.015

18/10/13 7.74 9.6 21.2 269 3.1 0.127 0.019

01/11/13 7.31 9.2 21.6 274 8.3 0.214 0.018

15/11/13 7.42 8.1 21.8 283 15 0.249 0.028

24/01/14 7.5 8 25.3 221 2.2 0.056 0.004

07/02/14 7.66 8.3 25.4 269 2.1 0.229 0.003

21/02/14 7.75 8.2 25.5 298 2.3 0.219 0.005

07/03/14 7.62 8.1 25.3 289 2.2 0.116 0.005

CHAPTER 05: Results

48

CHAPTER 5

RESULTS AND DISCUSSION ON ASSESSMENT OF RIVER WATER

QUALITY

5

5.1 OVERVIEW

This chapter presents the results of the assessment of the water quality of the Hawkesbury

Nepean River System (HNRS) using the water quality data obtained from Sydney Catchment

Authority and the data sampled as a part of this study. At the beginning, preliminary data

analyses were performed to explore the general characteristics of the water quality parameters

along the HNRS. Principal components and factor analyses were then performed to identify

the most significant water quality monitoring stations, water quality parameters and the

correlations among the water quality parameters. Thereafter, long term water quality trends

were identified by performing Mann–Kendall statistical test and Sen’s slope analysis.

Afterwards, prediction equations for various water quality parameters were developed using

multiple linear regression analysis and finally, the water quality index method was used to

make an overall assessment of the water quality in the HNRS.

5.2 PRELIMINARY WATER QUALITY DATA ANALYSIS

5.2.1 pH

Figure 5.1 presents the box plot of the observed pH values along the Hawkesbury Nepean

CHAPTER 05: Results

49

River System (HNRS). It can be seen that station N21 has the highest observed pH value

(11.40). Overall, N92 shows the highest levels of pH values, where median value has

exceeded the ANZECC trigger value (upper limit). Higher pH refers to a higher alkaline

condition, which is generally attributed to numbers of factors such as weathering of concrete,

pavement and other building materials into smaller particles, that are then washed off from

the landscape into streams. This could also be partially linked to higher algal growth in the

river. Excess alkalinity can cause ammonia toxicity and algal blooms, altering water quality

and harming aquatic life.

The lowest observed pH (4.32) can be seen at station N42. Furthermore, pH values are found

to be below the ANZECC trigger value (lower limit) for stations E851, N21, N42, N44, N57,

N641, N67 and N92. Lower pH indicates an acidic condition, which can be caused by acid

rain, leaching of surrounding acid rocks, mining activities within the catchment and certain

wastewater discharges. Low pH can allow toxic elements and compounds to become more

mobile and available for uptake by aquatic plants and animals.

The spread (i.e. standard deviation) of the measured pH values is the highest for station N92,

followed by N57, while it is the lowest for N86. The skewness is the highest for station N86,

followed by N881. Generally, skewness of pH data is very low for most of the stations.

Overall, the observed pH values mostly fall within the ANZECC guideline recommended

upper and lower limits; however, there are more cases where pH values are higher than the

ANZECC recommended trigger value (upper limit) compared with the recommended lower

limit. The worst case is seen for station N92 (Figure 5.1). The causes for observed higher and

CHAPTER 05: Results

50

lower pH values have not been specifically identified in this study.

N92

N881N8

6N8

5N7

5N6

7N6

41N64

N57

N44

N42

N35

N21

N14

E851

12

11

10

9

8

7

6

5

4

pH ANZECC upper limit

ANZECC lower limit

Figure 5.1. Box plot of pH values at different measuring stations along the Hawkesbury

Nepean River System.

5.2.2 Temperature

Figure 5.2 presents the box plot of observed temperature values along the HNRS.

Temperature does not show any outlier, and much skewness, which implies that the observed

variability might be due to seasonal variations. However, the medians and range of the

temperature values vary along the river, which can be attributed to changes in weather,

shading stream bank vegetation, impoundments, discharge of cooling water, urban

CHAPTER 05: Results

51

stormwater and groundwater inflows to the stream in different parts of the river system. The

highest range of temperature (whiskers to 8°C and 33.7°C) can be seen at station N57 and the

smallest range can be seen at station N881 (whiskers to 9.4°C and 25.4°C). Station N57 has

the highest observed temperature (33.7°C) and station N75 has the lowest (7°C). Temperature

values do not show much skewness. It is interesting to note that the river temperature does

not follow the extremes of the surrounding land temperature which exceeds 40°C and falls

below 2°C a few times in a year. Temperature mainly governs the biological activity in a

river e.g. the higher the temperature the greater the biological activity.

N92N881N86N85N75N67N641N64N57N44N42N35N21N14E851

35

30

25

20

15

10

Te

mp

era

ture

(D

eg

C)

Figure 5.2. Box plot of measured temperature along the Hawkesbury Nepean River System.

CHAPTER 05: Results

52

5.2.3 Dissolved Oxygen

Figure 5.3 presents the box plot of dissolved oxygen (DO) along the HNRS. At all the

stations, the 25th

percentile DO values are higher than the minimum ANZECC recommended

value (5mg/l) of DO, which implies a good water quality condition (in terms of organic

pollution) along the river system. In a few cases, the DO values are found to be below the

ANZECC recommended value. The low DO conditions might have been caused by higher

sewage discharge, agricultural runoff containing higher organic load and failing septic

systems in the rural parts of the catchment.

N92

N881N8

6N8

5N7

5N6

7N6

41N64

N57

N44

N42

N35

N21

N14

E851

20

15

10

5

0

Dis

so

lve

d O

xy

ge

n (

mg

/L)

ANZECC lower limit

Figure 5.3. Box plot of DO along the Hawkesbury Nepean River System.

CHAPTER 05: Results

53

It can be seen that station N92, which is the most upstream station, has the highest observed

DO (18.8 mg/l). The spread of the boxes are very similar for all the stations and no

remarkable skewness is noticed for any of the monitored stations.

5.2.4 Conductivity

Figure 5.4 and Figure 5.5 present the box plots of conductivity along the HNRS. Site N14

shows a much higher range and a median value of conductivity as compared with other

stations. It shows a highest observed reading of conductivity of 48.400 mS/cm and whiskers

of 0.009 mS/cm and 28.7 mS/cm. Also, station N21 shows a comparatively high conductivity

values. The lowest observed conductivity (0.031 mS/cm) can be seen at N57. Minimum range

of data distribution can be seen at stations N86 (whiskers to 0.08 and 0.11). Sampling

stations, N42, N57, N64, N85, N86 and N881 show comparatively better conductivity values.

Sites N14, N21, N35, N67, N75 and N92 show a higher median value exceeding the

ANZECC recommended value. Discharges to streams can change the conductivity depending

on the water chemistry. A failing sewage system would raise the conductivity because of the

presence of chloride, phosphate and nitrate. In contrast, an oil spill would lower the

conductivity. Site N14 needs to be further investigated to find the sources of pollutants which

contribute to the observed higher conductivity values. However, this was not done in this

thesis as it falls beyond its scope.

CHAPTER 05: Results

54


50

40

30

20

10

0

Co

nd

ucti

vit

y

(mS

/cm

)

Figure 5.4. Box plot of conductivity along the Hawkesbury Nepean River System for all

sampling stations (showing all the observed data range).

CHAPTER 05: Results

55

N92

N881N8

6N8

5N7

5N6

7

N641

N64N57N44N42N35N21N14

E851

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Co

nd

ucti

vit

y

(mS

/cm

)

ANZECC upper limit (0.35 mS/cm)

Figure 5.5. Box plot of conductivity along the Hawkesbury Nepean River System for all

sampling stations (with the scale up to only 3 mS/cm).

5.2.5 Turbidity

Figure 5.6 and Figure 5.7 present the box plot of the observed turbidity data along the HNRS.

It can be seen that all the median turbidity values are below the ANZECC trigger value.

Sampling stations N14, N35 and N64 show comparatively high turbidity values during the

considered period of study. Site N57 has the highest observed turbidity (437 NTU) value.

There are many outliers at all the sites above the ANZECC trigger value, which indicates that

at many instances, the turbidity values in the HNRS are too high. These high turbidity values

can occur in wet weather conditions. Turbidity often increases sharply during rainfall,

CHAPTER 05: Results

56

especially in developed watersheds, which typically have relatively high proportions of

impervious surfaces. The flow of stormwater runoff from impervious surfaces rapidly

increases stream velocity, which increases the erosion rates of stream-banks and channels,

which can increase turbidity. Turbidity can also rise sharply during dry weather if earth-

disturbing activities are occurring in or near a stream without erosion control practices in

place, where atmospheric deposition can increase the turbidity in the river. High turbidity

values can also arise from large numbers of bottom feeders and excessive algal growth.

The lowest observed turbidity of 0 NTU can be seen at some stations. When the outlier data

is ignored, sampling stations N14 and N35 show a comparatively higher range of data

distribution (N14: whiskers to 0 and 34.3 and N35: whiskers to 5.2 and 39.1), minimum

range of data distribution can be seen at N85 (whiskers to 0.77 and 5.67). At all the stations,

the turbidity values are positively skewed.

CHAPTER 05: Results

57

N85N75N67N641N64N57N44N42N35N21N14E851

500

400

300

200

100

0

Tu

rbid

ity

La

b/

Fie

ld (

NTU

)

Figure 5.6. Box plot of turbidity along the Hawkesbury Nepean River System (showing all

the observed data range).

CHAPTER 05: Results

58

N85

N75

N67

N641N6

4N5

7N4

4N4

2N3

5N2

1N1

4E8

51

100

80

60

40

20

0

Tu

rbid

ity

La

b/

Fie

ld (

NTU

)

ANZECC upper limit (20 NTU)

Figure 5.7. Box plot of turbidity along the Hawkesbury Nepean River System (with the scale

up to only 100 NTU).

5.2.6 Phosphorus

Figure 5.8 and Figure 5.9 present the box plot of total phosphorus and filterable phosphorus

along the HNRS, respectively. Stations N21, N35 and N44 show comparatively high total

phosphorus and filterable phosphorus values. High phosphorus values can occur due to both

natural and human factors. These include soil and rocks, wastewater treatment plants, runoff

from fertilized lawns and cropland, failing septic systems, runoff from animal manure storage

areas, disturbed land areas, drained wetlands, water treatment and commercial cleaning

preparations. Since phosphorus is a key nutrient in most fresh water bodies, even a modest

CHAPTER 05: Results

59

increase in phosphorus can, under the right conditions, set off a whole chain of undesirable

effects in a stream including accelerated plant growth, algae blooms, low dissolved oxygen,

and the death of certain fish, invertebrates, and other aquatic animals (Boman et. al., 2002).

It can be seen that station N35 has the highest, observed, total phosphorus value (0.380). The

lowest, observed phosphorus (0.005) can be seen at many stations. When the outliers are

overlooked, sampling station N35 shows a comparatively high range of data distribution

(whiskers to 0.005 and 0.123). The minimum range of data distribution can be seen at stations

N86 and N881 (whiskers to 0.005 and 0.014). All the total phosphorus data are positively

skewed. Except stations N14, N21, N35 and N44, the total phosphorus values at all the other

stations lie below the ANZECC trigger value.

CHAPTER 05: Results

60


0.4

0.3

0.2

0.1

0.0

Ph

osp

ho

rus T

ota

l (m

g/

L)

Figure 5.8. Box plot of total phosphorus along the HNRS (showing all the observed data

range).

CHAPTER 05: Results

61

N92

N881N8

6N8

5N7

5N6

7N6

41N64

N57

N44

N42

N35

N21

N14

E851

0.20

0.15

0.10

0.05

0.00

Ph

osp

ho

rus T

ota

l (m

g/

L)

ANZECC upper limit (0.05 mg/l)

Figure 5.9. Box plot of total phosphorus along the Hawkesbury Nepean River System (with

the scale up to only 0.2 mg/L).

The sampling station N92 has the highest, observed, filterable phosphorus value (0.237). The

lowest, observed, filterable phosphorus (0.001) can be seen at many stations. When the

outliers are overlooked, sampling station N35 shows a comparatively higher range of data

distribution (whiskers to 0.001 and 0.46). The minimum range of data distribution can be

seen at stations N881 and N881 (whiskers to 0.001 and 0.006). All the filterable phosphorus

data are positively skewed.

CHAPTER 05: Results

62

N92N881N86N85N75N641N64N57N44N42N35N21N14E851

0.25

0.20

0.15

0.10

0.05

0.00

Ph

osp

ho

rus F

ilte

rab

le (

mg

/L)

Figure 5.10. Box plot of filterable phosphorus along the Hawkesbury Nepean River System

(showing all the observed data range).

CHAPTER 05: Results

63

N92N881N86N85N75N641N64N57N44N42N35N21N14E851

0.05

0.04

0.03

0.02

0.01

0.00

Ph

osp

ho

rus F

ilte

rab

le (

mg

/L)

Figure 5.11. Box plot of filterable phosphorus along the Hawkesbury Nepean River System

(with the scale up to only 0.05 mg/L).

5.2.7 Nitrogen

Figures 5.12, 5.13, 5.14, and 5.15 present the box plots of total nitrogen (TN), nitrogen

oxidised, ammoniacal nitrogen and total kjeldahl nitrogen (TKN) along the HNRS,

respectively.

Figures 5.12 and 5.13 show that station N75 has the highest observed TN (6.73) value. The

lowest observed TN (0.01) can be seen at stations N851, N42, N85, N86 and N881. The

sampling station N75 shows a comparatively higher range of data distribution (whiskers to

0.01 and 5.9). The minimum range of data distribution can be seen at N86 (whiskers to 0.1

CHAPTER 05: Results

64

and 0.4). At all the stations, TN does not show any notable skewness. Most of the values lie

above the ANZECC trigger value. Stations N85, N86 and N881 exhibit comparatively better

values.


7

6

5

4

3

2

1

0

Nit

rog

en

To

tal (m

g/

L)

Figure 5.12. Box plot of total nitrogen along the Hawkesbury Nepean River System.

CHAPTER 05: Results

65

N92

N881N8

6N8

5N7

5N6

7N6

41N64

N57

N44

N42

N35

N21

N14

E851

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Nit

rog

en

To

tal (m

g/

L)

ANZECC upper limit (0.35mg/l)

Figure 5.13. Box plot of total nitrogen along the Hawkesbury Nepean River System.

Figure 5.14 shows that station N75 has the highest observed Nitrogen oxides (NOx) (5,900

mg/L) value. The lowest observed NOx (0.002) can be seen at some stations. When the

outliers are overlooked, sampling stations N35 and N75 show a comparatively higher range

of data distribution (N35: whiskers to 0.01 and 3.00 and N75: whiskers to 0.002 and 5.00).

The minimum range of data distribution can be seen at N881 (whiskers to 0.002 and 0.233).

The median values of NOx at stations N35, N44 and N75 are above the ANZECC trigger

value. Data does not show any notable skewness.

CHAPTER 05: Results

66

N92

N881N8

6N8

5N7

5N6

7N6

41N64

N57

N44

N42

N35

N21

N14

E851

6

5

4

3

2

1

0

Nit

rog

en

Oxid

ise

d (

mg

/L)


Figure 5.14. Box plot of nitrogen oxidised along the Hawkesbury Nepean River System.

For nitrogen ammonical (NH4-N), (Figure 5.15 and 5.16) station N75 has the highest

observed NH4-N value (1.070 mg/l). The lowest observed NH4-N (0) can be seen at N35.

When the outliers are ignored, sampling station N75 shows a comparatively higher range of

data distribution (whiskers to 0.006 and 0.13). The minimum range of data distribution can be

seen at N641 (whiskers to 0.005 and 0.03). NH4-N data is positively skewed and median

values at all the stations are below the ANZECC recommended trigger value.

CHAPTER 05: Results

67

N92N881N86N85N75N641N64N44N35

1.2

1.0

0.8

0.6

0.4

0.2

0.0

Nit

rog

en

Am

mo

nia

ca

l (m

g/

L)

Figure 5.15. Box plot of ammoniacal nitrogen along the Hawkesbury Nepean River System


CHAPTER 05: Results

68

N92N881N86N85N75N641N64N44N35

0.5

0.4

0.3

0.2

0.1

0.0

Nit

rog

en

Am

mo

nia

ca

l (m

g/

L)


Figure 5.16. Box plot of ammoniacal nitrogen along the Hawkesbury Nepean River System

(with the scale up to only 0.5 mg/L).

For nitrogen TKN, Figure 5.17 shows that station N75 has the highest observed value (5.40

mg/L). The lowest observed TKN (0) can be seen at N35. Sampling station N75 shows a

comparatively higher range of TKN values (whiskers to 0.01 and 1.00), while minimum

range of data distribution can be seen at N881 (whiskers to 0.01 and 0.3). TKN data does not

exhibit any notable skewness.

CHAPTER 05: Results

69

N92N881N86N85N75N641N64N57N44N42N35N21N14E851

6

5

4

3

2

1

0

Nit

rog

en

TK

N (

mg

/L)

Figure 5.17. Box plot of nitrogen TKN along the Hawkesbury Nepean River System.

In all the forms of nitrogen, the sampling stations N35 and N75 show fairly high median and

range. This can be due to the discharge from wastewater treatment plants, runoff from

fertilized lawns and cropland, failing on-site septic systems, runoff from animal manure

storage areas and industrial discharges that contain corrosion inhibitors. Though nitrates are

essential plant nutrients, in excess amounts they can cause significant water quality problems.

Together with phosphorus, nitrates in excess amounts can accelerate eutrophication, causing

dramatic increase in aquatic plant growth and changes in the types of plants and animals that

live in streams. This, in turn, affects dissolved oxygen, temperature, and other indicators. In

terms of water treatment, algal growth is highly undesirable as it makes the water toxic. The

CHAPTER 05: Results

70

cost of treating water with algal content is too high and hence water authorities are highly

vigilant to identify any early sign of nitrogen and phosphorous increase in raw water.

5.2.8 Alkalinity

Figure 5.18 presents the box plot of alkalinity along the HNRS. Alkalinity in streams is

influenced by surrounding rocks and soils, salts, certain plant activities, and industrial

wastewater discharges. Except stations N851, N86 and N881, alkalinity readings at all the

other stations are higher than the ANZECC trigger value. These high alkalinity values can be

due to many factors such as dissolved compounds in rain, soil, sediments, and bedrock and

by-products from biological processes in the stream. It can be seen that station N92 has the

highest observed alkalinity value (298 mg CaCO3/L). Also, it shows the highest range of data

distribution (whiskers to 1.00 and 298.00). The lowest observed alkalinity (1 mg CaCO3/L)

can be seen at many stations. The minimum range of alkalinity can be seen at N86 (whiskers

to 5.00 and 16.00)

CHAPTER 05: Results

71

N92

N881N8

6N8

5N7

5N6

7

N641N6

4N5

7N4

4N4

2N3

5N2

1N1

4E8

51

300

250

200

150

100

50

0

Alk

alin

ity

(m

gC

aC

O3

/L)

ANZECC upper value (20 mgCaCO3/l)

Figure 5.18. Box plot of alkalinity along the Hawkesbury Nepean River System.

5.2.9 Suspended solids

Figure 5.19 and Figure 5.20 present the box plot of suspended solids (SS) along the HNRS. It

can be seen that many higher concentration of SS values have been plotted as outliers. As the

turbidity and SS often increase sharply during rainfall, especially in developed watersheds,

which typically have relatively high proportions of impervious surfaces. The flow of

stormwater runoff from impervious surfaces rapidly increases stream velocities which

increase the erosion rates of stream-banks and channels. This can also rise sharply during dry

weather if earth-disturbing activities are occurring in, or near, a stream without erosion

control practices in place. It can be seen that station N67 has the highest observed SS value

CHAPTER 05: Results

72

(360 mg/L). The lowest observed SS (.05 mg/l) can be seen at N75. When the outliers are

ignored, sampling stations N14 and N35 show a comparatively higher range of data

distribution (whiskers to 1.00 and 32.00), while minimum range of data distribution can be

seen at N57 (whiskers to 1.00 and 4.00). Most of the SS values are found to be below the

ANZECC recommended trigger value.


400

300

200

100

0

Su

sp

en

de

d S

olid

s (

mg

/L)

Figure 5.19. Box plot of suspended solids along the Hawkesbury Nepean River System


CHAPTER 05: Results

73

N92

N881N8

6N8

5N7

5N6

7N6

41N64

N57

N44

N42

N35

N21

N14

E851

50

40

30

20

10

0

Su

sp

en

de

d S

olid

s (

mg

/L)

ANZECC upper limit (20 mg/l)

Figure 5.20. Box plot of suspended solids along the Hawkesbury Nepean River System (with

scale up to only 50 mg/L).

5.2.10 Algae and chlorophyll-a

Figures 5.21 to 5.24, displaying box plots of total algal count and chlorophyll-a, clearly show

the relationship between algal count and chlorophyll-a. It can be seen that the sampling

stations N21 and N35 show a comparatively higher median and a range.

When considering Figures 5.21 and 5.22, it can be seen that station N92 has the highest,

observed algal count (633,800 cells/mL). The lowest, observed algal count (314 cells/mL)

can be seen at N14. When the outliers are overlooked, sampling stations N21 and N35 show a

comparatively higher range of data distribution (N21: whiskers to 1900 and 180303 and N35:

CHAPTER 05: Results

74

whiskers to 1011 and 163269). The minimum range of data distribution can be seen at

E851(whiskers to 1744 and 12097).

N92N85N75N67N641N64N57N44N42N35N21N14E851

700000

600000

500000

400000

300000

200000

100000

0

Alg

al To

tal C

ou

nt

(ce

lls/

mL)

Figure 5.21. Box plot of algal total count along the Hawkesbury Nepean River System


CHAPTER 05: Results

75

N92N85N75N67N641N64N57N44N42N35N21N14E851

200000

150000

100000

50000

0

Alg

al To

tal C

ou

nt

(ce

lls/

mL)

Figure 5.22. Box plot of algal total count along the Hawkesbury Nepean River System (with

the scale up to only 200,000 cells/mL).

Considering Figure 5.23 and Figure 5.24, it can be seen that station N21 has the highest

observed chlorophyll-a (253.1 ug/L). The lowest observed chlorophyll-a (0 ug/L) can be seen

at N57, N85 and N92. When the outliers are overlooked, sampling stations N21 and N35

show a comparatively higher range of data distribution (N21: whiskers to 0.1 and 46.3 and

N35: whiskers to 0.2 and 49.5). Except at stations N85 and N881, chlorophyll-values are

positively skewed, and also, most of the observed values are above the ANZECC trigger

value.

CHAPTER 05: Results

76


250

200

150

100

50

0

Ch

loro

ph

yll-

a (

ug

/L)

Figure 5.23. Box plot of chlorophyll-a along the Hawkesbury Nepean River System


CHAPTER 05: Results

77

N92

N881N8

6N8

5N7

5N6

7N6

41N64

N57

N44

N42

N35

N21

N14

E851

50

40

30

20

10

0

Ch

loro

ph

yll-

a (

ug

/l)

ANZECC upper limit (5 ug/l)

Figure 5.24. Box plot of chlorophyll-a along the Hawkesbury Nepean River System (with the

scale up to only 50 ug/L).

CHAPTER 05: Results

78

5.3 RESULTS FROM PRINCIPAL COMPONENT ANALYSIS (PCA)

When 15 monitoring stations were reduced to three principal components, it explained 95.2%

of the total variance and the rest of the 12 components only accounted for 4.8%. Further, the

first, second and third components (PC 1, PC 2 and PC3) accounted for about 79.6%, 8.8%

and 6.6% of the total variance in the data set, respectively. Therefore, only the first three

principal components are focused in this thesis as they contain the bulk of the data

information.

Item PC 1 PC 2 PC 3

Eigenvalue 11.960 1.3328 0.993

Variance (%) 79.731 8.855 6.620

Cumulative variance (%) 79.731 88.586 95.206

Table 5.1: Principal components with eigenvalues > 1

Item PC 1 PC 2 PC 3

Eigenvalue 11.960 1.3328 0.993

Variance (%) 79.731 8.855 6.620

Cumulative variance (%) 79.731 88.586 95.206

The first component has almost equal loadings on all the stations (Table 5.2). Therefore, it is

a measure of overall performance of the stations. It also shows an extremely high correlation

with the stations. It accounts for 79.7% of the data variance (Table 5.1). Similarly, the second

and third components have different loadings on different stations. Hence, PC 2 and PC 3

represent a difference among the stations. Loading reflects only the relative importance of a

variable (station) within a component, and does not reflect the importance of the component

CHAPTER 05: Results

79

itself (Davis, 1986).

The results of the first PCA identify three important components that account for 95.2% of

the variance in the dataset.

Table 5.2: Component score coefficients for first three PCs (for monitoring stations)

Variable/station PC 1 PC 2 PC 3

E852 0.218 0.316 -0.386

N14 0.265 -0.221 -0.176

N21 0.243 -0.116 -0.338

N35 0.272 -0.081 0.098

N42 0.287 0.09 0.03

N44 0.249 0.094 0.489

N57 0.229 0.145 0.572

N64 0.279 -0.106 -0.197

N641 0.284 -0.031 -0.061

N67 0.278 -0.17 0.15

N75 0.274 -0.261 0.022

N85 0.284 -0.073 0.076

N86 0.234 0.489 0.006

N881 0.21 0.547 -0.207

N92 0.251 -0.374 -0.119

Table 5.3 demonstrates the rotated factor correlation coefficient (obtained from factor

analysis) for 15 water quality monitoring stations. In this study, the factor correlation

coefficient is considered to be significant if the value is greater than 0.7. This conservative

CHAPTER 05: Results

80

criterion is selected because the study area is relatively large and the HNRS is deemed to be

highly non-linear and dynamic in nature. As it can be clearly seen from Table 9.8 , water

quality monitoring stations N14, N64, N641, N67, N75, N85, N86, N881 and N92 have

coefficient values greater than 0.70, and hence these are considered to be the most important

water quality monitoring stations.

Table 5.3: Varimax rotated factor loadings (for first 5 factors)

Variable/station Factor 1 Factor 2 Factor 3 Factor 4 Factor 5

E852 0.378 0.672 0.125 -0.19 0.594

N14 0.766 0.265 0.27 -0.452 0.081

N21 0.582 0.339 0.147 -0.717 0.098

N35 0.555 0.267 0.566 -0.524 0.118

N42 0.621 0.536 0.498 -0.266 0.07

N44 0.404 0.303 0.85 -0.128 0.058

N57 0.293 0.288 0.904 -0.099 0.033

N64 0.818 0.432 0.248 -0.27 0.094

N641 0.768 0.473 0.373 -0.185 0.07

N67 0.776 0.244 0.537 -0.174 0.082

N75 0.842 0.189 0.424 -0.244 0.116

N85 0.749 0.368 0.498 -0.175 0.11

N86 0.261 0.846 0.43 -0.12 0.037

N881 0.195 0.926 0.238 -0.184 0.074

N92 0.946 0.119 0.227 -0.16 0.111

CHAPTER 05: Results

81

The results of PCA on the water quality parameters dataset give eight principal components

with eigenvalues > 1, explaining about 72.7% of the total variance in the data set. The first

PC (PC 1) accounts for 24.1% of the total variance of the data, which is highly correlated

(loading > 0.7) with total iron (TI), true color (TCOL), turbidity, aluminum total and UV

absorbent. Whereas, the other seven PCs, although account for 12.7%, 8.3%, 7.3%, 6.6%,

5.2%, 4.4% and 3.8% variances, respectively, show little correlation (loading > 0.7) with

none of the parameters (Table 5.4 and 5.5).

Principal components extracted for water quality parameters do not have a strong correlation

when comparing with principal components extracted for the water quality monitoring

stations. Monitoring stations are primarily controlled by hydrological conditions, while water

quality parameters are controlled by a combination of hydrological, chemical, physical and

biological conditions, so it is expected that the monitoring stations would have a higher

correlation than the water quality parameters.

Table 5.4: Explained variance and eigenvalues (for water parameters)

Item PC 1 PC 2 PC 3 PC 4 PC 5 PC 6 PC 7 PC 8

Eigenvalue 6.75 3.55 2.34 2.06 1.85 1.46 1.24 1.07

Variance (%) 24.13 12.71 8.36 7.38 6.61 5.21 4.44 3.83

Cumulative

variance (%) 24.13 36.84 45.20 52.59 59.20 64.41 68.85 72.69

CHAPTER 05: Results

82

The stations N14, N64, N641, N67, N75, N85, N86, N881 and N92 were found to be the

most significant sampling stations explaining the most variation in the water quality data in

the Hawkesbury-Nepean River System. This result might be used to reduce the number of

sampling stations in the river system. Principal component analysis allowed deriving three

principal components which explained more than 90% of the total variance in the data set.

The stations N14, N64, N641, N67, N75, N85, N86, N881, N92, N57 and N21were found to

be the most significant sampling stations explaining the most variation in the water quality

data in the Hawkesbury- Nepean River System. This result might be used to reduce the

number of sampling stations in the river system. Principal component analysis allowed the

derivation of three principal components which explained more than 90% of the total

variance in the data set.

CHAPTER 05: Results

83

Table 5.5: Component loadings for first eight PCs (water quality parameters)

Variable PC 1 PC 2 PC 3 PC 4 PC 5 PC 6 PC 7 PC 8

PH -0.404 0.450 -0.092 0.065 -0.148 -0.036 0.351 -0.464

LOR 0.052 0.402 0.080 -0.599 0.517 0.092 0.207 -0.112

TI 0.907 -0.133 0.080 -0.078 -0.007 0.179 -0.016 -0.030

PHA 0.124 0.378 0.032 -0.295 0.045 0.081 -0.186 0.202

TKN 0.322 0.515 -0.239 -0.102 -0.038 -0.218 0.081 0.042

TEMP 0.059 0.168 0.487 -0.315 -0.260 -0.629 0.089 0.019

CHLA 0.085 0.507 0.080 -0.630 0.502 0.089 0.141 -0.059

ECOL 0.459 0.169 0.250 0.324 -0.167 0.170 0.426 0.183

FI 0.504 -0.554 -0.220 -0.236 -0.011 0.050 0.059 -0.161

TCOL 0.754 -0.342 -0.207 -0.016 0.195 -0.296 0.061 -0.134

TN 0.172 0.618 -0.665 -0.017 -0.110 -0.063 -0.073 0.250

TUR 0.748 0.294 0.288 0.203 0.013 0.205 0.101 0.055

ALK -0.236 0.482 -0.157 -0.003 -0.424 -0.122 0.143 -0.448

TA 0.802 0.220 0.170 0.272 0.138 0.052 -0.080 -0.064

TM 0.553 -0.207 0.038 -0.535 -0.383 0.264 0.017 0.017

DO -0.243 -0.042 -0.491 0.258 0.380 0.530 0.134 -0.213

ECOCC 0.450 0.201 0.225 0.328 -0.188 0.156 0.504 0.180

TP 0.700 0.428 0.063 0.107 0.066 -0.025 -0.166 -0.087

SS 0.605 0.409 0.361 0.033 0.076 0.220 -0.295 -0.069

NO 0.061 0.510 -0.694 0.026 -0.117 0.022 -0.123 0.283

FA 0.487 -0.288 -0.198 0.208 0.332 -0.232 0.008 -0.254

FM 0.486 -0.430 -0.171 -0.415 -0.405 0.287 0.096 -0.027

EC -0.046 0.116 0.305 0.046 -0.127 0.193 -0.554 -0.219

NH-N 0.498 -0.038 -0.345 -0.222 -0.477 0.044 -0.059 -0.144

FP 0.527 0.395 -0.082 0.275 -0.110 -0.132 -0.213 -0.259

RS 0.570 -0.352 -0.272 0.106 0.165 -0.202 -0.033 0.079

DOC 0.117 0.039 -0.041 -0.014 0.092 -0.170 0.004 0.266

UV 0.742 -0.155 -0.122 0.010 0.211 -0.303 0.096 -0.036

CHAPTER 05: Results

84

5.4 LONG TERM TRENDS IN WATER QUALITY DATA

Median values of various water quality parameters are compared against the ANZECC

(2000) guidelines for fresh water. The rank-based non-parametric Mann–Kendall (MK)

statistical test is used to assess the trends in water quality parameters. The MK test is

performed at a significance level of 0.05.

The median water quality parameters and the corresponding ANZECC (2000) trigger values

are presented in Table 5.6, where the medians above the trigger values are marked in red. The

trend test results for all the water quality parameters at each station are summarised in Table

5.7, in which the detected trends are represented by arrows, with an upward arrow to indicate

an upward trend and a downward arrow for a downward trend. Dash (-) designates no

detected significant trend.

CHAPTER 05: Results

85

Table 5.6: Median values of water quality parameters and ANZECC (2000) guidelines

Variable

N14 N21 N35 N42 N44 N57 N67 N75 N92

ANZECC

trigger

values

PH 7.47 7.60 7.50 7.69 7.70 7.77 7.79 7.88 8.21 6-8

TEMP 20.90 21.10 20.90 20.50 20.90 20.70 20.50 20.70 19.05

DO 7.85 8.70 8.00 9.00 8.60 9.15 8.40 9.00 9.47 mini 5

EC 6.28 0.34 0.40 0.25 0.34 0.28 0.57 0.54 0.36 0.35

SS 12.00 8.00 12.00 2.00 2.00 2.00 4.00 3.00 2.00 20

TUR 11.00 8.31 13.60 2.42 2.70 1.76 5.13 3.80 1.40 20

TCOL 8.00 10.50 12.00 10.00 11.00 9.00 10.00 11.00 10.00 15

TN 0.40 0.42 0.82 0.50 0.70 0.35 0.61 1.20 0.43 0.350

NO 0.12 0.05 0.44 0.26 0.40 0.07 0.26 0.77 0.20 0.250

NH-N 0.01 0.01 0.02 0.01 0.02 0.01 0.01 0.02 0.01 1.000

TKN 0.27 0.30 0.40 0.24 0.34 0.25 0.36 0.44 0.25

TP 0.02 0.02 0.04 0.01 0.02 0.01 0.02 0.02 0.01 0.050

FP 0.01 0.01 0.01 0.01 0.01 0.00 0.01 0.01 0.01 20

CHLA 8.55 18.90 18.20 5.10 6.20 3.80 5.80 8.50 3.00 5

ALK 44.75 32.75 46.00 32.00 49.00 40.00 81.50 83.50 122.00 20

DOC 4.00 4.30 5.00 4.00 4.90 4.40 5.10 5.20 4.00

TI 0.43 0.34 0.53 0.28 0.18 0.19 0.24 0.19 0.15 0.300

FI 0.05 0.05 0.05 0.12 0.07 0.08 0.05 0.05 0.08

TA 0.24 0.14 0.25 0.04 0.04 0.02 0.10 0.06 0.03 0.200

FA 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.020

TM 0.04 0.04 0.06 0.03 0.04 0.03 0.07 0.03 0.02 0.100

FM 0.01 0.00 0.01 0.01 0.02 0.01 0.02 0.00 0.00

RS 1.21 0.80 1.40 2.23 1.40 1.82 1.50 1.93 1.59

ECOL 12.00 5.00 23.00 11.00 46.00 55.00 22.00 23.00 9.00

ECOC 6.00 6.00 26.00 20.00 53.00 50.00 40.00 20.00 11.00

CHAPTER 05: Results

86

Table 5.7: Mann-Kendal test results and yearly Sen’s slope

pH TEMP DO EC SS TUR TCOL TN NO NH-N TKN TP FP CHLA ALK DOC TI FI TA FA TM FM RS ECOL ECOC

N14 ↓ ↑ ↓ ↓ ↓ ↑ ↑ ↑ ↑ − − − − ↑ ↑ ↑ ↑ ↑ − − − − ↓ ↓ ↑

0.065 0.307 0.109 3.502 0.317 1.040 1.391 0.003 0.003

1.625 6.742 0.442 0.039 0.005

0.029 1.131 0.723

N21 ↓ − ↓ ↓ − ↑ ↑ − − − ↓ − − ↑ ↑ ↑ ↑ ↑ ↑ − ↓ − ↑ ↑ ↑

0.073

0.148 0.060 0.372 1.084

0.013

0.026 2.569 0.143 0.036 0.018 0.003 0.003

0.008 0.406 0.499

N35 ↓ ↑ ↓ ↓ ↑ ↑ ↑ ↓ ↓ − ↓ − − ↓ ↑ ↑ ↑ ↑ ↑ − ↑ ↑ ↑ ↑ ↑

0.078 0.130 0.317 0.023 0.801 1.438 0.998 0.109 0.081

0.018 0.751 2.954 0.325 0.062 0.016 0.013 0.008 0.008 0.239 3.000 1.009

N42 ↓ ↑ ↓ ↓ − ↑ ↑ ↓ ↓ − ↓ − − ↑ ↑ ↑ ↑ ↑ ↑ − ↑ ↑ ↑ ↑ ↑

0.047 0.143 0.060 0.023 0.484 1.084 0.049 0.047

0.008

0.497 2.785 0.122 0.042 0.018 0.013 0.003 0.003 0.122 1.856 3.206

N44 ↑ ↑ ↑ ↓ − ↑ ↑ ↓ ↓ ↓ ↓ − − ↑ ↑ ↑ ↑ ↑ ↑ − ↑ ↑ ↑ ↓ ↓

0.013 0.281 0.185 0.031 0.380 1.040 0.096 0.081 0.003 0.010 1.022 4.228 0.096 0.036 0.016 0.008 0.003 0.003 0.096 1.999 2.600

N57 ↓ − ↑ ↓ − ↑ ↑ ↑ ↑ − − − − ↑ ↑ ↑ ↑ ↑ ↑ − ↑ − ↑ ↑ ↓

0.073

0.083 0.010 0.364 0.634 0.013 0.003

0.736 2.642 0.224 0.031 0.013 0.005 0.003

0.101 12.667 0.702

N67 ↓ − − ↓ − ↑ ↑ ↓ ↑ ↓ ↓ − − ↑ ↑ − ↑ ↑ ↑ − ↑ ↑ ↑ ↑ ↑

0.065

0.070 0.983 0.650 0.039 0.003 0.003 0.029 0.679 10.517

0.055 0.016 0.010 0.005 0.003 0.096 3.588 4.703

N75 ↓ ↑ ↓ ↓ − ↑ ↑ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↓ ↑ ↑ ↑ ↑ − ↑ − ↑ ↑ ↑

0.125 0.354 0.231 0.081 0.676 0.541 0.530 0.421 0.005 0.104 0.003 0.003 0.198 12.467 0.075 0.055 0.021 0.013 0.005

0.026 6.323 3.182

N92 ↓ ↑ ↑ ↓ − ↓ ↑ ↓ ↓ ↓ ↓ − − ↓ ↑ ↑ ↑ ↑ ↑ − ↑ − ↑ ↑ ↑

0.114 0.195 0.109 0.068 0.052 1.019 0.075 0.039 0.003 0.023

0.406 25.732 0.216 0.029 0.234 0.003 0.003

0.213 1.547 2.062

CHAPTER 05: Results

87

The median values of pH at all the stations except N92 are within the ANZECC

recommended trigger values (i.e. between 6 and 8). At station N92, the pH is 2.6% above the

upper limit of the trigger value. The change in the median value of pH along the river is

presented in Figure 5.25, which shows that pH reduces from upstream to downstream of the

HNRS. This result shows an increasing acidification from upstream to downstream of the

HNRS. The pH shows a decreasing trend for all the stations except for station N44. The

maximum decrease in trend is found for station N75 (0.125 per year). The overall decreasing

trend of pH indicates an increasing acidification of water in the HNRS over the last decade.

Figure 5.25. Median values of pH along the Hawkesbury Nepean River System.

The median values of dissolved oxygen (DO) at all the 9 stations are above the ANZECC

trigger value. The DO has a decreasing trend for stations N14, N21, N35, N42 and N75. Its

maximum decreasing trend of 0.317 mg/L per year can be seen for station N35 (Figure 5.26).

CHAPTER 05: Results

88

The upstream of N35 is affected by quality and magnitude of flows coming from the South

Creek and discharge from North Richmond sewage treatment plant (STP). The dominant land

use in this part of the catchment includes rural, grazing, commercial gardening, intensive

agriculture and urban and industrial activities. These land uses can be attributed to the

decreasing trend of DO at N35. The increasing trends of DO for N92 and N57 demonstrate

the influence of natural undeveloped catchment conditions at upstream of these two stations.

Figure 5.26. Decreasing trend of DO at station N35.

The median values for electrical conductivity are higher than the trigger values for stations

N14, N35, N67, N75 and N92. In particular, the median value of station N14 is 10 times

higher than the ANZECC (2000) trigger value, which is significantly higher than any of the

0

2

4

6

8

10

12

14

N35 - Dissolved Oxygen (mg/L)

CHAPTER 05: Results

89

other stations. Figure 6 shows that the electrical conductivity value at station N14 has

decreased significantly over time and the most current results in year 2012 are much smaller

than those of 2002 to 2008. Electrical conductivity has a decreasing trend for all the stations,

with a maximum decreasing trend of 3.5 mS/cm per year at station N14 (Figure 5.27). Its

overall decreasing trend for all the 9 stations demonstrates an overall improvement of the

HNRS water quality with time, in terms of the total solids dissolved in water.

Figure 5.27. Decreasing trend of EC at station N14.

The median suspended solids (SS) are within the ANZECC (2000) trigger value for all the 9

stations. For most of the stations, SS does not show any trend; however, it has a decreasing

trend for station N14 (0.317mg/L per year) and an increasing trend for N35 (0.801 mg/L per

0

10

20

30

40

50

60

N14 - Conductivity Field (mS/cm)

CHAPTER 05: Results

90

year). The low SS levels in the river indicate that the river water is not notably polluted with

particulate matter, which is a positive aspect of the water quality of the HNRS.

The median values for turbidity are well within the ANZECC (2000) trigger value for all the

9 stations. However, it has an increasing trend at all the stations except N92. It should be

noted that station N92 is located at the most upstream part of the river among all the 9

stations. This part of the river has the lowest level of anthropogenic activity as it has the

smallest degree of urbanisation and industrialisation. As a result, it has the lowest turbidity

level (1.40 NTU) and an overall decreasing trend. The increasing trend of turbidity for the 8

out of 9 stations demonstrates the influence of increasing urbanisation and industrialisation

within the downstream parts of the catchment that has intensified over recent times.

The median values for total nitrogen (TN) are above the ANZECC (2000) trigger value for 8

stations out of 9, which are N14 (14.2%), N21 (18.5%), N35 (134.2%), N42 (42.8%), N44

(100%), N67 (74.2%), N75 (242.8%) and N92 (2.8%). Also, the TN shows an increasing

trend for stations N14 and N57. At stations N35, N42, N44, N67 and N74, the median values

for oxidised nitrogen are above the ANZECC (2000) trigger value by 76.0%, 3.6%, 60.0%,

4.0% and 208.0% respectively. It has an increasing trend for N14, N57 and N67. Ammonical

nitrogen shows a decreasing trend or no trend for all the stations. All the median values are

within the ANZECC (2000) guidelines. Nitrogen TKN has decreasing trends for all the

stations except for stations N14 and N57; the maximum slope of 0.104 mg/L can be seen for

station N75. Considering the median values, it may be stated that NOx is the main

contribution for the high value of TN. The median value of TN for N75 is 242% higher than

CHAPTER 05: Results

91

the ANZECC (2000) trigger value, which appears to be associated with the intensive

agricultural activities in the upstream catchment parts of station N75. The reduction of TN

from N75 to N67 by 49%, and from N67 to N57 by 43%, can be attributed to the natural

pristine undeveloped condition of the HNRS in between stations N75 to N67 and N57.

Furthermore, the agricultural activities at upstream of N44 has possibly increased the TN

value at N44. The overall TN levels in the HNRS are notably higher than the ANZECC

(2000) trigger value, which is likely to make the river prone to eutrophication.

The medians total phosphorus and filterable phosphorus levels are within the ANZECC

(2000) trigger value for all the stations; however, for station N35, the total phosphorus level

is very close to the trigger value (0.04 versus 0.05). No station shows a significant trend for

total phosphorus except N75, which shows a decreasing trend.

The determination of photosynthetic chlorophyll pigments and their degradation products is

one of the most frequently performed analyses in aquatic ecology (Gitelson, 1992). The

median values for chlorophyll-a are above the ANZECC (2000) trigger value for 7 stations

out of 9. Stations N92 and N57, which flow through natural undeveloped parts of the

catchment, have median values within the ANZECC (2000) trigger value. The median value

of chlorophyll-a for station N75 is 70% higher than the trigger value. It has been reduced by

68% while flowing through the natural undeveloped parts of the catchment between N75 and

N67. It is expected that water quality would be improved at N67 because of nutrient

assimilation and loss processes while traveling this section of the catchment without further

input of nutrients. The median value for chlorophyll-a has been further improved at station

CHAPTER 05: Results

92

N57 demonstrating further assimilation of nutrients while flowing through a pristine

catchment part which is largely undeveloped. The Warragamba River joins the Nepean River

in this section, carrying discharge from the Wallacia STP as well as environmental flow

release from the Warragamba dam. Nutrients that enter via Matahil Creek from the West

Camden plant and via the Warragamba River from Wallacia plant experience long residence

time and distance for assimilation, as well as dilution by low nutrient water from

Warragamba dam. When considering the median values for chlorophyll-a at stations N35 and

N21 (which are 374% and 378% higher than the ANZECC (2000) trigger value), it can be

seen that, industrialization, urban developments and agricultural activities in the catchment

have contributed in degrading the water quality. The land use at upstream parts of the

catchment of N35 predominantly includes rural, grazing and market gardening, intensive

agriculture, such as poultry farming, and both urban and industrial activities. Also, it receives

water from the South Creek tertiary treated wastewater discharges originated from three

STPs. High nutrient levels, tidal influences, high residence times and low flows make the

streams ideal for excessive algal growth and hence very high chlorophyll-a levels are noticed

at N35 and N21. Figure 5.28 presents how the median values of chlorophyll-a have changed

along the HNRS, which shows a remarkably high peak at stations N35 and N21. However, it

is a good sign that station N35 shows a downward trend for clorophyll-a.

CHAPTER 05: Results

93

Figure 5.28. Median values of chlorophyll-a along the Hawkesbury Nepean River System.

Station N14 is located just before the confluence with the Macdonald River. The water

quality of this station is influenced by flow from the Colo River and downstream of the

Hawkesbury River. The Colo River catchment is the best in terms of nutrient enrichments

among all the other sub-catchments of the HNRS because it consists primarily of pristine and

undisturbed catchment areas. About 80% of these catchments are national parks of the Blue

Mountains world heritage area. There are also limited upstream areas that support agricultural

activities. Water quality at station N14 has been improved as expected because of dilution by

high quality inflows from the Colo River and the undisturbed upstream catchment. Algae

growth, and thus chlorophyll-a level, has directly been affected by the amount of nutrients in

CHAPTER 05: Results

94

the river (e.g. Station 35 has very high chlorophyll-a level and it has the highest total

phosphorous level and the second highest total nitrogen level among the 9 stations). Low

levels of chlorophyll-a suggest a good river health; however, high levels are not necessarily

bad; it is the long-term persistence of high levels that is a problem (NLWRA, 2008). It should

also be noted that 6 out of 9 stations show an increasing trend for chlorophyll-a, indicating an

overall deterioration of water quality in the HNRS over the last decade.

The median values of alkalinity are found to be above the ANZECC trigger value for all the 9

stations, N14 (123.7%), N21 (63.7%), N35 (130%), N42 (60%), N44 (145%), N57 (100%),

N67 (307.5%), N75 (317.5%) and N92 (510%). It has an increasing trend for 8 of the

stations out of 9, with a high Sen’s slope. The maximum trend is found for station N92

(25.7mg/L per year) (Figure 5.29), which has a median value of 510% above the ANZECC

trigger value. It should be noted that station N92 is located the most upstream among all the 9

stations, and the highest level of alkalinity at this station is somewhat unexpected, which

needs further investigation (but not done in this study).

CHAPTER 05: Results

95

Figure 5.29. Increasing trend of alkalinity at station N92.

Dissolved organic carbon shows an increasing trend for 8 out of the 9 stations. Organic

carbon occurs as a result of decomposition of plant or animal materials. Total aluminium has

an increasing trend for all the stations except for N14. The median values are within the

ANZECC (2000) trigger value for all the stations except N14 and N35, which are 20% and

25% above the ANZECC (2000) trigger value, respectively. Aluminium filtered does not

show any trend for most of the stations. It was found in a study of “the water quality of

Roanoke River, Virginia”, that the sewage treatment plants were the most significant

anthropogenic contributors of aluminium to the river (Butcher, 1988). Total manganese

shows an increasing trend at all the stations except for N21 and N14. Its median values are

0

10

20

30

40

50

60

70

N92 - Alkanility (mgCaCO3)

CHAPTER 05: Results

96

within the ANZECC (2000) trigger value for all the stations. Manganese filtered shows an

increasing trend for most of the stations. Reactive silicate shows an increasing trend for all

the stations except for N14. It has the maximum increasing trend at N35 (Figure 5.30). The

ratios between silicate and phosphorous, and silicate and nitrogen largely determine which

algae would dominantly be present in the river water. Water moving over and through natural

deposits is expected to dissolve a small amount of various silicate minerals. The overall

increasing trends of aluminium, manganese and reactive silicate demonstrate the influence of

intensified land use in recent years that has occurred along the HNRS.

Figure 5.30. Increasing trends of reactive silicate at station N35.

0

1

2

3

4

5

6

7

N35 - Silicate Reactive (SiO2 mg/L)

CHAPTER 05: Results

97

Trend analysis has been done for 9 water quality parameters and 25 sampling stations. Only

dominant water quality parameters and stations of concern have been discussed in detail;

however, the trends and their significance levels are presented in Table 5.7 for all the

parameters and stations.

5.5 RESULTS FROM REGRESSION ANALYSIS FOR DEVELOPING

PREDICTION EQUATIONS FOR WATER QUALITY PARAMETERS

Pearson correlation coefficients among various water quality parameters are provided in

Table 5.8. There are a number of high correlations which are of significance, as noted below.

Nitrogen total is highly correlated with nitrogen (oxidized) (Pearson correlation coefficient, r

= 0.976), which implies that most of the nitrogen in water remains in oxidized form. Nitrogen

(oxidized) has a strong negative correlation (r = -0.716) with temperature, which implies that

nitrogen (oxidized) reduces as temperature increases. Total nitrogen (TN) is highly correlated

with conductivity (r = 0.698), which implies that dissolved minerals in water are largely of

nitrogen-based. Algal (total count) is highly correlated with suspended solids (SS) (r =

0.611). This implies that total SS in water contains a notable proportion of algae.

CHAPTER 05: Results

98

Table 5.8: Correlations among water quality parameters at station N44 of the HNRS

pH

Nitrogen TKN

Temperature

Chlorophyll-a

Nitrogen Total

Dissolved Oxygen

Phosphorus Total

Suspended Solids

Nitrogen Oxidised

Conductivity

Nitrogen Ammonia

cal

Algal Total Count

Phosphorus

Filterable

pH 1

Nitrogen TKN 0.166 1

Temperature 0.216 0.335 1

Chlorophyll-a -0.114 0.08 0.072 1

Nitrogen Total -0.138 0.182 -0.632 -0.162 1

Dissolved Oxygen 0.443 -0.222 -0.29 0.387 0.049 1

Phosphorus Total -0.255 0.268 0.219 0.355 0.263 -0.047 1

Suspended Solids -0.048 -0.108 0.007 0.552 0.402 0.226 0.303 1

Nitrogen Oxidised -0.177 -0.037 -0.716 -0.185 0.976 0.097 -0.321 -0.386 1

Conductivity 0.301 0.455 -0.223 -0.185 0.698 -0.005 -0.365 -0.382 0.608 1

Nitrogen Ammoniacal 0.049 0.472 0.113 -0.199 0.124 -0.342 0.129 -0.188 0.021 0.343 1

Algal Total Count 0.103 0.023 0.213 0.819 -0.317 0.358 0.263 0.611 -0.329 -0.180 -0.135 1

Phosphorus Filterable -0.026 0.306 0.303 -0.425 0.032 -0.361 0.219 -0.204 -0.034 0.080 0.422 -0.406 1

CHAPTER 05: Results

99

Prediction equations are developed using multiple linear regression analysis for chlorophyll-

a, total nitrogen (TN) and total phosphorous (TP). The prediction equation for chlorophyll-a

is presented by Equation 5.1. The multiple R of this equation is 0.827, coefficient of

determination (R2) is 0.683 and the standard error of estimate is 4.312. The prediction

equation for TN is expressed by Equation 5.2; the multiple R of this equation is 0.656, R2 is

0.430 and the standard error of estimate is 0.24. The prediction equation for TP is expressed

by Equation 5.3, the multiple R in this case is 0.767, R2 is 0.589 and the standard error of

estimate is 0.007. The plots of standardized residuals and predicted values of these three

equations are shown in Figures 5.31, 5.32 and 5.33, respectively. Equations 5.4.1 to 5.4.3

show quite high R2 values. The plots of standardized residuals and predicted values do not

show any trend, which indicate that the developed prediction equations satisfy the

assumptions of the least squares regression quite well.

Chl-a = 3.995 + 80.051(TP) – 556.155(FP) – 3.274(EC) + 1.594(SS) + 10.128(TKN) (5.1)

TN =1.296 + 0.940(EC) + 0.003(SS) – 0.038(temp) (5.2)

TP = 0.031 – 0.004(pH) + 0.001(temp) + 0.003(SS) + 0.015(EC) (5.3)

CHAPTER 05: Results

100

Figure 5.31. Plot of standardized residuals against estimate for Chlorophyll-a.

Figure 5.32. Plot of standardized residuals against estimate for total nitrogen.

CHAPTER 05: Results

101

Figure 5.33. Plot of standardized residuals against estimate for total phosphorous.

5.6 RESULTS OF WATER QUALITY ASSESSMENT BY USING WATER

QUALITY INDEX

For the calculation of CCME WQI, 12 water quality parameters were selected based on the

importance and the availability of data. These selected water quality parameters and

Australian and New Zealand Guidelines for Fresh and Marine Water Quality (ANZECC) are

presented in Table 5.9.

CHAPTER 05: Results

102

Table 5.9: Water quality parameters and ANZEC Guidelines for Fresh and Marine Water

Quality

Water Quality Parameter Non-compliance

if: Value1 Value 2 Unit

pH <> 6 8

Nitrogen Total > 0.35 mg/L

Phosphorous Total > 0.05 mg/L

Chlorophyll > 5 µg/L

Dissolved Oxygen < 5 mg/L

Turbidity > 20 NTU

Iron Total > 0.3 mg/L

Aluminium Total > 0.2 mg/L

True Colour > 15

Alkalinity > 20

Suspended Solids > 20

Conductivity > 0.35 mS/cm

WQIs were primarily developed for each year for the 9 sampling locations to investigate the

water quality changes along the HNRS over time. Figure 5.34 shows an improvement of

water quality over time for most of the stations. Also, it shows a marginal water quality with

WQI in between 45 and 64 for all the stations except N14 and N35, which have WQIs less

than 40 at many years.

Medians of CCME WQI values for the 21 years range from 33 to 57. All the monitoring

stations indicate marginal or poor water quality. Water quality at N21, N42, N44, N57 and

N92 is frequently threatened or impaired. WQIs at N14, N35, N67 and N75 are below 40 and

thus indicate that water quality is almost always threatened or impaired at these stations

(Figure 5.35).

CHAPTER 05: Results

103

Figure 5.34. Change in WQI over time for 9 monitoring stations in HNRS (Reproduced

from: http://www.lahistoriaconmapas.com/atlas/map-river/Cook-Islands-river-map.htm).

CHAPTER 05: Results

104

Figure 5.35. Average WQI along the HNRS.

Scope, frequency and amplitude values at the 9 monitoring stations are presented in Figure

5.36. At N35, nearly 90% of water quality values are beyond the guidelines. N35 shows the

highest frequency among the 9 monitoring stations; it also shows high amplitude (46.3). The

upstream of N35 is affected by quality and magnitude of flows coming from the South Creek

(which carries discharges from St. Marys Sewage treatment plant (STP), Riverstone STP,

Quakers Hill STP, McGraths Hill STP, and South Windsor STP) and discharge from North

Richmond STP. The dominant land use in this part of the catchment includes rural, grazing,

commercial gardening, intensive agriculture and urban and industrial activities. The low

WQI at N35 can be attributed to these land uses. .

05

1015202530354045505560

N14 N21 N35 N42 N44 N57 N67 N75 N92

WQ

I

Water quality monitoring stations

Marginal Poor

CHAPTER 05: Results

105

Figure 5.36. Scope, frequency and amplitude values at 9 monitoring stations in HNRS.

At N14, 81% of water quality data are outside the guidelines. Also, it has an amplitude of

70%. From 1993 to 2008, amplitudes are greater than 60%. Table 5.10 presents the

amplitudes at 9 stations in different years. The years with higher amplitudes (greater than

60%) are indicated in red.

Further data exploration was done at N14 as it shows the worst WQI among the 9 stations.

Table 5.11 presents details of percentage failed tests for different water quality parameter (the

total number of tests, number of failed tests, and percentage failed for each parameter for

different year). Total nitrogen, chlorophyll-a, total iron, total aluminium, alkalinity and

conductivity are the water quality parameters which exceeded the guidelines on many

occasions.

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

N14 N21 N35 N42 N44 N57 N67 N75 N92

F1

F2

F3

CHAPTER 05: Results

106

Table 5.10: Amplitudes at 9 stations in different years

Index

Period N14 N21 N35 N42 N44 N57 N67 N75 N92

2013 39.0 31.2 39.9 29.5 34.8 30.1 35.5 24.7 20.2

2012 44.4 32.1 43.4 31.4 35.4 23.0 26.4 20.6 16.8

2011 49.9 24.8 34.7 7.2 14.8 18.8 21.9 15.5 7.5

2010 50.8 40.4 35.7 17.1 18.7 22.9 25.1 24.4 22.3

2009 52.3 33.1 37.4 10.8 22.3 35.4 31.1 36.8 31.3

2008 65.9 32.0 42.6 24.9 30.6 49.7 41.0 53.4 43.5

2007 71.0 38.5 43.6 22.3 33.4 50.5 46.4 57.2 55.8

2006 81.6 43.2 45.3 15.9 22.2 41.5 47.0 60.9 50.7

2005 76.3 43.4 44.7 13.9 23.8 32.3 41.1 53.9 39.3

2004 82.9 45.2 46.6 15.2 26.7 35.3 37.0 55.4 38.9

2003 80.2 39.3 43.7 17 26.7 41.3 37.7 54.8 43.3

2002 78.7 37.7 41.4 19.2 26.3 29.7 30.6 52.8 40.1

2001 78.2 35.3 38.9 16.7 23.4 3.6 29.4 40.0 26.0

2000 87.1 35.2 41.0 16.3 22.0 6.9 25.8 49.3 30.2

1999 65.1 42.7 50.4 22.5 38.2 18.8 35.5 55.6 19.9

1998 75.0 39.9 50.7 19.5 22.8 10.6 30.6 49.2 20.7

1997 81.2 46.4 58.3 21.5 27.0 7.4 34.5 59.1 19.7

1996 69.2 39.0 54.7 20.6 30.4 9.4 33.3 60.1 13.5

1995 80.4 37.7 58.0 18.2 24.3 12.2 27.8 50.3 7.9

1994 85.3 60.1 60.9 20.1 20.7 2.0 18.0 51.4 17.8

1993 74.8 61.5 31.4 29.2 4.1 23.4 24.5 18.9

CHAPTER 05: Results

107

Table 5.11: Water quality results at N14

25 < percentage Failed < 50 = , , PercentageFailed >= 50 =

Index

Period

Number

of Tests

Number

of Failed

Tests

Percent

Failed

(%)

Number

of Tests

Number

of Failed

Tests

Percent

Failed

(%)

Number

of Tests

Number

of Failed

Tests

Percent

Failed

(%)

Number

of Tests

Number

of Failed

Tests

Percent

Failed

(%)

Number

of Tests

Number

of Failed

Tests

Percent

Failed

(%)

Number

of Tests

Number

of Failed

Tests

Percent

Failed

(%)

2013 4 3 75.0 4 2 50.0 4 3 75.0 4 4 100.0 4 0 4 3 75.0

2012 14 10 71.4 14 8 57.1 14 5 35.7 14 11 78.6 14 2 14.3 13 10 76.9

2011 13 9 69.2 13 7 53.8 13 5 38.5 13 9 69.2 13 2 15.4 12 8 66.7

2010 12 9 75.0 12 7 58.3 12 5 41.7 12 11 91.7 12 1 8.3 9 7 77.8

2009 13 8 61.5 13 6 46.2 13 3 23.1 13 13 100.0 13 1 7.7 11 8 72.7

2008 12 9 75.0 12 9 75.0 12 5 41.7 12 12 100.0 12 1 8.3 11 10 90.9

2007 12 9 75.0 12 8 66.7 12 3 25.0 12 11 91.7 12 4 33.3 13 12 92.3

2006 13 5 38.5 13 6 46.2 13 0 13 13 100.0 13 2 15.4 12 11 91.7

2005 12 7 58.3 12 6 50.0 12 2 16.7 12 12 100.0 12 4 33.3 12 12 100.0

2004 13 6 46.2 13 5 38.5 13 2 15.4 13 13 100.0 13 1 7.7 13 13 100.0

2003 13 6 46.2 13 6 46.2 13 0 13 13 100.0 13 4 30.8 12 12 100.0

2002 14 10 71.4 14 11 78.6 14 4 28.6 14 12 85.7 14 8 57.1 14 13 92.9

2001 6 5 83.3 6 5 83.3 6 1 16.7 6 6 100.0 17 8 47.1 17 17 100.0

2000 0 0 0 0 0 0 0 0 27 1 3.7 21 18 85.7

1999 0 0 0 0 0 0 0 0 26 1 3.8 12 11 91.7

1998 0 0 0 0 0 0 0 0 24 7 29.2 24 20 83.3

1997 0 0 0 0 0 0 0 0 23 3 13.0 23 23 100.0

1996 0 0 0 0 0 0 0 0 26 6 23.1 19 19 100.0

1995 0 0 0 0 0 0 0 0 26 4 15.4 26 25 96.2

1994 0 0 0 0 0 0 0 0 23 8 34.8 24 24 100.0

1993 0 0 0 0 0 0 0 0 16 1 6.3 8 8 100.0

Aluminium Total True Colour Alkalinity Suspended Solids ConductivityIron Total

CHAPTER 05: Results

108

From Table 5.11, it can be seen that water quality at N14 is poor with respect to Nitrogen,

Chlorophyll a, Iron, Aluminium and conductivity. Nitrogen is a nutrient used by plants within

natural ecosystems, with minimal leakage into surface or groundwater (Vitousek et al., 2002).

Nitrogen concentrations in streams generally increase due to discharge of sewage water,

pollutant wash off from urban and agricultural land, and atmospheric deposition. Increased

nitrogen may result in overgrowth of algae, which can decrease the dissolved oxygen content

of water, thereby harming or killing fish and other aquatic species. Controlling of nitrogen

load in the urban river systems is viewed as a priority by many river management authorities

as this affects algal growth. The HNRS has seen a number of episodes of algal blooms in the

past, causing public concerns. For examples, the shallow mid Nepean River section was

affected heavily by aquatic weed Egeria densa (Roberts et al., 1999). The Berowra Creek

estuarine section of the river was infested by toxic dinoflagellate algal blooms (SMEC,

1997).

The long-term persistence of elevated levels of Chlorophyll-a is a concern to water

authorities. An excessive growth often leads to poor water quality, noxious odours, oxygen

depletion, human health problems and fish kills. It may also be linked to harmful (toxic) algal

blooms. Poor water quality associated with high chlorophyll concentrations needs to be

distinguished from the natural variation observed with the seasons, with latitude, and those

associated with hydrodynamic features (e.g. upwelling). However, there is very little

information to make this distinction (Ward, 1998). Observed increases in the concentrations

of chlorophyll may be related to increased nutrient concentrations, decreased flow/changed

hydrodynamics (increased residence times) and/or decreased turbidity (increased light

CHAPTER 05: Results

109

penetration) (i.e. the increasing eutrophication status). The high Chlorophyll-a level at N14

needs to be investigated to find the possible reasons and to devise controlling measures.

If the alkalinity level is too high, the water can be cloudy, which inhibits the growth of

underwater plants i.e. it may restrict algal growths. A higher alkalinity may raise the pH

level, which in turn, can harm or kill fish and other aquatic organisms which are too sensitive

to higher pH levels. High alkalinity may result from the presence of the bicarbonate ion,

which is derived from the dissolution of carbonates by carbonic acids due to factors such as

weathering of limestone and dolomite rocks mainly composed of calcite. The high alkalinity

level at N14 in HNRS needs further investigation.

There are a number of factors that can lead to high conductivity levels. For examples, streams

that run through clay catchments may have a higher conductivity level due to the presence of

clay particles that ionize when enter into the river system. Groundwater inflows can have the

same effects if it contains clay particles. An underperforming STP could raise the

conductivity level in the effluent because of the presence of chloride, phosphate and nitrate.

CHAPTER 05: Results

110

5.7 COMPARISON OF MEASURED WATER QUALITY DATA WITH SCA DATA

Measured water quality data were compared with SCA data (Figure 5.37 – 5.57) considering

the closest site for different sampling location as follows.

S1 – Blaxland Crossing with N67

S2 – at M4 Bridge with N57

S3 - Weir Reserve with N44

When compared, measured pH data with SCA data, N57 (SCA station) and S2 (self-

monitoring) show similar values for the monitored year. The other 2 stations do not show

similar values. Measured and SCA data for DO show a considerable similarity for all the 3

stations. Only S2 and N57 show similar values for electrical conductivity. Turbidity values

for S1 and S2 sampling stations are similar to the SCA data. At S3, measured data does not

show comparatively higher values in June at S3 as compared to SCA data. For nitrogen

oxides and nitrogen as ammonia, only S2 and N57 show similar results. Measured and SCA

temperature values are almost similar for all the 3 sampling stations. Sampling stations for

self-monitoring are not exactly the SCA sampling locations, though they are the closest

points. This may be the reason for the variations of water quality data. Differences in

sampling procedures (ex, water depth) and collection time also can cause variations in the

water quality data. However, when considering overall data sets, they show a considerable

similarity.

CHAPTER 05: Results

111

5.7.1 pH

Figure 5.37. pH values at S1 and N67.


6.4

6.6

6.8

7

7.2

7.4

7.6

7.8

8

pH

Date

S1 Vs N67

6

6.2

6.4

6.6

6.8

7

7.2

7.4

7.6

7.8

8

pH

Date

S2 Vs N57

CHAPTER 05: Results

112


5.7.2 Dissolved Oxygen


7

7.1

7.2

7.3

7.4

7.5

7.6

7.7

7.8

7.9

8

pH

Date

S3 Vs N44

0

2

4

6

8

10

12

DO

Date

S1 Vs N67

CHAPTER 05: Results

113



0

2

4

6

8

10

12

14

DO

Date

S2 Vs N57

0

2

4

6

8

10

12

14

DO

Date

S3 Vs N44

CHAPTER 05: Results

114

5.7.3 Electrical Conductivity

Figure 5.43. Electrical conductivity at S1 and N67.


0

100

200

300

400

500

EC

Date

S1 Vs N67

0

100

200

300

400

EC

Date

S2 Vs N57

CHAPTER 05: Results

115


5.7.4 Turbidity

Figure 5.46. Turbidity at S1 and N67.

0

100

200

300

400

EC

Date

S3 Vs N44

0

10

20

30

40

50

60

70

80

Tu

rbid

ity

Date

S1 Vs N67

CHAPTER 05: Results

116



0

5

10

15

20

25

30

Tu

rbid

ity

Date

S2 Vs N57

0

5

10

15

20

25

30

35

40

Tu

rbid

ity

Date

S3 Vs N44

CHAPTER 05: Results

117

5.7.5 Nitrogen Oxides

Figure 5.49. Nitrogen oxides at S1 and N67.


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

NO

x m

g/L

Date

S1 Vs N67

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

NO

x m

g/L

Date

S2 Vs N57

CHAPTER 05: Results

118


5.7.6 Ammonical Nitrogen

Figure 5.52. Ammonical nitrogen at S1 and N67.

0

0.1

0.2

0.3

0.4

0.5

0.6

NO

x m

g/L

Date

S3 Vs N44

0

0.01

0.02

0.03

0.04

0.05

0.06

Nh

H-N

mg/L

Date

S1 Vs N67

CHAPTER 05: Results

119



0

0.004

0.008

0.012

0.016

0.02

NH

3-N

mg/L

Date

S2 Vs N57

0

0.01

0.02

0.03

0.04

0.05

0.06

NH

3-N

mg/L

Date

S3 Vs N44

CHAPTER 05: Results

120

5.7.7 Temperature

Figure 5.55. Temperature at S1 and N67.


0

5

10

15

20

25

30

Tem

per

atu

re

Date

S1 Vs N67

0

5

10

15

20

25

30

Tem

per

atu

re

Date

S2 Vs N57

CHAPTER 05: Results

121


5.8 CHAPTER SUMMARY

The results of the assessment of the water quality in the Hawkesbury Nepean River system

(HNRS) have been presented in this chapter. It has been found that the water quality

parameters vary along the length of the HNRS. Preliminary analyses from the box plots and

principal component analysis of the water quality parameters have shown that many of the

water quality parameters are highly correlated and some of the monitoring stations do not

provide any independent information.

From the trend analysis, a general pattern of downward trends of pH, nitrogen TKN,

alkalinity, dissolved oxygen and electrical conductivity has been detected. Total iron,

filterable iron, true colour, total aluminium, reactive silicate and dissolved organic carbon

0

5

10

15

20

25

30

Tem

per

atu

re

Date

S3 Vs N44

CHAPTER 05: Results

122

demonstrate increasing trends at most of the stations, and total phosphorus, suspended solids,

filterable aluminium, ammonia nitrogen and filterable phosphorus do not show any trend at

most of the stations. The median values for chlorophyll-a, total nitrogen and alkalinity are

above the ANZECC (2000) trigger values for most of the stations. The increasing trend of

turbidity, chlorophyll-a, alkalinity, dissolved organic carbon, total iron, total aluminium, total

manganese and reactive silicate and the exceedance of the ANZECC (2000) trigger values for

chlorophyll-a, total nitrogen and alkalinity indicate an overall water quality deterioration in

the HNRS during the last decade. The parameters such as phosphorus, suspended solids and

ammonia nitrogen do not show any marked change over the period of this study. Although an

improvement in water quality can be seen at some stations downstream of the undisturbed

parts of the catchment, there has been an overall water quality deterioration in the HNRS

during the last decade.

Three prediction equations have been developed for three important water quality parameters

(chlorophyll-a, total nitrogen and total phosphorous) for the HNRS. These equations

generally present a high co-efficient of determination values and satisfy the assumptions of

least squares regression analysis. These equations can be used to estimate chlorophyll-a, total

nitrogen and total phosphorous from easily measurable water quality parameters.

Application of Canadian Water Quality Index method has shown that water quality at the 9

stations fall under either poor or marginal category, based on the Canadian Water Quality

Index (CWQI) categorisation where the CWQI values are found to be in the range of 33 to

57. Marginal water quality is found for 5 stations and poor water quality is found for the

CHAPTER 05: Results

123

remaining 4 stations. None of the stations were found to have good quality water. Stations

N14 and N35 were found to be the most polluted stations in the HNRS among the 9 stations.

With detailed investigation at station N14, it was found that the higher values of water quality

parameters: Nitrogen, Chlorophyll a, Iron, Aluminium, Alkalinity and Conductivity have

contributed to the poor water quality condition at N14.

Comparison of self-monitored water quality data SCA data obtained from nearby sampling

stations show a considerable similarity.

CHAPTER 06: Conclusion

124

CHAPTER 6

SUMMARY AND CONCLUSIONS

6

6.0 SUMMARY

The Hawkesbury Nepean River System (HNRS) is one of the most important rivers in

Australia as it supplies water to over 4 million people in Sydney. HNRS has multiple and

complex land uses. Hence, the water quality of this river is of great significance. In this study,

a total of 9 water quality parameters have been used from 15 water quality monitoring

stations plus one-year self-monitoring to assess the quality of water in the HNRS.

6.1 PRELIMINARY WATER QUALITY DATA ANALYSIS

From the preliminary data analysis, it has been found that the average concentrations of some

water quality parameters, such as total nitrogen and alkalinity, are higher than those

recommended by the Australian and New Zealand Environment and Conservation Council

(ANZECC) guidelines at most of the monitoring stations along the HNRS. The higher levels

of total nitrogen might be attributed to runoff from agricultural lands, urban areas and sewage

treatment plants. A higher value of total phosphorus at station N35 is observed. Station N14

shows notable higher conductivity values and N35 shows higher values of nitrogen oxides

and chlorophyll-a levels. Also, station N21 has high levels of chlorophyll-a which is found to

be above the recommended guidelines. High nitrogen, phosphorus and chlorophyll-a levels at

many stations appear to be a sign of deteriorated water quality in the HNRS.


125

6.2 TREND ANALYSIS

From the trend analysis, a general pattern of downward trends of pH, nitrogen TKN,

alkalinity, dissolved oxygen and electrical conductivity has been detected. Total iron,

filterable iron, true colour, total aluminium, reactive silicate and dissolved organic carbon

demonstrate an increasing trend at most of the stations. In addition, total phosphorus,

suspended solids, filterable aluminium, ammonia nitrogen and filterable phosphorus do not

show any trend at most of the stations. The median values for chlorophyll-a, total nitrogen

and alkalinity are above the ANZECC (2000) trigger values for most of the stations. The

increasing trend of turbidity, chlorophyll-a, alkalinity, dissolved organic carbon, total iron,

total aluminium, total manganese and reactive silicate and the exceedance of the ANZECC

(2000) trigger values for chlorophyll-a, total nitrogen and alkalinity indicate an overall water

quality deterioration occurred in the HNRS during the last decade. The parameters such as

phosphorus, suspended solids and ammonia nitrogen do not show any marked changes over

the period of this study. Although an improvement in water quality can be seen at some

stations at downstream of the undisturbed parts of the catchment, trend analysis shows an

overall water quality deterioration in the HNRS during the last decade.

6.3 REGRESSION ANALYSIS

Using the regression analysis, three prediction equations have been developed for three

important water quality parameters (chlorophyll-a, total nitrogen and total phosphorous) for

the HNRS. These equations generally present a high co-efficient of determination values and


126

satisfy the assumptions of least squares regression analysis. These equations can be used to

estimate chlorophyll-a, total nitrogen and total phosphorous from easily measurable water

quality parameters.

6.4 APPLICATION OF CANADIAN WATER QUALITY INDEX METHOD

Application of Canadian Water Quality Index (CWQI) method shows the water quality at the

9 stations fall under either poor or marginal category based on the CWQI categorisation,

where the CWQI values are found to be in the range of 33 to 57. Marginal water quality is

found for 5 stations and poor water quality for the remaining 4 stations. None of the stations

are found to have good quality water. Stations N14 and N35 are found to be the most polluted

stations in the HNRS among the 9 stations. With detailed investigation at the station N14, it is

found that the higher values of water quality parameters: Nitrogen, Chlorophyll a, Iron,

Aluminium, Alkalinity and Conductivity have contributed to the poor water quality condition

at N14. There are many sewage treatment plants that discharge reed waste water to upstream

of station N35. Also, the dominant land use in this part of the catchment includes rural,

grazing, commercial gardening, intensive agriculture and urban and industrial activities. The

low WQI at N35 can be attributed to these land uses. Water quality at station N14 should be

improved because of dilution by high quality inflows from the Colo River and the

undisturbed upstream catchment. The high pollutant levels at N14 need to be investigated to

find the possible reasons and to devise controlling measures. `


127

6.5 COMPARISON OF MEASURED WATER QUALITY DATA WITH SCA DATA

In general, the self-monitored water quality data is similar to SCA data obtained from nearby

sampling stations.

6.6 CONCLUSION

The following conclusions can be drawn from this study.

The concentrations of total phosphorus, nitrogen oxides and chlorophyll-a are higher

than those recommended by the Australian and New Zealand Environment and

Conservation Council (ANZECC) guidelines.

An increasing trend has been detected for turbidity, chlorophyll-a, alkalinity,

dissolved organic carbon, total iron, total aluminium, total manganese and reactive

silicate for majority of the monitoring stations.

Application of Canadian Water Quality Index method shows the water quality at 9

stations fall under either poor or marginal category.

Stations N14 and N35 are found to be the most polluted stations in the HNRS among

the 9 stations.

Although an improvement in water quality can be seen at some stations at

downstream of the undisturbed parts of the catchment, there has been an overall water

quality deterioration in the HNRS during the last decade.

The developed prediction equations for three important water quality parameters

(chlorophyll-a, total nitrogen and total phosphorous) can be used to predict these

water quality parameters for the HNRS.


128

6.7 LIMITATIONS OF THE STUDY

The study has a number of limitations as noted below:

Self-monitoring was not conducted for all the selected water quality parameters due to

limited laboratory facilities.

Water quality modelling could not be conducted due to the complex land use and the

large catchment size.

The seasonality effects on water quality were not investigated.

6.8 RECOMMENDATIONS FOR FUTURE RESEARCH

An artificial intelligence based model can be developed to increase the overall

prediction accuracy of various water quality parameters along the HNRS where easily

measurable water quality parameters can be used to predict other water quality

parameters.

Specific land use data can be obtained and a close monitoring program can be

developed to link water quality and land use characteristics.

Automatic water sampling probes and the telemetry system should be developed to

provide real-time assessment of water quality for this very important river system.

A random sampling technique should be developed by a joint group of water

authorities and universities to uncover any major water quality deterioration in the

HNRS in future.


129

References

Alberto, W. D., Marıa del Pilar, D., Marıa Valeria, A., Fabiana, P. S., Cecilia, H. A. & Marıa

de los Ángeles, B. (2001). Pattern recognition techniques for the evaluation of spatial and

Temporal variations in water quality. A case study: Suquıa River basin (Córdoba–Argentina).

Water Research, 35(12), 2881-2894.

Al-Janabi, Z. Z., Al-Kubaisi, A., & Alobaidy, A. H. M. J. (2012). Assessment of water

quality of Tigris River by using water quality index (CCME-WQI). Journal of Al-Nahrain

University, 15(1), 119-126.

ANZECC (2000). Australian and New Zealand guidelines for fresh and marine water quality,

Australian and New Zealand Environment and Conservation Council.

http://www.environment.gov.au/water/publications/quality/nwqms-guidelines-4-vol1.html

Astaraie-Imani, M., Kapelan, Z., Fu, G. & Butler, D. (2012). Assessing the combined effects

of urbanisation and climate change on the river water quality in an integrated urban

wastewater system in the UK. Journal of Environmental Management, 112, 1-9.

Astel, A., Tsakovski, S., Barbieri, P. & Simeonov, V. (2007). Comparison of self-organizing

maps classification approach with cluster and principal components analysis for large

environmental data sets. Water Research, 41(19), 4566-4578.

Atech (2000). Cost of algal blooms. LWRRDC Occasional paper 26/99. Land and Water

Resources Research and Development Corporation, Canberra, ACT, Australia, 42 pp.

Baginska, B., Pritchard, T. & Krogh, M. (2003). Roles of land use resolution and unit-area

load rates in assessment of diffuse nutrient emissions. Journal of Environmental

Management, 69(1), 39-46.

Bartholomew, D. J., Steele, F., Galbraith, J., & Moustaki, I. (2008). Analysis of multivariate

http://www.environment.gov.au/water/publications/quality/nwqms-guidelines-4-vol1.html


130

social science data. CRC press.

Boman, B. J., Wilson, P. C., & Ontermaa, E. A. (2002). Understanding water quality

parameters for citrus irrigation and drainage systems. University of Florida IFAS Extension,

Circ, 1406.

BOM (2013). Australian government – bureau of meteorology.

http://www.bom.gov.au/water/nwa/2011/sydney/

Bowes, M. J., Gozzard, E., Johnson, A. C., Scarlett, P. M., Roberts, C., Read, D. S., ... &

Wickham, H. D. (2012). Spatial and temporal changes in chlorophyll-a concentrations in the

River Thames basin, UK: are phosphorus concentrations beginning to limit phytoplankton

biomass?. Science of the Total Environment, 426, 45-55.

Boyacioglu, H. (2010). Utilization of the water quality index method as a classification tool.

Environmental Monitoring and Assessment, 167(1-4), 115-124.

Brown, P. J., Le, N. D. & Zidek, J. V. (1994). Multivariate spatial interpolation and exposure

to air pollutants. Canadian Journal of Statistics, 22(4), 489-509.

Chang, H. (2008). Spatial analysis of water quality trends in the Han River basin, South

Korea. Water Research, 42(13), 3285-3304.

Commonwealth of Australia (1996). Australia State of the Environment.

Crosa, G., Froebrich, J., Nikolayenko, V., Stefani, F., Galli, P. & Calamari, D. (2006). Spatial

and seasonal variations in the water quality of the Amu Darya River (Central Asia). Water

Research, 40(11), 2237-2245.

Davis, J.C. (1986). Statistics and data analysis in geology. New York: Wiley, 2, 646.

Dawson, E. J. & Macklin, M. G. (1998). Speciation of heavy metals on suspended sediment

http://www.bom.gov.au/water/nwa/2011/sydney/


131

under high flow conditions in the River Aire, West Yorkshire, UK. Hydrological Processes,

12(9), 1483-1494.

Dede, O. T., Telci, I. T., & Aral, M. M. (2013). The use of water quality index models for the

evaluation of surface water quality: a case study for Kirmir Basin, Ankara, Turkey. Water

Quality, Exposure and Health, 5(1), 41-56.

Draper, N. R., Smith, H., & Pownell, E. (1966). Applied regression analysis. New York:

Wiley, 3, 85-96.

Drápela, K. & Drápelová, I. (2011). Application of Mann-Kendall test and the Sen's slope

estimates for trend detection in deposition data from Bílý Kríz (Beskydy Mts., the Czech

Republic) 1997-2010. Beskydy, 4(2), 133-146.

Diamond, R. (2004). Water and Sydney’s future-balancing the value of our rivers and

economy: Final report of the Hawkesbury–Nepean River Management Forum. Department of

Infrastructure, Planning and Natural Resources, Sydney, NSW.

Dixon, W. & Chiswell, B. (1996). Review of aquatic monitoring program design. Water

Research, 30(9), 1935-1948.

Egodawatta, P., Thomas, E. & Goonetilleke, A. (2009). Understanding the physical processes

of pollutant build-up and wash-off on roof surfaces. Science of the Total Environment,

407(6) 1834-41.

Eroksuz, E. & Rahman, A. (2010). Rainwater tanks in multi-unit buildings: A case study for

three Australian cities. Resources, Conservation and Recycling, 54(12), 1449-1452.

Erturk, A., Gurel, M., Ekdal, A., Avsan, C., Ugurluoglu, A., Seker, D. Z., Tanik, A. &

Ozturk, I. (2010). Water quality assessment and meta modal development in Melen

watershed – Turkey. Journal of Environmental Management, 91, 1526-1545.


132

Gburek, W. J. & Folmar, G. J. (1999). Flow and chemical contributions to streamflow in an

upland watershed: A base flow survey. Journal of Hydrology, 217(1), 1-18.

Gehrke, P. C. & Harris, J. H. (1996). Fish and fisheries of the Hawkesbury-Nepean River

System. Final Rapport to the Sydney Water Corporation.

Gitelson, A. (1992). The peak near 700 nm on radiance spectra of algae and water:

Relationships of its magnitude and position with chlorophyll concentration. International

Journal of Remote Sensing, 13(17), 3367-3373.

Haddad, K., Egodawatta, P., Rahman, A. & Goonetilleke, A. (2013). Uncertainty analysis of

pollutant build-up modelling based on a bayesian weighted least squares approach. Science of

the Total Environment, 449, 410-417.

Hanrahan, G., Gledhill, M., House, W. A. & Worsfold, P. J. (2003). Evaluation of

phosphorus concentrations in relation to annual and seasonal physico-chemical water quality

parameters in a UK chalk stream. Water Research, 37(15), 3579-3589.

Horton, R. K., (1965). An index-number system for rating water quality. Journal of Water

Pollution Control Federation. 37(3), 300–306.

Interlandi, S. J. & Crockett, C. S. (2003). Recent water quality trends in the Schuylkill River,

Pennsylvania, USA: A preliminary assessment of the relative influences of climate, river

discharge and suburban development. Water Research, 37(8), 1737-1748.

Kaika, M. (2003). The water framework directive: A new directive for a changing social,

political and economic European framework. European Planning Studies 11, 299-316.

Karr, J. R. (1999). Defining and measuring river health. Freshwater Biology, 41(2), 221-234.

Kendall, C. A. R. O. L., Elliott, E. M. & Wankel, S. D. (2007). Tracing anthropogenic inputs

of nitrogen to ecosystems. Stable Isotopes in Ecology and Environmental Science, 2(1), 375-


133

449.

Khan, A. A., Paterson, R., & Khan, H. (2004). Modification and application of the Canadian

council of ministers of the environment water quality index (CCME WQI) for the

communication of drinking water quality data in Newfoundland and Labrador. Water Quality

Research Journal of Canada, 39(3), 285-293.

Kovács, J., Tanos, P., Korponai, J., Székely, I. K., Gondár, K., Gondár-Sőregi, K. & Hatvani,

I. G. (2012). Analysis of water quality data for scientists. Water Quality Monitoring and

Assessment. 65-94.

Lack, T. J. (1971). Quantitative studies on the phytoplankton of the Rivers Thames and

Kennet at Reading. Freshwater Biology, 1(2), 213-224.

LeBlanc, R. T., Brown, R. D. & FitzGibbon, J. E. (1997). Modeling the effects of land use

change on the water temperature in unregulated urban streams. Journal of Environmental

Management, 49(4), 445-469.

Lumb, A., Sharma, T. C., Bibeault, J. F., & Klawunn, P. (2011). A comparative study of USA

and Canadian water quality index models. Water Quality, Exposure and Health, 3(3-4), 203-

216.

Ma, J., Ding, Z., Wei, G., Zhao, H. & Huang, T. (2009). Sources of water pollution and

evolution of water quality in the Wuwei basin of Shiyang River, North Chaina. Journal of

Environmental Management, 90, 1168-1177.

Markich, S. J. & Brown, P. L. (1998). Relative importance of natural and anthropogenic

influences on the fresh surface water chemistry of the Hawkesbury–Nepean River, south-

eastern Australia. Science of the Total Environment, 217(3), 201-230.

Marsden, T. J., & Gehrke, P. C. (1996). Fish passage in the Hawkesbury-Nepean River

system. Fish and Fisheries of the Hawkesbury-Nepean River System. NSW Fisheries


134

Research Institute, Sydney, 200-218.

Massart, D. L., Vandeginste, B. G. M., Deming, S. N., Michotte, Y. & Kaufman, L. (1988).

Chemometrics: A textbook, 20-21.

Meybeck, M. (2002). Riverine quality at the Anthropocene: Propositions for global space and

time analysis, illustrated by the Seine River. Aquatic Sciences, 64(4), 376-393.

NLWRA (2008). The national land and water resources audit final report: 2002-2008.

Australian government. http://lwa.gov.au/files/products/national-land-and-water-resources-

audit/pn21537/pn21537.pdf.

Norris, R. H., & Thoms, M. C. (1999). What is river health?. Freshwater Biology, 41(2), 197-

209.

Osborne, L. L. & Wiley, M. J. (1988). Empirical relationships between land use/cover and

stream water quality in an agricultural watershed. Journal of Environmental Management,

26(1), 9-27.

Pinto, U., Maheshwari, B. L. & Grewal, H. S. (2010). Effects of greywater irrigation on plant

growth, water use and soil properties. Resources, Conservation and Recycling, 54(7), 429-

435.

Pinto, U. & Maheshwari, B. L. (2011). River health assessment in peri-urban landscapes: An

application of multivariate analysis to identify the key variables. Water Research, 45(13),

3915-3924.

Pinto, U., Maheshwari, B. L., Shrestha S. & Morris C. (2012). Modelling eutrophication and

microbial risks in peri-urban River systems using discriminant function analysis .Water

Research, 46, 6476-6488.

Pinto, U., Maheshwari, B. L. & Ollerton, R. L. (2013). Analysis of long-term water quality

http://lwa.gov.au/files/products/national-land-and-water-resources-audit/pn21537/pn21537.pdf

http://lwa.gov.au/files/products/national-land-and-water-resources-audit/pn21537/pn21537.pdf


135

for effective river health monitoring in peri-urban landscapes: A case study of the

Hawkesbury–Nepean River system in NSW, Australia. Environmental Monitoring and

Assessment, 185(6), 4551-4569.

Rahman, A., Thomas, E. C., Bhuiyan, S., & Goonetilleke, A. (2002). Modelling pollutant

washoff from south east Queensland catchments Australia.

Rahman, A., Keane, J., & Imteaz, M. A. (2012). Rainwater harvesting in Greater Sydney:

Water savings, reliability and economic benefits. Resources, Conservation and Recycling, 61,

16-21.

Roberts, D. E., Church, A. G. & Cummins, S. P. (1999). Invasion of Egeria into the

Hawkesbury-Nepean River, Australia. Aquatic Plant Management, 37, 31-34.

Russell, D. W. (2002). In search of underlying dimensions: The use (and abuse) of factor

analysis in personality and social psychology bulletin. Personality and Social Psychology

Bulletin, 28(12), 1629-1646.

Sarkar, C. & Abbasi, S. A. (2006). A new software for generating water quality indices.

Environmental Monitoring and Assessment, 119, 201-231.

Sen, P. K. (1968). Estimates of the regression coefficient based on Kendall's tau. Journal of

the American Statistical Association, 63(324), 1379-1389.

SMEC (1997). Options for sewage treatment and effluent disposal for West Hornsby and

Hornsby Heights sewage treatment plants. SMEC Australia.

Solbe, J. F. (1986). Effects of land use on fresh waters: Agriculture, forestry, mineral

exploitation, urbanisation.

Sousounis, P. J., & Bisanz, J. M. (2000). Preparing for a changing climate: The potential

consequences of climate variability and change. Great Lakes Overview.


136

Stanhill, G., & Cohen, S. (2001). Global dimming: a review of the evidence for a widespread

and significant reduction in global radiation with discussion of its probable causes and

possible agricultural consequences. Agricultural and forest meteorology, 107(4), 255-278.

Stroomberg, G. J., Freiriks, I. L., Smedes, F., & Cofino, W. P. (1995). Quality assurance and

quality control of surface water sampling. Quality Assurance in Environmental Monitoring.

Sampling and Sample Pretreatment, 51.

Stubblefield, A. P., Reuter, J. E., Dahlgren, R. A. & Goldman, C. R. (2007). Use of

turbidometry to characterize suspended sediment and phosphorus fluxes in the Lake Tahoe

basin, California, USA. Hydrological Processes, 21(3), 281-291.

Suhr, D. D. (2005). Principal component analysis vs. exploratory factor analysis. SUGI 30

proceedings, 203, 230.

Tabari, H., Marofi, S. & Ahmadi, M. (2011). Long-term variations of water quality

parameters in the Maroon River, Iran. Environmental Monitoring and Assessment, 177(1-4),

273-287.

Thoms, M. C., Parker, C. R., & Simons, M. (2000). The dispersal and storage of trace metals

in the Hawkesbury River valley. River Management: The Australasian Experience, 197-219.

Tian, J., & Fernandez, G. (1999). Seasonal trend analysis of monthly water quality data.

Proceedings of the 7th Annual Western users of SAS software regional users group, 229-234.

Turner, L. & Erskine, W. D. (2005). Variability in the development, persistence and

breakdown of thermal, oxygen and salt stratification on regulated rivers of southeastern

Australia. River Research and Applications, 21(2‐3), 151-168.

Vega, M., Pardo, R., Barrado, E. & Debán, L. (1998). Assessment of seasonal and polluting

effects on the quality of river water by exploratory data analysis. Water Research, 32(12),

3581-3592.


137

Van der Sterren, M., Rahman, A., Shrestha, S., Barker, G. & Ryan, G. (2009). An overview

of on-site retention and detention policies for urban stormwater management in the greater

Western Sydney region in Australia, Water International, 34(3), 362-372.

Van der Sterren, M., Rahman, A. & Dennis, G.R. (2012). Implications to stormwater

management as a result of lot scale rainwater tank systems: a case study in Western Sydney,

Australia, Water Science and Technology, 65.8, 1475-1482.

Van der Sterren, M., Rahman, A. & Dennis, G. (2013). Quality and quantity monitoring of

five rainwater tanks in Western Sydney, Australia. Journal of Environmental Engineering,

139, 332-340.

Van der Sterren, M., Rahman, A. & Ryan, G. (2014). Modeling of a lot scale rainwater tank

system in XP-SWMM: A case study in Western Sydney, Australia, Journal of Environmental

Management, 141, 177-189.

Van der Sterren, M., & Rahman, A. (2015). Single lot on site detention requirements in New

South Wales Australia and its relation to holistic storm water management. Sustainability of

Water Quality and Ecology, 6, 48-56.

Vitousek, P. M., Cassman, K., Cleveland, C., Crews, T., Field, C. B., Grimm, N. B. & Sprent,

J. I. (2002). Towards an ecological understanding of biological nitrogen fixation.

Biogeochemistry, 57(1), 1-45.

WBC (1994). Water Board (Corporatisation) Act.

http://www.austlii.edu.au/au/legis/nsw/num_act/wba1994n88323.pdf.

Walsh, C. J., Roy, A. H., Feminella, J. W., Cottingham, P. D., Groffman, P. M., & Morgan,

R. P. (2005). The urban stream syndrome: Current knowledge and the search for a cure.

Journal of the North American Benthological Society, 24(3), 706-723.

Ward, S. N. (1998). On the consistency of earthquake moment release and space geodetic

strain rates: Europe, Geophysical Journal International, 135(3), 1011-1018.

http://www.austlii.edu.au/au/legis/nsw/num_act/wba1994n88323.pdf


138

Wenning, R. J. & Erickson, G. A. (1994). Interpretation and analysis of complex

environmental data using chemometric methods. TrAC Trends in Analytical Chemistry,

13(10), 446-457.

Whitehead, P. G., & Hornberger, G. M. (1984). Modelling algal behaviour in the River

Thames. Water Research, 18(8), 945-953.

WQA (2013). eWater Ltd. Water Board Corporatisation Act, (1994).

http://www.austlii.edu.au/au/legis/nsw/num_act/wba1994n88323.pdf.

ung, T.F. & Sanzone, S. (2002). A Framework for assessing and reporting on ecological

condition. US Environmental Protection Agency, EPA Science Advisory Board.

Yue, S., Pilon, P. & Cavadias, G. (2002). Power of the Mann–Kendall and Spearman's rho

tests for detecting monotonic trends in hydrological series. Journal of Hydrology, 259(1),

254-271.

Yu, Y. S., Zou, S. & Whittemore, D. (1993). Non-parametric trend analysis of water quality

data of rivers in Kansas. Journal of Hydrology, 150(1), 61-80.

http://www.austlii.edu.au/au/legis/nsw/num_act/wba1994n88323.pdf

water quality assessment in the hawkesbury …

Documents