SRTRATEGIC ANALYSIS OF SPATIAL AND TEMPORAL WATER QUALITY OF RIVER CHENAB AND ITS MANAGEMENT
submitted By
MUHAMMAD TOUSIF BHATTI (2005-PhD-CEWRE-07)
FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY IN WATER RESOURCES ENGINEERING
CENTRE OF EXCELLENCE IN WATER RESOURCES ENGINEERING University of Engineering and Technology LAHORE, PAKISTAN
i SRTRATEGIC ANALYSIS OF SPATIAL AND TEMPORAL WATER QUALITY OF RIVER CHENAB AND ITS MANAGEMENT
By
Muhammad Tousif Bhatti 2005-Ph.D.-CEWRE-07
A thesis submitted in fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY IN WATER RESOURCES ENGINEERING
Thesis examination date: 30-05-2009
______Prof. Dr. Muhammad Latif Prof. Dr. M.A.Q Jahangir Durrani Research Advisor & Internal Examiner External Examiner NWFP University of Engineering and Technology, Peshawar, Pakistan
______Prof. Dr. Muhammad Latif Director, CEWRE
Thesis Submitted on:
CENTRE OF EXCELLENCE IN WATER RESOURCES ENGINEERING University of Engineering and Technology, Lahore, Pakistan
2009
ii
This work is dedicated to my mother she could not wait to see the completion of this work but her prayers, love and care enabled me to do it all May her soul rest in Heavens (Aameen)
iii ABSTRACT
Water quality of many rivers in the developing countries is under serious threat of degradation and Pakistan is no exception to this. The river water may be polluted by the effluents stemming from industrial, municipal, agricultural or mining activities. The most affected rivers are those flowing through the urban areas and subjected to anthropogenic activities. The river Chenab, traversing near the industrial cities and municipalities, is largely used for constant disposal of untreated effluents in the Punjab province of
Pakistan. Consequently water quality of the river degrades particularly in the low flow months.
This study was conducted to monitor, assess and model the water quality (WQ) of river Chenab over a length of 292 km from its entrance in Pakistan at Marala. The monitoring program was conducted during low flow months (October to March) of years
2006-7 and 2007-8. Water samples were collected from seven locations along the river and all the contributing drains as well. These samples were analyzed for a variety of physical, chemical and biological quality parameters.
The data collected from monitoring as well as from secondary sources were utilized in three phases of analysis. In the first phase water quality indices (WQIs) were calculated using CWQI 1.0 model developed by Canadian Council of Ministers of the
Environment (CCME). Three intended uses of river water i.e. drinking, aquatic life and irrigation were incorporated for WQI calculations at selected points along the river. In the second phase, mathematical model (MIKE 11 model developed by Danish Hydraulic
Institute (DHI), Denmark) was formulated to simulate a conservative WQ parameter
v (salinity of river water). Two non-conservative WQ parameters (dissolved oxygen (DO) and biochemical oxygen demand (BOD)) were modeled in third phase of the analysis using MIKE 11 model.
The results of WQI revealed that the lower river reach (185 to 233 km) was more polluted than the upper 185 km segment. In this river reach, overall WQI ranking were poor for drinking and marginal for both irrigation and aquatic life. The WQIs for all three uses were ranked poor at sampling point located at 233 km below Marala headworks. The calibrated model for salinity simulated the most saline condition in the river during the months with minimum flow (i.e. November and December). The results also depicted high salinity in the downstream river reach receiving polluted effluents from two major drains (Faqirian Sillanwali and Chakbandi drain). Finally the model was calibrated and validated for DO and BOD. The results of simulations indicated DO depletion and high
BOD levels in the downstream river reaches particularly from 200 to 270 km.
Different scenarios were also tested to predict the river water salinity by varying discharge of the drains. The salinity of river water was found highly sensitive to the amount of effluents added by the surface drains. The study of management scenarios for
BOD suggested that the maximum water quality improvement can be achieved if there is no diversion of flow from the river coupled with 60 percent reduction in BOD of the drain effluents through treatment.
vi ACKNOWLEDGEMENTS
All praise to ALMIGHTY ALLAH, the most compassionate, the most merciful; who blessed me to complete my studies and humblest and deepest gratitude to the greatest educator of mankind, the Holy Prophet MUHAMMAD (Peace be
Upon Him).
My deepest gratefulness is due to my Research Advisor Prof. Dr.
Muhammad Latif (Director, Centre of Excellence in Water Resources Engineering,
University of Engineering & Technology, Lahore) whose intellectual inspiration, guidance, encouragement and regular discussions have been invaluable to me. His continual willingness to listen, discuss and render critical judgments helped me to produce this work in its present shape.
I wish to acknowledge my external examiners; Prof. Dr. M.A.Q. Jahangir
Durrani (Pakistan), Prof. Dr. Faisal Khan (Memorial University of Newfoundland,
Canada) and Prof. Dr. Shahbaz Khan (UNESCO, France) for assessing my PhD dissertation in spite of their professional and administrative engagements. I am thankful for their useful suggestions, personal interest, encouraging attitude and devoted guidance. Their critical readings and suggestions were helpful in further improving the dissertation.
I express my heartiest thankfulness to my father Muhammad Idrees Bhatti whose motivation, support, love and prayers have always been with me in my life. I
vii also acknowledge my beloved sisters and brothers for their continuous encouragement and care for me.
My thanks are due to my friends especially Adnan Khan Niazi, Zeeshan
Shehzad and Nadeem Abbas. Their love, sincerity and support are priceless.
The whole remains incomplete, if I do not record my sincerest thanks to all faculty members and staff of CEWRE for their continued help and cooperation in completion of my research work and this manuscript. I also acknowledge the contribution of all those who helped me in any way during data collection, analysis and other phases of this research study.
Eng. Muhammad Tousif Bhatti Dated: June 4, 2009
viii TABLE OF CONTENTS
ABSTRACT ...... v AKNOWLEDGEMENTS...... vii TABLE OF CONTENTS ...... ix LIST OF TABLES ...... xiii LIST OF FIGURES ...... xv LIST OF ABREVIATIONS ...... xviii
CHAPTER
I INTRODUCTION...... 1
1.1 BACKGROUND OF THE STUDY...... 1 1.2 PROBLEM STATEMENT AND SCOPE OF WORK...... 4 1.3 RESEARCH HYPOTHESIS ...... 6 1.4 OBJECTIVES ...... 7
II REVIEW OF LITERATURE ...... 8
2.1 LOW FLOW ANALYSIS...... 8
2.1.1 Low-Flow Measures and Indices and Their Estimation from Stream-Flow Time Series ……… ...... 9
2.2 WATER QUALITY MONITORING AND ASSESSMENT...... 14
2.2.1 Global Perspective………...... 14 2.2.2 Local Perspective………...... 27
2.3 WATER QUALITY INDEXING...... 33 2.4 WATER QUALITY MODELING...... 38
2.4.1 Global Perspective……………...... 38 2.4.2 Local Perspective…………… ...... 48
2.5 DERIVED CONCLUSION ...... 49
III MONITORING AND ASSESSMENT OF RIVER WATER QUALITY ...... 52
3.1 THE RIVER CHENAB...... 52
3.1.1 Historical Background…...... 52 3.1.2 Location…...... 52 3.1.3 Chenab River Basin…...... 53
ix Table of Contents (Continued)
3.1.4 River Topology………...... 54 3.1.5 Hydraulic Infrastructures…...... 56 3.1.6 Travel and Lead Times…...... 68 3.1.7 Historical Discharges...... 70
3.2 SELECTION OF RIVER REACH…...... 70 3.3 DESIGN METHODOLOGY...... 72
3.3.1 Monitoring Strategy ...... 74 3.3.2 Network Design...... 74 3.3.3 Sample Collection...... 76 3.3.4 Laboratory analysis...... 78
IV WATER QUALITY INDEXING...... 80
4.1 THEORATICAL CONCEPTS ...... 80
4.1.1 Structure of Various Water Quality Indices...... 81 4.1.2 CWQI Model Development…...... 84
4.2 INPUT DATA FOR CWQI 1.0 MODEL ...... 89 4.3 STEPS FOR WQI CALCULATION BY USING CWQI 1.0 MODEL…...... 92
4.3.1 Start Page...... 92 4.3.2 Data Page ...... 93 4.3.3 Criteria Page ...... 95 4.3.4 Report Page...... 97 4.3.5 The WQI Page ...... 98 4.3.6 The WQI Chart …...... 98
V HYDRODYNAMIC AND WATER QUALITY MODELING ...... 100
5.1 THEORATICAL CONCEPTS AND WATER QUALITY MODELS ...100 5.2 MIKE 11 MODEL...... 106
5.2.1 The Conceptual Model…...... 106 5.2.2 Processes of the Model...... 108 5.2.3 Data Requirements of the Model...... 111 5.2.4 Operation of the Model …...... 112 5.2.5 Outputs of the Model …...... 113
x
Table of Contents (Continued)
5.3 COEFFICIENTS USED IN MIKE 11 MODEL FORMULATION ...... 113
5.3.1 Deoxygenation Coefficient ...... 113 5.3.2 Dispersion Coefficient (D)...... 116
5.4 STEPS FOR MODEL FORMULATION IN MIKE 11...... 117
5.4.1 The Simulation Editor...... 118 5.4.2 The Network Editor...... 121 5.4.3 The Cross-Section Editor...... 123 5.4.4 The Boundary Editor…… ...... 124 5.4.5 The Hydrodynamic Parameter Editor…...... 128 5.4.6 The Advection-Dispersion Editor……...... 129
5.5 MIKE VIEW…...... 131 5.6 MODEL EVALUATION STATISTICS ...... 132
5.6.1 Root Mean Square Error (RMSE) ...... 132 5.6.2 Relative Mean Absolute Error (MAE)rel …...... 133 5.6.3 Percent Bias (PBIAS)...... 134 5.6.4 Nash-Sutcliffe Efficiency (NSE)...... 135 5.6.5 Coefficient of Determination (R2)……...... 136
VI RESULTS AND DISCUSSION…...... 137
6.1 LOW FLOW ANALYSIS OF RIVER CHENAB...... 137 6.2 WATER QUALITY ASSESSMENT...... 140 6.3 WATER QUALITY INDEXING...... 144 6.4 MODELING OF RIVER WATER SALINITY...... 147
6.4.1 Model Calibration...... 147 6.4.2 Simulation of Salinity during Low Flow Months...... 151
6.5 MODELING OF DISSOLVED OXYGEN AND BIOCHEMICAL OXYGEN DEMAND ...... 155
6.5.1 Model Calibration …… ...... 155 6.5.2 Model Validation and Testing…...... 159
6.6 MODEL APPLICATION AND MANAGEMENT SCENARIOS...... 161
6.6.1 Application of Model for Salinity ……...... 162 6.6.2 Application of Model for BOD…...... 163
xi
Table of Contents (Continued)
VI SUMMARY, CONCLUSIONS AND RECOMMENDATIONS...... 166 7.1 SUMMARY ...... 166 7.2 CONCLUSIONS ...... 167 7.3 RECOMMENDATIONS ...... 169
REFERENCES ...... 171
APPENDICES
Appendix-I Analytical Techniques for the Analysis of Water and Wastewater Samples ...... 185 Appendix-II Selected Cross-Sections of Chenab River Used in the Formulation of Mike 11 (HD) Model ...... 197 Appendix-III Calculation of Flow Duration Curve for River Chenab at Marala ...... 201 Appendix-IV Calculation of Low Flow Frequency Curve (LFFC) ...... 217
VITA………………...... 220
xii LIST OF TABLES
Table Description Page #
2.1 Location of sampling points and analytical data of Chenab river...... 28
2.2 Sampling stations at various rivers of Pakistan ...... 29
2.3 Industrial effluent discharging into river Chenab...... 29
3.1 Observed wave travel time and potential lead time in river Chenab ...... 69
3.2 Mean monthly discharge of river Chenab at Marala for 60 year period (1946-47 to 2006-07)…… ...... 70
3.3 Water quality monitoring stations selected for the water quality study on river Chenab……...... 76
3.4 Summry of special sampling and handling requirements...... 77
3.5 The selected water quality parameters for laboratory and in-situ analysis...... 78
4.1 CWQI categorization schema…………………...... 89
4.2 Water quality standards for different water uses…………...... 91
4.3 Format for the data entry on data page of CWQI model ...... 94
5.1 An overview of some important software products for water quality modeling...... 105
5.2 Estimation of deoxygenation constant for selected reaches of river Chenab…… ...... 116
6.1 Summery of water quality analysis of river sampling during low flow season of 2006-7 and 2007-8………………………………...... 141
6.2 Scope, Frequency and Amplitude for different water uses of river Chenab……145
6.3 Model evaluation statistics for calibration and verification of discharge during different months ……...... 149
6.4 Model evaluation statistics for calibration and verification of water levels during different months ……...... 151
xiii
Table Description Page #
6.5 Model evaluation statistics for calibration and verification of salinity during different months ……...... 153
6.6 Variation of salinity in Chakbandi and Faqirian Sillanwali drains during lean flow months………………………..………...... 154
6.7 Characteristics of various point sources of pollutions along river Chenab ...... 158
6.8 Model evaluation statistics for calibration of discharge, water level, DO and BOD during October 2007 ……...... 158
6.9 Model evaluation statistics for validation of DO and BOD during December 2007 ……...... 160
6.10 Management scenarios tested for salinity of river Chenab...... 162
6.11 Management scenarios tested for BOD in the river Chenab...... 164
xiv LIST OF FIGURES
Figure Description Page #
3.1 The Chenab river basin……………………………………………………… ....55
3.2 Layout of Marala barrage…………………………………………………….....59
3.3 Layout of Qadirabad barrage……………………………...... 63
3.4 Layout of Trimmu barrage………………………...... 65
3.5 Map of selected river reach of river Chenab…………………………………....71
3.6 Schematic diagram of selected reach of river Chenab...... 72
3.7 Monitoring cycle developed by UN/ECE …………………...... 73
3.8 Water quality sampling network at river Chenab……………………...... 75
4.1 CCME water quality index formulation…………...... 85
4.2 The start page of CWQI model………………...... 92
4.3 The data page of CWQI model………………………… ...... 93
4.4 The criteria page of CWQI model…………… ...... 95
4.5 The report page of CWQI model……………...... 97
5.1 Subdivisions of water-quality models in common use ...... 101
5.2 Description of different complexity levels of water quality component of MIKE 11……………………………………...... 107
5.3 The input tab of simulation editor in MIKE 11 model…………… ...... 119
5.4 The simulation tab of simulation editor in MIKE 11 model……………...... 120
5.5 The start tab of simulation editor in MIKE 11 model……………...... 121
5.6 The network editor of MIKE 11 model…………………...... 122
5.7 The cross-section editor of MIKE 11 model……………...... 124
xv Figure Description Page #
5.8 The boundary editor of MIKE 11 model…………………… ...... 128
5.9 The hydrodynamic parameter editor of MIKE 11 model...... 129
5.10 The advection-dispersion editor of MIKE 11 model ...... 131
6.1 Flow duration curve for river Chenab at Marala using mean monthly flow data of sixty year period (1947-48 to 2006-07)………………………………. ..137
6.2 Long term average of mean monthly flow in river Chenab at Marala during different months of the year ……………………………………...... 138
6.3 Low flow frequency curve for river Chenab at Marala …… ...... 139
6.4 Graphical presentation of different water quality parameters of river Chenab during low flow months of 2006-7 and 2007-8………………………………. ..142
6.5 Water Quality Indices according to different water uses at selected stations along river Chenab……………………………………...... 146
6.6 Comparison of observed and simulated results of MIKE 11 hydrodynamic module for discharge during lean flow months of 2006-07………...... 148
6.7 Comparison of observed and simulated results of MIKE 11 hydrodynamic module for water levels during lean flow months of 2006-07………………….150
6.8 Comparison of observed and simulated results of MIKE 11 AD module for salinity during flow months of 2006-07…………………………………...... 152
6.9 Simulated salinity profiles for lean flow months of 2006-07 ...... 154
6.10 Comparison of observed and simulated discharge and water level for October 2007…………………………………...... 156
6.11 Calibration of MIKE 11 AD module for October 2007...... 157
6.12 Validation of MIKE 11 AD module for December 2007 ...... 159
xvi Figure Description Page #
6.13 Simulated DO profiles for different months of low flow season 2007-08 ...... 160
6.14 Simulated BOD profiles for different months of low flow season 2007-08...... 161
6.15 Salinity profiles in the polluted river reach under different scenarios...... 163
6.16 Simulated BOD profiles under different management scenarios ...... 165
xvii LIST OF ABBREVIATIONS
µS/cm Micro Siemens per centimeter 0C Degrees Centigrade AD Advection-dispersion ADE Advection-Dispersion Equation asl above sea level BCM Billion Cubic Meter BOD Biochemical Oxygen Demand CCME Canadian Council of Ministers of the Environment Cd Cadmium cfs Cubic Feet per Second Cl Chloride COD Chemical Oxygen Demand Cr Chromium Cu Copper DHI Danish Hydraulic Institute DO Dissolved Oxygen F Fluoride FAO Food and Agriculture Organization ft Foot (feet) GEMS Global Environment Monitoring System GoP Government of Pakistan HD Hydrodynamic HW Headworks IWT Indus Waters Treaty K Potassium km2 Square Kilometers lbs pounds m meter (s) m2/s Square Meter per Second
xviii m3/d Cubic Meters per Day m3/s Cubic Meters per Second meq/l Milli Equivalent per Liter mg/l Milli Grams per Liter MW Mega Watt Na Sodium NEQS National Environmental Quality Standards NGOs Non Government Organizations
NH4–N Ammonical Nitrogen Ni Nickel
NO3–N Nitrate Nitrogen Pb Lead
PO4 Phosphate RSC Residual Sodium Carbonate SAR Sodium Adsorption Ratio TC Total Coliforms TDS Total Dissolved Solids TKN Total Kjeldahl Nitrogen WQ Water Quality WQI Water Quality Index Zn Zinc
xix CHAPTER I
INTRODUCTION
1.1 BACKGROUND OF THE STUDY
Fresh water resources are the most valuable assets of any human civilization.
They serve a pivotal role in overall economy of a country due to inevitable demand of
water in all sectors of life. Surface water quality is influenced by various natural
processes and anthropogenic activities. In many developing countries, wastewater is
disposed into the natural water bodies due to their capacity to assimilate and dilute the
harmful constituents of the effluents. As municipal and industrial demand for freshwater
rises, increasing effluents of low quality are dumped unchecked into the natural water
bodies resulting in further degradation of their water quality. Consequently, human health
and crop yields are being affected or threatened in many cases. “Each water body can
assimilate a certain amount of effluents depending on numerous factors. The water
quality management attempts to protect the uses of water bodies facing the threat of
pollution”. (McBride, 2002; Mohammed et al., 2002; de Azevedo et al., 2000; Somlyódy et al., 1998).
Pakistan is one of the world’s most arid countries, with an average annual rainfall of under 240 mm. Throughout the history, people of this region have adapted to low and poorly distributed rainfall by depending on an annual influx of about 180 billion cubic
meters (BCM), into the Indus river system (including Indus, Jehlum, Chenab, Ravi, Bias
and Sutluj rivers). The rivers flowing in Pakistan largely emanate from the neighboring
countries and are mostly derived from snowmelt in the Himalayas. Alteration of natural
1 flows of the trans-boundary rivers, as happened in the subcontinent of Pakistan and India,
threatens their water quality. The Governments of Pakistan and India signed Indus
Waters Treaty (IWT) in 1960. According to this treaty, Pakistan was authorized to use
water of three western rivers namely Indus, Chenab and Jehlum while India got rights on
the waters of eastern rivers i.e. Ravi, Beas and Sutluj. Soon after the treaty, India constructed a number of storage dams on the eastern rivers that caused severe water shortage in the downstream areas of Pakistan. The most fertile agricultural lands in the
Punjab province of Pakistan were previously irrigated from these eastern rivers. To irrigate these lands after IWT many link canals were constructed in the country that supply water from western to eastern rivers. These link canals traverse through industrial towns and big municipalities and hence face the risk of pollution addition (World Bank,
2005).
The current population of Pakistan is about 161 million and expected to increase up to 208 million in 2025 with about 50 percent of this population living in urban centers.
Urbanization is taking place in the country at the rate of 32.5 percent. (GoP, 2008)
Approximately 59 percent of the total population of Pakistan is covered by sanitation facilities (WHO/UNICEF, 2006). Most of the cities and industries in Pakistan are without wastewater treatment facilities. Large quantities of untreated municipal sewage and industrial effluent are being discharged directly to surface water resulting in serious pollution. Generally, urban wastewater consists of high quantities of sewage (millions of pathogenic microorganisms causing enteric infections), industrial wastes (toxic ions and heavy metals) and carcinogens (Morishita, 1988). The rapid industrialization also
2 adversely effects the environment directly and indirectly. Industrial development
manifested due to setting up of new industries or expansion of existing industrial
establishments results in the generation of industrial effluents. The present method of
transportation of these effluents, their ultimate disposal and treatment facilities for
making effluents innocuous and safe are inadequate, unplanned and suffers negligence
and shortage of funds (Kulkarni, 1979). The net result is large scale pollution of the water
bodies which may act as a source of water supply for domestic use of inhabitants of
localities. This loss of water quality may cause health hazards to human and livestock,
death of aquatic life, crop failure and loss of aesthetics.
The industrial waste and domestic sewage is dumped through surface ditches
(drains) into water bodies of the country. A recently conducted nation-wide wastewater
assessment showed that total waste water supply in Pakistan is 4.6x106 m3/day, and of total 7.85 million m3/day of wastewater (30 percent of the total) is used for irrigating an
area of 32500 hectares. It has also been estimated that 64 percent of the total wastewater
is disposed off either into rivers or in Arabian sea. Similarly, 400,000 m3/day wastewater
is additionally added to canals. These practices threaten both human health and the
environment at downstream and more importantly reduce the effective availability of
Pakistan’s already short water supplies (Ensink et al., 2004). Indiscriminate disposal of sewage and industrial effluent has seriously affected the quality of surface water.
A recent report of WWF (2007) stated that the quantity as well as quality of water resources of Pakistan is strongly affected by tremendous increase in population,
3 urbanization and unsustainable water consumption. There is very little separation of municipal and industrial effluents in the country. Both the effluents flow directly into the nearby natural water bodies (rivers or canals) through open drains. Unfortunately, no surface water quality standards have been established in Pakistan. National standards are available only for wastewater (industrial and municipal effluents) but these are rarely enforced.
In this scenario, the water quality issue in Pakistan has not yet got its due importance. A comprehensive water quality monitoring program is indispensable to assess the water quality status of the national rivers. After the collection of monitoring data on water quality, it is needed to convert it into an understandable format that can be easily interpreted. Another important issue to be addressed is the little attention given to formulate surface water quality standards according to different water uses. National
Environmental Quality Standards (NEQS) established in 1993 are available only for municipal and liquid industrial effluents and do not provide any guideline for the receiving water bodies. PSI (1987) and PCRWR (2002) have drafted water quality standards for drinking and irrigation waters but their enforcement is still pending.
1.2 PROBLEM STATEMENT AND SCOPE OF WORK
Rivers and streams are important component of natural environment. They have many values such as economic (fishing, electricity generation, transport and irrigation), aesthetic (recreation), ecological (biodiversity), water for consumption (water supply for domestic and industrial uses) and conveying wastewater discharges (treated or untreated).
4 To maintain these values and their sustainable use, given water quality standard must be met.
For every use of the river water different set of contaminants or water quality parameters play deterministic role for water quality assessment. For irrigation use dissolved solids (TDS), pH, sodium adsorption ratio (SAR) and residual sodium carbonate (RSC) are the most important. For other uses dissolved oxygen (DO),
Biochemical Oxygen Demand (BOD), carbonaceous oxygen demand (COD), inorganic nitrogen (ammonia and nitrite), phosphorus, suspended solids, hazardous substances, organic pollutants (e.g. petroleum and hydrocarbons) and heavy metals (e.g. mercury and cadmium) are also considered. The contamination by hazardous substance can pose risk to human health in particular via the food chain. However, it becomes more and more difficult to meet such water quality standards because of continuous economic expansion, urban development and growing population pressure.
Without appropriate assessment of existing water quality status, management of water quality is out of question. As a second step after water quality assessment, use of mathematical models can be extremely helpful for detailed analysis of the existing situation and proposition of management options.
The present study is an effort to monitor, assess and model the water quality of the Chenab, a western river of Pakistan. The river traverses through a number of densely populated and industrial cities (e.g. Sialkot, Gujranwala, Gujrat, Hafizabad, Faisalabad,
5 Sargodha and Jhang) and receive effluents through a network of surface drains as depicted in Figure 3.1.
1.3 RESEARCH HYPOTHESIS
The water quality degradation is evident as a result of effluents added in a water body. But the response of water body in terms of its quality for different uses may vary due to numerous factors. These factors include the nature of water body (lake, river, reservoir, sea etc.), water quality status of the receiving water body, availability of flow, alteration in natural flow, quantity and quality of the added effluents, climatic factors and so on. Some of these factors may be the key determinant of water quality in one condition but they may have a little or no impact in a different condition.
The hypothesis of the present was that the flow condition in the river Chenab
(temporal factor) and the distance from entry points of the surface drains (spatial factor) significantly affect the water quality condition of the river in context of its different uses.
To test this hypothesis a comprehensive monitoring program was initiated. The samples were collected from varying locations keeping in view the network of contributing drains. Monthly variations in the flow of river and the contributing drains were observed during low flow months (October to March). The water quality assessment showed the variation of individual water quality parameters and their comparison with the water quality guidelines for different uses. As a readily understood indicator, water quality index, was calculated using CWQI 1.0 model showing the spatial variations of the
6 river water for three intended uses (irrigation, drinking and aquatic life). Simulations with
MIKE 11 model provided detailed profiles of water quality parameters along the river and for different months as well. The hypothesis was tested with measured and simulated values of different water quality parameters.
1.4 OBJECTIVES
The specific objectives of the study are as follows:
• Monitoring and analysis of the water quality status of river Chenab and the
drains polluting it.
• Quantification and assessment of spatial variation of water quality in the
selected reaches of the river during low flow months of the year.
• Calculation of water quality indices along the river for different water uses
and identification of the most polluted reaches.
• Modeling and simulation of both conservative and non-conservative
components of water quality on spatial and temporal basis.
• Study and propose various management scenarios for the reduction of
pollution buildup in the river.
7 CHAPTER II
REVIEW OF LITERATURE
Literature review provides necessary insight and helps in conceptualization of the problems. The research required to complete this project was multidisciplinary in nature.
The following chapter outlines past research and basic principles on low flow analysis, water quality monitoring, water quality indices, and hydraulic and water quality modeling in the streams and rivers. The literature is divided into four main sections i.e. low flow analysis, water quality monitoring and assessment, water quality indices and water quality modeling.
2.1 LOW FLOW ANALYSIS
Low flow is a term used in different meanings by different interest groups. It may be considered as the actual flows in a river occurring during the dry season of the year, it may also be regarded as the length of time and the conditions occurring between flood events (e.g. in erratic and intermittent semi-arid flow regimes). Yet the effects of changes in the total flow regime of a river on sustainable water yield or riverine and riparian ecology may be another aspect of this term.The definition of low flow according to
International glossary of hydrology (WMO, 1974) is as follows:
‘Flow of water in a stream during prolonged dry weather’.
However, this definition does not make a clear distinction between low flows and droughts. Low flow is a seasonal phenomenon, and an integral component of a flow regime of any river. Drought, on the other hand, is a natural event resulting from a less than normal precipitation for an extended period of time (Smakhtin, 2001).
8 There are natural and anthropogenic factors which influence the various aspects
of the low-flow regime of the river. The natural factors include the distribution and infiltration characteristics of soils, the hydraulic characteristics and extent of the aquifers, rate, frequency and amount of recharge, evapotranspiration rates from the basin, distribution of vegetation types, topography and climate. On the other hand, anthropogenic factors include groundwater abstraction within the sub-surface drainage area, artificial drainage of valley bottom soils for agricultural or building construction purposes, changes to the vegetation regime in valley bottom areas through clearing or planting, deforestation, urbanization, direct river abstractions for industrial, agricultural or municipal purposes, direct effluent flows into river channels from industrial or municipal sources, irrigation return flows from agricultural fields and construction of dams and subsequent regulation of a river flow regime.
2.1.1 Low-Flow Measures and Indices and Their Estimation from Stream-Flow Time Series
Low-flow regime of a river can be analyzed in a variety of ways dependent on the type of data initially available and the type of output information required. Consequently
there exist a variety of low-flow measures and indices. The term ‘low-flow measure’ used
here, refers to the different methods that have been developed for analyzing, often in
graphic form, the low-flow regime of a river. The term ‘low-flow index’ is used
predominantly to define particular values obtained from any low-flow measure
Sometimes, it is however rather difficult to separate one from another (Smakhtin, 2001).
In the present study, two measures i.e. flow duration curve (FDC) and low flow
frequency curve (LFFC) and one index (i.e. 30Q10), have been used for the analysis of
9 low flow in river Chenab. A brief overview of these low flow measures and index are as
follows:
A) Low Flow Measures
Flow duration curve (FDC): A flow duration curve (FDC) is very useful methods of displaying the complete range of river discharges from low flows to flood events. It is a relationship between any given discharge value and the percentage of time that this discharge is equaled or exceeded, or a relationship between magnitude and frequency of stream-flow discharges (Smakhtin, 2001).
To construct a FDC, the flow time series data is arranged in decreasing order of magnitude. The data is ranked and plotted against percentage of time based on the total number of time steps in the record. A FDC is usually plotted on a log-normal scale which allows FDCs in some cases to be linearized. The FDC may be constructed using different time resolutions of stream-flow data: annual, monthly or daily. Daily flow time series, when used for FDC, provide the most detailed information regarding duration characteristics of a river. FDCs may also be constructed based on data with some other time intervals, e.g. m-day or m-month averages. Details on FDC construction and interpretation are available in literature (e.g. Searcy, 1959; Institute of Hydrology, 1980 and McMahon and Mein, 1986).
The ‘low-flow section’ of a FDC is of most interest for low-flow studies. It is that part of FDC ‘which may be arbitrarily determined as part of the curve with flows below
MF (which corresponds to the discharge equaled or exceeded 50% of the time—Q50).
10 FDC illustrates the frequency distribution of flows in a stream with no regard to their sequence of occurrence (Smakhtin, 2001).
The present study included the construction of FDC for river Chenab using mean monthly flow data at Marala for sixty year period (from 1947-48 to 2006-07). In order to relate FDC with the sequence of occurrence of stream flow discharges, average (sixty year) monthly flow data during different months of the year were also plotted in connection with the information obtained from FDC.
Low-flow frequency analysis: Unlike the FDC, which shows the proportion of time during which a flow is exceeded, a Low-flow Frequency Curve (LFFC) shows the proportion of years when a flow is exceeded (or equivalently the average interval in years
(‘return period’ or ‘recurrence interval’) that the river falls below a given discharge).
The annual flow minima (daily or monthly minimum discharges or flow volumes), which are extracted from the available original continuous flow series (one value from every year of record) are normally used to construct a LFCC. Similarly to
FDCs, LFFCs may also be constructed using the flow minima series of different averaging intervals (Smakhtin, 2001). In the case of daily data, the minima of 1, 3, 7, 10,
15, 30, 60, 90, 120, 150, 180 or 183, 273 and 284 may be analyzed (e.g. Characteristics of low flows, 1980; Musiake et al., 1984; McMahon and Mein, 1986; FREND, 1989;
Zalants, 1992 and Harris and Middleton, 1993). In the case of monthly data, averaging intervals of 1, 3, 6 and 9 months may be selected (Midgley et al., 1994).
11 The available observed flow records are normally insufficient for reliable frequency quantification of extreme low-flows events and, therefore, different types of theoretical distribution functions are used to extrapolate beyond the limits of ‘observed’ probabilities and to improve the accuracy of low-flow estimation. The ‘true’ probability distributions of low flows are unknown and the practical problem is to identify a reasonable ‘functional’ distribution and to quantify its parameters. The procedure includes fitting several theoretical distribution functions to observed low-flow data and deciding, by statistically based and graphically based tests, which distribution best fits the data. Among the distribution functions most frequently referred to in the literature in connection with low-flows are different forms of Weibull, Gumbel, Pearson Type III, log-normal distributions (Smakhtin, 2001).
B) Low-Flow Index
Numerous specific indices may be obtained from a LFFC. In USA, the most widely used indices are 7-day 10-year low flow (7Q10) and 7-day 2-year low flow (7Q2), which are defined as the lowest average flows that occur for a consecutive 7-day period at the recurrence intervals of 10 and 2 years, respectively (e.g. Characteristics of low flows, 1980). Some studies refer to different similar indices, e.g. 3 day 20 year low flow
(Hutson, 1988). A number of reports produced by the USGS contain a variety of minimum flows: annual means and extremes for selected periods ranging from 1 to 183 days and for recurrence intervals ranging from 2 to 50 years ( Hughes, 1981; Armentrout and Wilson, 1987; Zalants, 1992; Cervione et al., 1993; Giese and Mason, 1993 and
Atkins and Pearman, 1995).
12 In Russia and Eastern Europe, the widely used indices are 1-day and 30-day summer and winter low flows (either means, or flows with an exceedence probability of
50, 80, 90, 95% (Vladimirov, 1990; Balco, 1977; Szolgay, 1977; Walkowicz, 1978;
Amusja et al., 1988; Yevstigneev, 1990; Sakovich, 1990).
Low-flow frequency indices are widely used in drought studies, design of water supply systems, estimation of safe surface water withdrawals, classification of streams’ potential for waste dilution (assimilative capacity), regulating waste disposal to streams, maintenance of certain in-stream discharges, etc. (Chiang and Johnson, 1976; Refsgaard and Hansen, 1976; Male and Ogawa, 1982; Aron and Emmanuel, 1982; Biswas and Bell,
1984; Riggs, 1985; Paulson and Sanders, 1987; Cumming Cockburn Ltd, 1990).
LFFC for river Chenab at Marala was plotted using annual time series of low flow minima (based on 30 days average) of sixty years (from 1947-48 to 2006-07). Log- normal and log Pearson type III distribution functions were used to extrapolate the data.
Furthermore 30-day, 10-year low stream-flow (30Q10) index to determine mixing zones downstream from point-source discharges was calculated from LFFC. The 30Q10 is defined as the stream-flow below which the annual 30-day minimum falls in 1 year out of
10 as a long-term average. The recurrence interval of the 30Q10 is 10 years; the chance that the annual 30-day minimum flow will be less than the 30Q10 is 10 percent in any given year.
13 2.2 WATER QUALITY MONITORING AND ASSESSMENT
2.2.1 Global Perspective
The salinity of many rivers in US has increased through natural and anthropogenic activities. As compared to the other river basins, more salts are added to those located in western region of the United States particularly in the Colorado river basin. The salinity of river water depends mainly on the stream flow. Higher flow contains more salts but at the same it also causes substantial dilution resulting in lower salt concentration and vice versa (U.S. Department of the Interior, 2003).
Heejun (2005) conducted a water quality monitoring study of Han River and its tributaries in Seoul, Korea. Spatial and temporal variations of eight selected water quality parameters were examined for the 26 stations located within the city of Seoul from 1993 to 2002. Eight parameters were measured including water temperature, pH, dissolved oxygen (DO), biochemical oxygen demand (BOD), chemical oxygen demand
(COD), suspended solid, total nitrogen and total phosphorus. There was no significant increase or decrease in water quality parameters at four upstream stations. A striking longitudinal variation of all water quality parameters was detected at eight stations located in the midle reach of river Han. The water quality declined dramatically in the middle of the river where it received inputs from the polluted tributaries. All water quality parameters, except pH, exhibited better conditions in the main river group than its counterpart tributary group. The 26 stations were grouped into three main clusters based on water quality conditions: (1) the main river and its tributaries with relatively good
14 water quality, (2) the tributaries with medium water quality, and (3) the tributaries nearby
industrial sites that were heavily polluted.
Cruz (1997) presented the water quality status of Pasig river in Phillipine based
on a monitoring program. Estimation of pollution loadings from different sources and
different management approaches were also discussed in this article. The Pasig river runs through five cities and four municipalities. The waters of the Pasig river were analyzed
twice a month and the pollution levels were determined. The program used ten sampling
stations along the Pasig river system (including San Juan river, Marikina river, Manila
Bay and Laguna de Bay) to gauge the degree of pollution based on BOD, DO, coliform
bacteria counts, salinity, phosphates, nitrates and others. Traditionally, the municipalities
upstream were fishing communities relying mostly on the Pasig river and Laguna de Bay,
while the settlements downstream experienced rapid urbanization with the influx of trade
from other provinces and countries. Industrial pollution accounted for 45 per cent of the
total pollution in the Pasig River. About 315 of the 2,000 or more factories situated in the
river basin were determined as principal polluters of the river, dumping an average of 145
tons of BOD per day. This was established by determining the suspended solids in their
treated and untreated waste-waters. According to the records, textile and food
manufacturing industries were the greatest water polluters among those considered in the
study. The Pasig river has been historically known for its recreational and transport
functions. With its gradual degeneration, this aspect has been reduced to use for rowing
by some enthusiasts only.
15 The water quality of Piracicaba river in Brazil was determined on spatial and
temporal basis by Krusche et al. (1997). The area of river catchment is about 12400
square kilometers and included in the most urbanized and developed state (Sao Paulo) of
the country. The water quality was determined in terms of DO, BOD, total coliform (TC)
and nitrate. Nine sampling sites were selected along the river for water quality analysis.
The results showed that river water quality was deteriorated spatially in the downstream
direction in terms of all the selected water quality parameters. For temporal analysis,
water quality data of eighteen years was used. The results indicated a decrease in DO and
increase in BOD and TC levels at all sampling sites with the passage of time.
Spatial and temporal assessment of water quality of Axios/Vardar river was
carried out by Milovanovic (2005). The river is of trans-boundary type and drains
portions of many countries located in southeastern Europe e.g. Balkan, Serbia,
Macedonia and Greece. First the sources of pollution were identified and analyzed. Then
the long term data on river water quality was used to determine water quality trends. The data was obtained from 25 year (1979 to 2003) monitoring of 22 sampling sites. Water samples were collected on monthly basis and analyzed for BOD, nitrate, Cd, nitrite, Pb, ammonia, Cr, Zn and total phosphorus. This long term data was divided into five subsets, each of five years period i.e. 1979–83, 1984–88, 1990–95 and 1996–2003. Average values of water quality parameters were calculated for these subsets to determine water quality trends on spatial and temporal basis. For subjective interpretation of the data, interviews were also conducted with many scientists. The results showed that the river water quality was deteriorated due to pollution of varying nature and generated from
16 different sources. Wastewater with high heavy metal contents was added into the river mainly from smelter and fertilizer industries in the nearby cities. Other sources included untreated industrial wastewater being discharged into the river. The nutrient pollution was stemming from agricultural runoff and the point source pollution consisted of a number of illegal landfills and untreated domestic sewage of the surrounding municipalities.
Different multivariate statistical techniques are commonly used for the anlaysis of complex water quality data and understanding the water quality trends on spatial and temporal basis. Shrestha and Kazama (2007) applied statistical techniques of cluster analysis (CA), principal component analysis (PCA), factor analysis (FA) and discriminant analysis (DA) to evaluate water quality of Fuji river basin in Japan. The data was generated during eight years (1995-2002) monitoring of twelve water quality parameters of the river samples collected from thirteen different sites (total 14976 observations). The sampling sites were divided into three clusters through CA based on the their similar characteristics of water quality. These clusters were named as relatively less polluted (LP), medium polluted (MP) and highly polluted (HP) sites. Statistical techniques of FA and PCA were applied on these three clusters of sampling sites. The results showed total variances of 73.18 percent in LP, 77.61 percent in MP and 65.39 percent in HP sites. It was indicated from FA that the variations in water quality of three clusters of the sampling sites were caused by different parameters: 1) in LP areas these were river flow, temperature and organic pollution from point sources 2) organic wastewater from point and domestic sources caused more variation in MP areas and 3)
17 domestic sewage along with nutrients from agricultural activities were found responsible for water quality variations in HP areas. The spatial and temporal trends were better analyzed using DA technique. From DA, the data was significantly reduced to only six parameters (discharge, BOD, EC, temperature, DO, and nitrate-nitrogen) in temporal analysis and seven parameters (pH, EC, BOD, discharge, temperature, nitrate-nitrogen and ammonical-nitrogen) in spatial analysis. The results of DA afforded correct assignments of more than 85 percent and 81 percent respectively in temporal and spatial analysis of three clusters of sites. It was also found that few parameters were responsible for large water quality variations in the basin. The study illustrated that the multivariate statistical methods efficiently analyzed complex data on water quality, identified the factors affecting pollution and assessed water quality on spatial and temporal basis.
In a similar study Singh et al. (2005) applied CA, FA, PCA, and DA to analyze water quality variations in Gomti river in India. The data set consisted of total 9792 observations consisted of 34 water quality parameters analyzed for water samples collected from eight sampling sites during three years (1999-2001). In CA, these eight sampling sites were divided into three significant regions based on their similar characteristics. The regions or clusters were named as upper catchment (UC), middle catchment (MC) and lower catchment (LC). The multivariate statistical techniques of FA and PCA were applied to these three clusters and the results showed total variances of
74.3 percent in UC, 73.6 percent in MC and 81.4 percent in LC. The water quality parameters identified to be responsible for water quality variations were grouped into trace metals leaching from industrial waste sites, organic pollution generated from municipal and industrial sources, nutrients stemming from agricultural runoff and EC,
18 hardness, solids and alkalinity. During spatial and temporal analysis, the complex data
set was most efficiently reduced through DA. It used only five parameters (Sodium,
temperature, chloride, total alkalinity and potassium) in temporal analysis and ten
parameters (Cl, BOD, Zinc, F, TKN, ammonium nitrate, pH, PO4, F, ammonical-nitrogen and discharge of the river) in spatial analysis. The results of DA afforded correct assignations of more than 94 and 97 percents respectively in temporal and spatial analysis of three clusters. The study confirmed the usefulness of DA in reducing the dimensionality of large and complex data on water quality.
Principal Component Analysis (PCA) applied on water quality data of Passaic river in New Jersey by Bengraine and Marhaba, (2003). The data consisted of analytical results of physical, chemical and biological parameters of the water samples collected from twelve sampling sites along the river during year 1998. The multivariate statistical
technique of PCA helped in finding the factors responsible for spatial and temporal
variations of water quality. The trends of different water quality parameters were
determined during the study i.e. temperature, concentration of soluble contents, nutrients
and organic matter. The results of spatial analysis showed that two sampling stations
were more polluted than the others which possibly received pollution from point and non-
point sources. The study concluded that regular monitoring and application of statistical
tools on water quality data can be helpful for understanding complex water quality
problems.
19 A case study of Rous river catchment was carried out by Eyre and Pepperell
(1999) to monitor the river water quality on spatial basis. This catchment is located in the
northern part of New South Wales (NSW) in Australia. The study involved the collection of water quality data from a large number of sampling sites over a short period of time and suggested that despite a few potential limitations, the spatially intensive water quality monitoring methodology should allow environmental managers to, rapidly and cost- effectively (in the long term) identify the point and non-point source impacts on water quality. It was found from the results that three point sources i.e. the Murwillumbah
Sewage Treatment Plant, a dairy shed and horse stables had the largest impact on water quality in the Rous river catchment during base flow conditions. The poorest water quality in the Rous river catchment, due to non-point source inputs, was associated with cane land, which had elevated total nitrogen, total particulate nitrogen, and dissolved organic nitrogen concentrations and temperatures that were significantly greater than other land uses. Highly oxidized nitrogen concentrations were associated with bananas, most likely due to leaching of Nitrogen-fertilizers. The oxidized nitrogen concentrations in the pristine areas appeared high because oxidized nitrogen concentrations were low in other parts of the catchment (excluding horticulture areas) due to algal uptake and removal of inorganic nutrients. At the time of sampling, low flows were reflected by the dominance of in-stream processes which had converted most of the inorganic nutrients to organic nutrients. These findings evoked an immediate management response, where
Environmental Health Officers were sent into the field to inspect the dairy shed and horse stables. This was in contrast with previous routine water quality studies in Rous river that identified a water quality problem, but not the exact causes; as such there was no
20 immediate management intervention. It was concluded from the study that the long-term management efforts in the Rous River catchment need to be firstly directed at reducing point source inputs (particularly nitrogen), secondly at reducing non-point source inputs
(particularly nitrogen) from cane land and bananas and thirdly at improving the catchment water quality for human health by reducing direct cattle access to streams.
Xie et al. (2006) proposed a new approach for monitoring water quality based on quantitative remote sensing in Huangpu river, Shanghai. The inversion models for two typical water quality parameters (DO and secchi disk (SD)) were developed. Based on the derived models and multi-temporal remote sensing imagery, the spatial temporal analysis for the water quality variation was conducted. The results showed that the proposed models can detect effectively the temporal and spatial distribution of water quality.
The spatial and temporal trends of water quality, in a segment of San Antonio river in USA were analyzed by Anderson et al. (2007). Water quality as a function of land use was examined in the upper San Antonio river in the city of San Antonio, Texas.
Five selected water sampling sites representative of different point and non point pollution sources were spread over a distance of 3.2 km. Surface grab sampling was performed on a monthly basis between November 2004 and April 2005 excluding
December. The evaluated water quality parameters were pH, dissolved oxygen, temperature, total dissolved solids (TDS), total nitrate–nitrogen, total orthophosphate, turbidity, alkalinity and hardness using standard analytical protocols. Results were statistically analyzed by MANOVA. Findings were compared to state (Texas
21 Environmental Quality Commission) and/or federal (U.S. Environmental Protection
Agency) limits to establish whether or not parameters were in compliance with those
standards or guidelines. Of the routine water quality parameters examined, only turbidity
and nitrate–nitrogen exceeded specific standards or guidelines in the sampled segment of the San Antonio river. Turbidity and nitrate–nitrogen also showed spatial and temporal trends, which were possibly, affected by land use and local precipitation patterns.
An effort to characterize the water quality of Amu Darya river was made by Crosa et al. (2006). The river is located in central Asia and serves as an important water resource of the region. The study examined the water quality variations in the river with respect to time and space. The extent and causes of degradation in water quality were also investigated. The results of the study showed high salinity levels in the river water with sulphate and chloride as main constituents. High salt concentrations were noted at the sampling sites located in upstream segment of the river but the salinity level was not higher than the acceptable limits. The results of temporal analysis showed that the river water can be withdrawn below 450 km point only from May to September while maintaining salinity level under the acceptable limits. The low drainage density of river basin and snow melt in the upper catchment were identified as two main reasons for salinity variations on temporal basis. The salt concentrations were dependent on the runoff generated from irrigation and land washing activities during lean flow period.
Old et al. (2006) investigated the relationship between hydrological characteristics
of a catchment and urbanization. The understanding of possible hydrological changes in
22 response of urbanization is very important for the management of catchment. It was found during the study that urbanization may significantly affect the amount and patterns of runoff and the composition of suspended and dissolved water constituents. These factors may cause changes in flows, salt concentrations, water quality and habitats of aquatic life in the rivers or lakes receiving the runoff. The study area was located in the
Bradford catchment in England within which central part was highly urbanized. The data was collected from different monitoring stations located in the central part of catchment along Beck river. The samples were also collected from river stations upstream and downstream of the highly urbanized part of the catchment. The data consisted of river flow, turbidity and specific conductance, collected with a frequency of fifteen minutes during one year period (June 2000 to June 2001). The results from data analysis showed that high total rainfall in the catchment resulted in high flow and more suspended sediments in the river. The pattern of flow and sediments in the whole catchment was flashy and turned flashier in the downstream reaches. The part of Bradford subcatchment imparted large amount of suspended sediments in the downstream reach of river Aire.
The runoff in the catchment might be interrupted by the combined sewer system before reaching the river. The results also showed that highly urbanized part of the Bradford catchment was a main source of solutes transport in Beck river. It was concluded that effective monitoring and management of river can be achieved by close understanding of spatial and temporal changes in the river flow and both suspended and dissolved sediments.
23 The water quality of Pisuerga river was analyzed by Vega et al. 1998. The river is
a part of Duero river basin located in the centre-north of Spain and crosses the town of
Valladolid (major industrial centre of the region with a population of about 400000
persons). Municipal wastewater of about 57 million m3 is directly discharged into the river with out treatment. Moreover, although big industries settled in the area purify their wastewater, small industries are suspected to discharge residues into the river. The combination of both a high population density in the area and an extreme continental climate caused river hydrology and hence river pollution to be strongly influenced by seasonality. In this context, 22 physico-chemical variables were analyzed in the water samples collected every three months for two and a half years from three sampling stations located along a section of 25 km of a river affected by man-made and seasonal influences. Exploratory analysis of experimental data was carried out by box plots,
ANOVA, display methods (principal component analysis) and unsupervised pattern recognition (cluster analysis) in an attempt to discriminate sources of variation of water
quality. Spatial (pollution from anthropogenic origin) and temporal (seasonal and
climatic) sources of variation affecting river water quality and its hydrochemical
characteristics were differentiated and assigned to polluting sources. An ANOVA of the rotated principal components has demonstrated that (i) mineral contents were seasonal and climate dependent, thus pointing to a natural origin for this polluting form and (ii) pollution by organic matter and nutrients originated from anthropogenic sources, mainly as sewage.
24 A design of a water quality monitoring network for the Limpopo River basin in
Mozambique was proposed by Chilundo et al. (2008). According to this study measurement of chemical, physical and biological parameters is important for the characterization of streams health. The study suggested cost-effective and targeted water quality (WQ) monitoring programs for proper assessment, restoration and protection of river systems. The study area consisted of a region prone to severe droughts and prone to anthropogenic and natural driven processes. Hence, physico-chemical, biological and microbiological characteristics at 23 sites within the basin were studied in November
2006 and January 2007. The physico-chemical and microbiological samples were analyzed according to American Public Health Association (APHA) standard methods, while the biological monitoring working party method (BMWP) was used for biological assessment. The assessment of the final WQ condition at sampled points was done taking into account appropriate indexes, the Mozambican standards for receiving waters and the
WHO guidelines for drinking WQ. The assessed data indicated that sites located at proximities to the border with upstream countries (Botswana, South Africa and
Zimbabwe) were contaminated with heavy metals. The Elephants subcatchment was found with a relatively better WQ, whereas the Changane subcatchment together with the effluent point discharges in the basin were found polluted as indicated by the low dissolved oxygen and high total dissolved solids, electric conductivity, total hardness, sodium adsorption ratio and low benthic macroinvertebrates taxa. Significant differences
(p < 0.05) were found for some parameters when the concentrations recorded in
November and January were tested, therefore, indicating possible need for monthly monitoring of WQ. From this study it was concluded that a systematic WQ monitoring
25 network composed of 16 stations would fit the conditions of the LRB. Additional research at a Basin scale was also recommended to identify the major sources of pollution, their transport and impacts to the downstream ecosystem.
Spatial and temporal variations in the water quality of Alberche River, Spain were examined during two consecutive years. Principal component analysis was used to analyze the environmental factors associated with the physico-chemical variability. The first principal component corresponded to the variation of water solute content along the course of the river. In general, the conservative parameters alkalinity, sulphate, calcium, and chloride gradually increased in a downstream direction. However, nutrient variables showed marked differences depending on the location and the season. Two sections can be distinguished in the river. The first in an upstream zone with little human occupation, a siliceous substrate produces waters with low ionic composition and few nutrients.
However, the second, downstream zone, featured high levels of phosphate and dissolved inorganic nitrogen in summer and to a lesser extent in spring. The variation in nutrient content in this section of the river could be attributed to anthropogenic sources since in these seasons the presence of holiday-makers lead to a high population density in the residential buildings and recreation areas. Thus, in this area, seasonal increases in human activities at some locations cause high levels of nutrients (Perona et al., 1999).
Based on surveys and chemical analyses Ma et al. (2008) performed a case study of the surface water and groundwater quality in the Wuwei basin to understand the sources of water pollution and evaluation of water quality of Shiyang river in China.
26 Concentrations of major chemical elements in the surface water were related to the
distance downstream from the source of the river. The surface water quality in the
upstream reaches was good but the river from Wuwei city to the Hongya reservoir was
seriously polluted. There were 23 wastewater outlets that discharging a total of 22.4×106 m3/year wastewater into Shiyang river, which, combined with a reduction of inflow
water, were found to be the major causes of water pollution. It was concluded from this
study that the consumption of water must be decreased until it reaches the sustainable
level permitted by the available resources in the whole basin, and discharge of wastes
must also be drastically reduced.
2.2.2 Local Perspective
A surface water quality monitoring program was initiated in October 2006 by the
Directorate of Land Reclamation, Punjab, to tackle the alarming scenario of water quality
of rivers, canals and drains, which are receiving pollution load from different industries
and cites. In February 2007, twenty five samples from all the five rivers (Chanab, Indus,
Ravi, Sutluj and Jehlum rivers), 75 samples from 23 canals of the Punjab were collected.
These samples were analyzed for water quality parameters like pH, EC, SAR and RSC
and heavy metals like Cu, Ni, Pb and Zn in the laboratories of Directorate of Land
Reclamation in Lahore.
During this monitoring program, six samples were collected from the Chenab
River. The analytical data values are presented in Table 2.1. An increasing trend in the
irrigation parameter is observed as the river flows from Marala to Trimmu head works.
The values of pH, EC, SAR and RSC at four head works i.e. Marala, Khanki, Qadirabad
27 and Trimmu, at the River Chenab varied from 7.91 to 8.13, 0.34 to 0.98, 1.10 to 4.81 and
0.0 to 0.40 respectively. A similar trend in the case of trace metals was observed. The concentration of trace metals (Ni, Pb and Zn) ranged in safer limit. The water chemistry of river Chenab showed abrupt changes in pH, EC, SAR, RSC and Ni values at the sites where Faqirian Silanwali (FS) drain joins the river. The presence of Ni in the River
Chenab indicates the significant effect of this drain. This effect is neutralized to some extent as the river flows downstream towards Trimmu headworks (GoP, 2007).
Table 2.1: Location of sampling points and analytical data of Chenab river.
Sampling Analysis Parameters Trace metals
Points EC RSC Cu Ni Pb Zn pH SAR (dS/m) (me/l) (mg/l) (µg/l) (µg/l) (mg/l)
Marala Headworks 7.91 0.34 1.1 0 0.3 0 0.17 0.43
Khanki Headworks 8.1 0.38 1.17 0 0.26 0 0.19 0.4
Qadirabad Headworks 8.13 0.41 1.32 0.1 0.29 0 0.19 0.41
Trimmu Headworks 8.12 0.98 4.81 0.4 0.46 0.12 0.29 0.56
Avgerage 8.01 0.53 2.1 0.13 0.33 0.03 0.21 0.45
Three major rivers of Pakistan i.e. Indus, Ravi and Chenab were monitored at the selected locations by Institute of Environmental Engineering and Research (IEER) under
Global Environmental Monitoring System (GEMS) of United Nations Environment
Program (UNEP) for ten years (1977 to 1987). The sampling points at these rivers are given in Table 2.2. The results of ten years monitoring revealed that the water quality of river Chenab at selected sampling station was fit for irrigation (Ahmad, 1988).
28 Table 2.2: Sampling stations at various rivers of Pakistan.
Sr. No. Rivers City Sampling Station
1 Indus Hyderabad Kotri Barrage
2 Chenab Faisalabad Gogera Branch
3 Ravi Lahore Syphon
Lahore Balloki Headwork
The sources of wastewater mixing in the rivers of Pakistan were identified by
Hussain (1996). The municipal wastewater pollution loads were obtained from Punjab
Public Health Engineering Department or calculated on the basis of population of the city. Pollution loads from the industries were collected from Directorate of Industries,
Punjab. The results revealed that river Chenab receives pollution loads from major cities as shown in Table 2.3
Table 2.3: Industrial effluent discharging into river Chenab.
Cities Discharge (m3/s) BOD (tons/day) TDS (tons /day)
Gujrat 1.53 130.3 892.2
Sialkot 0.60 24.9 3607.5
Sargodha 0.16 13.4 64.8
Faisalabad 1.33 40..6 234.6
Jhang 0.21 6.3 42.9
Multan 1.69 42.2 1339.3
Muzaffargarh 0.09 2.4 24.8
Total 5.62 260.10 6213.10
29 Also the municipal pollution from various cities of Punjab to river Chenab was
estimated to an average discharge of 7.34 m3/s with total load of BOD and TDS as 185 and 744 tons /day respectively.
Ahmad and Ali (2000) conducted a study to evaluate the water quality of river
Ravi in Pakistan. Two sampling points (Syphon and Balloki headworks) were selected at
90 km apart. Sampling was done once a month. Sampling for river at Balloki headworks was done after 24 hours of the sampling at siphon which is located at upstream of the other point. Water samples were collected in five litter polythene containers. The results suggested that there was high variation in the flow with time during the year. The flow also varied with the length of the river due to link canals discharging into the river and water withdrawals at headworks through canals. Dissolved oxygen concentration varied within a year due to temperature and flow variations. Their values show a decreasing trend indicating increase in pollution. The study concluded that the flows in the river were highly variable with time which resulted in high variations of the pollution parameters in the river. A decreasing trend in DO levels and an increasing trend in BOD,
TDS, Total and feacal coliform were observed over the time. The water quality of river
Ravi met the chemical water quality of requirement for irrigation. However the water quality did not meet the coliform and feacal coliform criteria for most of the water uses.
Waste water assimilative capacity of the river Ravi was studied by Tariq and Ziai
(1980). The purpose of this study was to determine the assimilative capacity (allowable
BOD load) of the river Ravi based on minimum average seven consecutive days flow
30 with a probability of occurrence once in ten years. Five sampling locations were selected
within a river stretch of 76 km. Regular sampling surveys over the low flow period
commencing from the month of September to late May, were conducted to get the
maximum, average and minimum values of BOD and DO. The Streeter-Phelp’s equation
was selected as the basis for the assessment of organic pollution and DO of the stream as
the indication of the extent of the stream pollution and attainment of the water quality
objective. Classical Streeter-Phelp’s model for routing of DO deficit profile is given as:
−K t K1L0 1 −K2t −K2t D = (e − e ) + D0e (2.1) K 2 − K1
Where
D = DO deficit at time t.
D0 = initial deficit of DO.
K1 = de-oxygenation or de-aeration coefficient.
K2 = re-aeration coefficient
t = time of travel.
The allowable pollution load L0 that the stream may absorb is given as:
K 2 K1tc = L0 Dce (2.2) K1
Where
Dc = maximum DO deficit or design dissolved oxygen deficit. tc = time required to reach the critical point.
The study concluded that
31 • Assimilative capacity of the river Ravi based on present population organic pollution
load was 65,000 pounds of BOD per day.
• Wastewater discharging to the river system should contain not more than 37 mg/l and
25 mg/l of BOD by the year 1986 and 2000, respectively, based on an average BOD
of 230 mg/l.
• Based on hypothetical sewage treatment plant located near main outfall, wastewater
discharging to the river system should contain not more than 50 mg/l and 40 mg/l of
BOD based on 1986 and 2000 years flows, respectively.
• To protect the quality of the river, pollution loads emanating from sewered areas
should be limited and for 1980 the amount of BOD should be less than 65,592 pounds
per day.
• Sewage treatment plant at the first stage of construction should provide 70 percent
treatment of BOD by the year 1986 and should be upgraded to 80 percent in the
second stage by the year 2000.
• Based on the study the average river discharge of 5000 cusec has an assimilative
capacity of 349287 pounds of BOD per day. Whereas BOD pollution load by the year
1986 (as projected by the study) will be 383356 pounds of BOD per day. The study
recommended the seasonal treatment plants to operate during low flow and very low
flow periods.
Conditions of the stream by the provision of hypothetical sewage treatment plant effluent located at main outfall were investigated for 1986 and 2000. The results showed that by the end of 1986 the sewage treatment plant operated at 70 percent efficiency gave a DO level of 4.3 mg/l at the end of one day time of flow from main outfall. Similarly
32 this theoretical treatment plant operated at 80 percent efficiency gave a DO level of 5.1
mg/l at the end of the year 2000 after one day time of flow from main outfall. Both of these DO levels were well above the standard DO level of 4.0 mg/l.
2.3 WATER QUALITY INDEXING
The next step after collecting the data on water quality is to convert it into an understandable format that can be easily interpreted. For this purpose Water Quality
Indices (WQIs) are efficiently used and serve as a tool to communicate and translate data on water quality (Ball and Church, 1980). The communication of water quality data is especially challenging when the intended audience for the water quality data is general public who is not directly interested in water quality data. They are more interested in the information that the water quality data conveys and are even more interested in the knowledge that follows from the information (Khan et al., 2003). The water quality data are usually not available in simply understandable form. Their complex nature make it difficult to be reviewed by untrained people for example general public, political decision makers and water managers having no technical background. To fill this gap of communication, various water quality indices have been developed which reduce the large water quality data into easily interpretable values (Couillard and Lefebvre, 1985).
The water quality indices are very useful to reduce large data set on water quality into a single number that can objectively interpreted. These data sets may be composed of physical, chemical and biological parameters The WQIs can be used to compare spatial and temporal variations in water quality and to evaluate the integrated impact of
33 individual parameters on overall water quality. These indices can also illustrate the trends in water quality of the stream with reference to its intended uses (Gupta et al., 2003).
As WQIs reduce and summarize the raw data, they convey less information about the water quality than the original large data set. For this reason WQIs are usually calculated for comparing the water quality of different sites and to get an overall water quality condition of a water body. They are less effective in making site specific decisions on water quality. In short, an index is a useful tool for “communicating water quality information to the general public and to legislative decision makers” but it is not
“a complex predictive model for technical and scientific application” (McClelland, 1974).
As a comprehensive indicator, WQI provides overall summaries of water quality and possible trends on scientific basis (Kaurish and Younos,, 2007). Horton (1965) was the first scientist who proposed the concept of using indices for water quality analysis.
The usefulness of indices as an evaluation tool and their ability to communicate complex information in simple manner, made them widely accepted for water quality management. Many researchers (e.g. Brown et al., 1970; Prati et al., 1971; Dinius, 1972,
1987; Walski et al., 1974; Landwehr, 1979; Dunnette, 1979; Bhargava, 1983, 1985;
Smith, 1987, 1989, 1990; Swamee and Tayagi, 2000; Said et al., 2004; Kaurish and
Younos, 2007 etc) have developed their own rating schemes during the last four decades.
Some of the water quality indices that have been frequently employed in public domain for the purpose of water quality assessment are the National Sanitation
34 Foundations’ WQI, British Columbia Water Quality Index (BCWQI), Canadian Water
Quality Index (CWQI), Oregon WQI, and the Florida Stream WQI. (Said et al., 2004).
The British Columbia Ministry of Environment, Lands and Parks in Canada have developed the BCWQI. This index has been developed by great effort, over a long time span (Zandbergen and Hall, 1998). In 1997, the Water Quality Guidelines Task Group of the Canadian Council of Ministers of the Environment (CCME) decided to review various existing techniques for the calculation of WQIs. Their efforts led to the development of a standardized system i.e. a unified water quality index that has been employed in all parts of Canada for the assessment of water quality (CCME, 2001).
The Canadian Water Quality Index (CWQI) used the concept of BCWQI. The
CWQI employed three factors namely scope, frequency and amplitude and their values were scaled from zero to 100. These factors denoted variance from predefined objectives or permissible limits for individual water quality parameters. By combining three factors of variances a vector was obtained which showed the calculated value of index. The objectives may refer to as water quality guidelines used in Canada however site specific guidelines can also be used while calculating the index for non Canadian sites. The value of index was set to range between zero and 100. The index value equal to zero or close to it indicate poor water quality and contrarily excellent water quality is depicted from the index value near 100 (CCME, 2001; Khan et al., 2003). The index was subsequently revised to overcome the problems pointed out in the formulae used for estimation of frequency and amplitude. The revised index have been employed by Khan et al. (2003) to
35 analyze the water quality trends in three selected watersheds of Atlantic region i.e.
Mersey river, the Point Wolfe river, and the Dunk river sites.
Pesce et al. (2000) used numerous water quality parameters in the calculation of
WQIs to study the water quality variations on spatial and temporal basis. The selection of
water quality parameters was made according to the recommendations of GEMS/Water
UNEP program. Twenty parameters were used for the calculation of two WQIs i.e.
subjective (WQIsub) and objective (WQIobj). On the other hand only three water quality
parameters (turbidity, DO and EC/TDS) were used to calculate a third index named as
minimal water quality index (WQImin). All three WQIs were calculated to evaluate the effect of pollution loads being disposed into Suquía river (Argentina) from Cordóba city and nearby sites. The water quality of the river was facing deterioration due to anthropogenic activities particularly in the segment downstream of sewage disposal points in Cordóba city. In this segment WQIsub and WQIobj showed prominent variations of water quality with respect to time (i.e. from high flow to low flow period). Similar water quality variation was also noted from WQIobj in the river reach just upstream of sewage disposal point. The WQImin incorporated only three water quality parameters that
involved low cost of analysis. The trends of river water quality obtained from WQImin were similar to those obtained from WQIsub and WQIobj. However it was suggested to calculate WQImin in combination of WQIsub to get reliable results. The study also suggested a framework of calculation according to which all three WQIs should be evaluated for at least two years on monthly basis. After two year regular evaluation, the frequency may be changed to the following patterns: 1) calculate WQImin on monthly or
36 weekly basis and 2) calculate WQIsub twice during high and low flow periods,
respectively.
The water quality of the Mackenzie-Great Bear sub-basin was evaluated by Lumb
et al. (2006) using CCME water quality index model. The study presented the application
of the CCME Water Quality Index to monitor the changes in water quality at four
lcataions at Mackenzie River i.e. near Fort Providence, at Strong Point, at Norman Wells
and above Arctic Red River (Tsiighetchic) and one location at Great Bear River (outlet of
Great Bear Lake). All these sites are located in the Mackenzie-Great Bear sub-basin
which is the largest of the six sub-basins within the Mackenzie River basin in Canada.
The WQIs were calculated for three water uses i.e. overall, drinking and aquatic uses.
Based on CCME WQI model, the raw water quality in the basin was ranked as marginal
to fair along the Mackenzie River for all three selected water uses. The water quality
declined downstream to the Mackenzie Delta for the above uses. Major ions and nutrients
CCME WQI values were categorized as excellent to good. Physical water quality variables (turbidity, true color, suspended solids) and total (mostly particulate) trace metals were lowering CCME WQI values and categories.
The concentrations of trace metals higher than the permissible limits were found in the Mackenzie River basin during 1990s. This was mainly caused due to natural sources and to some extent anthropogenic activities in the basin. Another reason might be the presence of several mines in the surrounding area during that period generating runoff with high trace metal concentrations. As a management action these mines were
37 abandoned due to their polluting nature. The WQIs were calculated for different water
uses by incorporating both site specific guidelines and Canadian Water Quality
Guidelines (CWQGs). The results of WQIs showed similar conclusive results with the
two sets of guidelines and hence no preference was given to one on the other. It was concluded that this study could be helpful for water users (public), water suppliers
(municipal authorities), planners, policy makers, and scientists engaged in research on environment. The CCME WQIs are useful in illustrating “What is changing in the environment?” while they are unlikely to single-handedly answer the other associated in- depth questions like “Why is it happening?” or “What does it mean?” without additional scientific, traditional and local knowledge.
2.4 WATER QUALITY MODELING
2.4.1 Global perspective
The presently available water quality models are sophisticated computer based softwares that can simulate channel hydrodynamics, advection and dispersion processes and kinetics of pollution. These models have been evolved from basic mass balance concepts developed in the second decade of twentieth century by the Royal Commission on Sewage Disposal. The Commission first gave the concept of fixing sufficient dilution ratio for the discharge of wastewater into freshwater bodies. It was a simple approach applicable to limited conditions and dealt only with mass balance in the receiving water bodies without considering flow dynamics and reaction kinetics. A remarkable advancement in context of water quality modeling was the development of simulation equations for DO and BOD in rivers by Streeter and Phelps in 1925. Many water quality
38 models adopted these equations as basis of their calculation. The use of computer has made it possible to solve complex equations quickly with high precision in the last two decades. Significant developments have been made in the field of water quality modeling resulting in the form a vast range of models e.g. WASP5, QUAL2E, SIMCAT, SOBEK,
TOMCAT, HEC 5Q, QUASAR, AQUSIM, MIKE-11, DESERT, ISIS, etc. (Cox, 2003).
Even et al. (2007) investigated the relationship between combined sewer overflows (CSO) in Paris and variations in the water quality of river Seine in France. The study included the laboratory and on-site analysis of the wastewater generated from urban sources. After the characterization of pollution, PROSE model was used to simulate spatial and temporal transport and fate of effluents in river Seine. These effluents were of two types i.e. those added in the river constantly during the dry whether and the effluents of transient nature from CSO. The model simulations incorporated the pollution loadings from CSO in Paris during 1990s. The results of model showed that 50 km long segment of the river passing through the Paris city was permanently affected in terms of DO due to settlement of heavy organic matter loads at downstream of the sewer outlets. Twenty percent reduction in DO levels was caused by CSO but high phytoplankton growth in the river compensated the oxygenation process of the stream to some extent. The model predicted future DO levels to become higher than 7.3 mg/l, 90 percent of times as compared to that of 4.0 mg/l prevailed during 1990s. The oxygen demand due to both permanent dry weather effluents and CSO will be reduced due to strong phytoplankton activities in the river. The study concluded that the inclusion of CSO of the Paris city in
39 the model formulation can be of great importance to get reliable results on spatial and temporal water quality trends.
A basin wide stochastic model was developed by Prairie and Rajagopalan (2007) to assess the salinity of Upper Colorado River Basin located in the western part of US.
Large amounts of salts are added into the river through natural as well as anthropogenic activities. The model was developed to fulfill the input needs of decision support models that are typically used by the water managers to make strategies and policies for salinity abatement in the basin. The decision support models require basin-wide scenarios showing linkage between salinity and stream flow in statistically consistent form. To address this requirement, a basin-wide stochastic model was developed which generated results on salinity levels consistent with the river flow. The setup combined two nonparametric modeling approaches i.e. space-time desegregation that was applied to stream flow and salt concentration, and regression that simulated the magnitude of salts in the given natural stream flow. These techniques were simpler than the parametric modeling approaches. Due to non parametric nature no assumptions showing functional relationship among the parameters were required in the model development process. On the other hand, traditional models using parametric modeling techniques were data-driven and required certain assumptions dealing with the underlying relationship within parameters. The stochastic model tested various scenarios by considering different aspects of stream flow and statistics of salt concentration. Among these scenarios two were finally selected as representative approaches of the stochastic model. The model
40 was then applied to check the performance of these selected approaches with data on flow and salts collected from four sites along the river.
The MIKE 11 model: This model, developed by the Danish Hydraulics Institute
(DHI) in the early seventies, has been used worldwide since 1979 for predicting in-stream concentrations. The model has been efficiently used for water quality evaluation in the
South Asian Subcontinent where Kazmi and Hansen (1997) have applied it for Yamuna river in India and Kamal et al. (1999) for Buriganga river in Bangladesh. This model has also been applied by various researchers in other continents of the world. For example
Duvail and Hamerlynck (2003) have applied it in Mauritania (Africa), Post et al. (2003) in North Queensland (Austrailia) and Thompson et al. (2004) have used the model in
England (Europe). MIKE 11 is excessively used in England as a tool for water quality management and officially recommended in ‘urban pollution methodology’ to assess water quality of estuaries and rivers in response of pollution (Foundation for Water
Research, 1994).
The model has an integrated modular structure with a variety of add-on modules for different simulating phenomena related to river system. The core of the model is the hydrodynamic (HD) module that solves vertically integrated equations of conservation of continuity and momentum (the ‘Saint Venant´ equations) using an implicit finite difference scheme for the computation of the flow in the rivers. The advection-dispersion
(AD) module describes the basic processes of river water quality in areas influenced by human activities, e.g. oxygen depletion and BOD levels as a result of organic matter
41 loads. Concentrations of DO and BOD were calculated in MIKE 11 by taking into
consideration advection, dispersion and the most important biological, chemical and
physical processes. (DHI, 2008)
Kazmi and Hansen (1997) applied Mike 11 model for water quality modeling of
Yamuna river in India. Existing water quality status of the river was assessed and
simulations were made to predict possible changes in water quality in response of
different strategies for pollution control. The river segment from Mawi to Okhla barrage
in Delhi was selected for model formulation. The simulated results of the model showed
that large growth of phytoplankton caused 80 to 135 percent saturation of dissolved oxygen in the river reach from Mawi to Wazirabad downstream of which DO was considerably decreased. There was immense deterioration in water quality near Delhi, seizing the photosynthesis process in the stream. High BOD levels were also observed due to deposition of organic pollution at the river bed. Different scenarios were tested with the calibrated model. These scenarios used the pollution loads resulting from proposed strategies (Yamuna Action Plan) to control waste addition in the river. The water quality was ranked within four classes based on BOD level as: class A < 2 mg/l, class B < 3mg/l, class C < 4 mg/l and class D < 6 mg/l. The simulated results of the scenarios showed that BOD in the river reach upstream of Delhi was improved from class
B to A after pollution control measures. Considerable improvement in BOD was simulated at Delhi but it remained below class D. Then the model was run incorporating some management strategies for DO under Yamuna Action Plan. The results indicated improvement in BOD up to class D (water quality fit for aquatic life). Finally the scenario
42 assuming addition of 40 m3/s flow in the river also simulated improved BOD level in
Delhi segment up to class D.
A water quality study was conducted by Nhan (2005) in Nhieu Loc Thi Nghe (NL
TN) basin. This basin is the centre of commerce and culture of Ho Chi Minh (HCM) city located in the south of Vietnam. It covers about 33 km2 with more than one million people. Along with rapid urbanization in HCM city in general and NL TN particularly, surface water quality in NL TN canal system was increasingly polluted by mainly domestic wastewater which had impacts on public health and city aesthetics. Therefore, an assessment of water quality in NL TN canal system to help in finding solutions was more important. In this study, MIKE 11 was used as a tool to analyze parameters of water quality as BOD and DO. The canal network included main canal, NL TN canal (9.4 km of length) and two branches i.e. Bong canal (2.1 km) and Van Thanh canal (1.8 km). The calibration of models was carried out with observed data in April 2003. The results suggested that the BOD level in NL TN canal was very high. The BOD was around 100 mg/l, as four times as limit set of Vietnamese Surface Water Quality Standard for
Category B (VSWQS-B; type of water not used for domestic use). While DO was mostly depleted, especially at upstream canal when low tides occurred. At downstream canal, polluted water was diluted by water from Sai Gon River with 25 mg/l of BOD and 4 mg/l of DO, so here the water quality was better than that at upstream canal. The study also predicted that a wastewater collection system and pumping station being constructed at left canal bank near NL TN canal mouth, expected to complete in 2007, will improve water quality in the basin by collecting wastewater to convert to Sai Gon River. The
43 polluted water in NL TN system and its impact on Sai Gon River were simulated in 2007
before and after operation of pumping station. In 2007, before operation of pumping
station, wastewater was discharged to NL TN canal system; the BOD values were not
much higher than those in 2003. After having pumping station, the interaction of BOD and DO between NL TN canal system and Sai Gon River were also simulated. The discharge of pumps to Sai Gon River was 2.58 m3/s. With this discharge, the BOD level was below the limit set of VSWQS-B.
MIKE 11 model was applied to assess the water quality status of Buriganga river in Bangladesh by Kamal et al. (1999). The river traverses through the capital city of
Bangladesh. Dhaka is one of the most populated cities in the world but there are very few treatment plants for the wastewater generated in the city. As a result increasing amount of industrial wastewater and domestic sewage is discharged into the river without prior
treatment. The study included some cardinal parameters for water quality assessment out
of which DO was modeled using MIKE 11 model. The water samples were collected
during 1994-95 and subsequently analyzed for selected water quality parameters in
laboratory and on site as well. The model was then formulated using the data collected
during low flow months of the study period. Simulated results of the model showed very
low DO level in the river. Through the series of alternative scenario simulations with the
calibrated water quality model, it was found that discontinuation of any of the major
point sources of pollutants might not be adequate to improve minimum DO level in river
Buriganga. Also, discontinuation of these wastewater sources was hardly possible
physically. It was also noted that establishment of treatment plants with 60 percent BOD
44 removal efficiency at major point sources of pollutant might not be adequate to raise DO
level. On the other hand, a dramatic improvement of the minimum DO level in the river
was observed if all the major pollutant sources were treated for biodegradable material
and disposed at a location further downstream of its existing point of entry into the river.
According to Radwan and El-Sadek (2005) people have become increasingly
concerned with water resources problems and studied them much more intensively in the
recent years. A mathematical model can be considered as a major tool for the efficient
management of receiving waters. With such a model, a river or watercourse can be
simulated to analyze its observed state. In this context a study aimed to provide an insight
into the effect of control structures on the outcomes of hydrodynamic (HD) and water
quality (WQ) modules of MIKE 11 modeling system was conducted. The study was
performed for the Molenbeek brook, which is one of the main tributaries of the River
Dender basin in Belgium. The river has been regulated by hydraulic structures to protect
several areas from flooding. The model of their regulation and their effects on the river
water quality were presented in this paper. The model results were studied for DO, NH4-
N, NO3-N and BOD concentrations at different locations. The first location was the most
upstream point in the river, the second was at the measurements station (upstream all the
hydraulic structures), the third and fourth were in between the structures, and the last one
was the most downstream point. It was seen from these results that, the water quality
condition along the river was much better for the upstream section where no hydraulic
structures existed. Starting at the location of the first control structure, the water quality
condition was deteriorated (lower DO, higher BOD and higher ammonium
45 concentrations). This can be explained as the structures reduced the flow velocity and
consequently increased the water depth directly upstream. The combined effect of these
was a reduction in the amount of re-aeration (by an increase in depth and a reduction in water surface gradient). Moreover, as the current regulation of the control structures was uniform in time, even in dry conditions, the discharge downstream of these structures was very small. This very low discharge combined with continuous release of the untreated domestic wastes creates critical quality conditions at these locations.
This study also included study of different scenarios for the regulations. By changing the current regulation of the control structures during dry summer periods
(when the flood regulation is not relevant), the water quality was improved significantly in terms of higher DO, lower BOD and Ammonia (NH4-N) concentrations. This showed
the need for an integrated management and modeling of a hydrographic catchment,
considering both hydrodynamic (flooding) and water quality aspects.
Khan and Vongvisessomjai (2002) applied MIKE-11 HD and AD modules for the
simulation of salinity intrusion in the southwest region of Bangladesh. The study area
was divided into two zones (the inland and coastal zones) of different characteristics. In the first (inland) zone, irrigation was applied from shallow tube wells and surface water.
The surface water quality was not saline during the dry season in this zone. On the other hand, surface water was saline in the second (coastal) zone. The salinity level of the rivers in coastal region depended on many factors e.g. dispersion process responsible for the transport of salts, level of salinity and discharge at the river entrance facing Bay of
46 Bengal and storage volume and distribution of freshwater discharge into the river at upstream. The salinity intrusion from the sea started from November and extreme conditions prevailed in late March and early April. In this study, the tidal analysis and quantification of salinity intrusion was performed during different months. The flow routes of main rivers and fresh water were identified. Salinity profiles along these routes were simulated during different months. The level of salinity and extent of salinity intrusion in the river system was compared before commissioning of Farakka barrage and after Ganges Water Treaty. The model simulations under these conditions showed that the water became unsuitable for irrigation in the extended river stretch of 20 km in upstream direction. The results of model assuming maximum freshwater availability and no structural measure showed a small reduction in salinity level.
In a research study on Wimmera river in Australia, three different techniques
(field analysis, laboratory analysis and numerical modeling) were combined by Western
(1994) to quantify water salinity and stratification density of the river. Flow of the river remained highly variable during different seasons. The salinity of river was very high because of two factors: 1) the runoff in upper river catchment brought large amounts of salts in the river and 2) the salts were also added in upper and lower river reaches through ground water inflows. A number of deep and long pools were formed in the river during extremely dry periods. MIKE 11 model was applied to simulate flow and salinity in the river over a 200 km long river segment. First the network of channels was defined during model formulation. Then the cross-sections and other related input data was specified.
The model was calibrated and subsequently validated. The application of model helped in
47 understanding morphology of the river channel. Simulations were made for river flow and salinity under different conditions. The results showed that the model simulated the flow regimes and salinity of the river to adequately good extent.
2.4.2 Local Perspective
Jamil and Latif (2006) conducted a study to model the behavior of contaminant transport in river Indus from downstream of Ghazi to the confluence of Indus river with
Kabul river (about 40 km stretch). For this purpose QUAL 2K model developed by
United States Environment Protection Agency (USEPA) was used. The data for the study was acquired from Water and Power Development Authority (WAPDA). For simulation purposes three scenarios were developed to study the worse conditions of water quality in river Indus during the flow period. Scenario-1 was run with the present conditions of wastewater quantity and quality. In the scenario 3, flow was the same but the wastewater qualities were projected up to year 2063 with the same water quality in year 2002. In case of scenario-3, it was assumed that the wastewater quality will be further deteriorated to the extent that the BOD value will be in the range of 80 mg/l in the project life i.e. year
2063. The river flow and the wastewater quantities ware kept same in case of scenario-2.
The model results showed that the DO and BOD will remain within the permissible limits
(i.e. DO > 4mg.l and BOD < 8mg/l) in case of all three scenarios.
Similarly, Behzad (2002) applied QUAL 2E model to river Ravi. The model was formulated using existing conditions of river discharge (12.74 m3/sec) and a wastewater load of 22 m3/sec. The results of the model showed that the river was highly polluted
48 with maximum BOD of 19.21 mg/l and minimum DO of 1.13 mg/l at the reaches near the
sewage disposal points. Some scenarios were also tested in the model with different
combinations of flow and sewage loads. The results showed that when flow from Ravi
Syphon will be 12.74 m3/sec then a flow of 26 to 92 m3/sec will be required to get a
minimum DO level of 2 to 4 mg/l, respectively, throughout the river reach. When the
flow from Ravi Syphon will be 12.74 m3/sec with the predicted sewage loads of 35 m3/sec after 15 years i.e. in 2017, a flow of 57 m3/sec from Marala Ravi (MR) link canal
would be required to meet a minimum DO standard of 2 mg/l through the river reach. It was recommended that under present operating conditions, the dilution of wastewater through MR Link canal appears available solution to solve the environmental problem of river Ravi. A minimum DO of 4 mg/l can be achieved by dilution from M.R. Link canal presently and in future with the conditions that the concerned department must have an operational plan for the releases from M.R. Link ready to use keeping in view the river discharge and wastewater discharges being pumped into river Ravi.
2.5 DERIVED CONCLUSION
The literature review on low flow analysis provides an overview of different measures, indices and frequency distribution functions. The present study emphasized water and wastewater sampling only during the low flow season. Therefore, low flow analysis was the first step to proceed for detailed analysis. In Pakistan, more attention has been given to flood analysis of the rivers as compared to low flow analysis. Low flow analysis is usually included in the feasibility reports of hydraulic and hydropower
49 projects. Such analysis is not sufficient for the studies focused on river water quality.
Unfortunately the national literature lacks such specific studies.
The next section of the chapter provides literature on water quality monitoring and assessment. In global perspective, many studies conducted in different parts of the world have been included. The literature shows that these monitoring programs were designed with different sampling frequencies and included varying set of water quality parameters. In local perspective, few indigenous studies are included. It is evident from review of literature in this section that the monitoring of different rivers of Pakistan was spatially fragmented and temporally sparse. Very little work has been done to monitor river water quality on systematic and regular basis. In fact the water quality monitoring program on national scale include all the rivers and lakes of Pakistan and only a few stations can be selected on each river due to constraints of resources and time. Moreover the sampling is usually made on season basis (once in six months). The data from these monitoring programs remain insufficient for detailed analysis. A few projects on water quality of river Ravi have been completed by some research institutions to overcome the limitations of national monitoring programs. But still enormous efforts are needed to fill the knowledge gap on the issue of water quality monitoring.
The evolution of water quality indices, particularly the Canadian Water Quality
Index (CWQI), is discussed in third section of this chapter. It also includes some water quality studies in which CWQI 1.0 model has been applied to calculate WQIs.
50 Unfortunately water quality indices have not been calculated for any river in Pakistan except this study.
Finally in the last section (water quality modeling) literature on MIKE 11 model is presented. This model has been extensively used in all continents of the world for water quality analysis of rivers. MIKE 11 has also been applied successfully in south
Asian subcontinent (India and Bangladesh). MIKE 11 has not extensively used in
Pakistan for water quality studies of the rivers. Few local studies have been included where other than MIKE 11 water quality models (qual2K and qual2E) have been applied.
51 CHAPTER III
MONITORING AND ASSESSMENT OF RIVER WATER QUALITY
3.1 THE RIVER CHENAB
The present study was based on river Chenab, one of the largest rivers of the
Indus basin. Some important characteristics of this river are illustrated as follows:
3.1.1 Historical Background
The river Chenab has a very old history. In the ancient Vedic period this river was
named as ‘Ashkini’ or ‘Iskmati’ in India and to the ancient Greeks it was known as
‘Acesines’. The historical town of “Uch Sharif’ or ‘Mithankot’ (where five rivers of
Punjab join the Indus river) was allegedly founded by Alexander the Great in 325 BC.
The people of Punjab give great importance to river Chenab due to its beneficial uses and
legendary place in the history of this region. The river basin is known as the land of
lovers and it holds an iconic place in the tales and epics of Punjab like ‘Sohni Mahiwal’
and ‘Heer Ranjha’ (en.wikipedia.org).
3.1.2 Location
The river Chenab has its origin in Lahul and Spite districts in Himachal Pradesh
province of India where it is formed by the confluence of two streams i.e. Bhaga and
Chandra at Tandi in the state of Jammu and Kashmir (Indian occupied). The location of
the river source is 77°- 30’ E and 32°-50’ N. The river is also known as ‘Chandrabhaga’
in its upper reaches. Besides Bhaga and Chandra, Bhut and Maru nullahs (small natural
streams) also join the river from right side above Salal dam. Near the line of control, river
52 Chenab is joined by two major tributaries namely Munawar Tawi and Jammu Tawi.
These two tributaries drain an area of about 2,800 km2 on both sides of the river. The
Chenab river enters in Pakistan just upstream of Marala rim station (74°-29’ E and 32°-
40’ N) and travels a distance of approximately 598 km up to its confluence point with river Indus at Mithankot. It traverses in the mountainous slopes of Jammu and Kashmir and then flows in the plains of Punjab after entering in Pakistan. The interfluve between
Jehlum and Chenab rivers is named as ‘Chej Doab’ and that between Chenab and Ravi rivers is called ‘Rechna Doab’. The river Chenab is joined by Jehlum river just upstream of Trimmu barrage followed by river Ravi near the town of Ahmadpur Siyal. This combined stream of three rivers finally meet with the remaining two rivers of Punjab i.e.
Sutluj and Bias to form ‘Punjnad’ (combined stream of five rivers) that finally merges into river Indus (en.wikipedia.org). Pakistan has the exclusive rights of the waters of river
Chenab according to Indus Waters Treaty (IWT) between Pakistan and India.
3.1.3 Chenab River Basin
The Chenab river basin drains an area of about 67500 km2 before its confluence
with river Indus. From its source to its confluence with river Indus at Sarki village, some
35 km upstream of Mithankot, it traverses a length of about 1240 km. In the upper reach,
the river first flows in the north-westerly direction till Benswar, where it is joined by
Maru nullah. Downstream of Benswar, the river takes a sharp turn to the south while
traversing the Pir Punjal range of mountains and continues its way in the westerly
direction till Salal dam site, thereby, draining the southern slopes of Pir Punjal range.
Here it again turns to the south till it turns out into the plains at Akhnoor, 34 km from the
53 line of control between India and Pakistan, to change its course into south westerly
direction. The total catchment area upstream of Marala (the point of entry in Pakistan) is
about 28000 km2. From Marala barrage to the confluence with river Indus, the river
Chenab traverses through Punjab province for another 576 km. A number of smaller tributaries i.e. Halsi, Bhimber, Palkhu and Aik join the river between Marala and Khanki barrages, draining a total area of 3437 km2. A layout of the river Chenab basin is presented in Figure 3.1.
3.1.4 River Topology
The river slope from its source to the mouth varies greatly, with the steepest part, about 25 m/km, upstream of Tandi. From Tandi to Akhnoor the slope is about 5 m/km which further reduces when the river flows into the plains. It is about 0.40 m/km south of
Akhnoor and to 0.20 m/km near the river mouth. The river channel below Akhnoor is wide, and the flood plain is enormous. Downstream of Marala the river width varies from
700 to 1,400 m, whereas the flood plain is about ten times as wide as the river. The topographical features of rather flat slope and very wide over bank lead to low flood wave celerity, in the lower reach of Chenab river. Measured along the flood plain from
Marala barrage to Panjnad the Chenab river traverses a distance of 534 km. The average slope ranges from 0.33 m/km in the Marala-Khanki reach and 0.17 m/km in the Trimmu-
Panjnad reach. The sinuosity (ratio of length of river meandering to the straight length) of river Chenab in the reaches from Marala barrage to its confluence with river Indus is less than 1.1. Therefore, no corrections for low flows are required.
54 55
Figure 3.1: The Chenab river basin. 3.1.5 Hydraulic Infrastructures
Following is a brief description of different hydraulic infrastructures on river
Chenab.
A. Dams
Two dams have been constructed on river Chenab in the disputed state of Jammu and
Kashmir discussed as follows:
i) Salal Dam: Salal hydropower dam has been constructed at Dhyangarh located at
about 73 km upstream of Marala. It is a concrete dam with a maximum height of 113 m.
The construction work commenced in 1970. In the first stage of the project, a power station of 345 MW capacity was commissioned in 1987. In the second stage, the total capacity of this station was doubled to 690 MW by 1995. Its live storage is small, giving the dam little potential to affect the river regime.
ii) Baglihar Dam: The disputed Baglihar hydropower dam is located in the southern
Doda district in the state of Jammu and Kashmir. The project was conceived during 1992 and the Indian government gave its approval in 1996. The construction work started in
1999 and by the year 2004 its first phase was completed. The dam was commissioned on
10th October 2008.
The design of Baglihar dam voilated the Indus Waters Treaty (IWT) between
India and Pakistan signed in 1960. This dispute could not be resolved after many rounds of talks between the two countries during 1999-2004. Pakistan raised the issue to the
56 World Bank (a guarantor of IWT) in January 2005 who appointd a Swiss technical expert
to judge the issue. The final verdict of the neutral expert was released in February 2007.
According to the decision some of the objections raised by Pakistan were partially
upheld. The decision recommended a reduction in pondage capacity by 13.5 percent and
in dam height by 1.5 meters. It was also recommended to raise the power intake tunnels
by three meters. From these changes in the previous design, its capabilities to control
river flow were reduced. Some objections of Pakistan regarding the height and gates of
spillway were rejected in this verdict (en.wikipedia.org).
Some salient features of the Baglihar dam as presented during the site visit by the
Pakistani engineers in October 2005 are given as follows (GoP, 2005):
Type of dam Concrete Gravity Dam
Height above deepest foundation 144.50 m
Catchment area 17325 km2
Mean annual inflow 25000 million m3
Length of dam crest 317 m
Crest elevation 84450 m asl
Maximum discharge capacity 16500 m3/s
Installed capacity of power house 450 MW
57 B. Barrages
Four barrages have been constructed in Pakistan i.e. at Marala, Khanki, Qadirabad
and Trimmu. Some salient features of these barrages are presented as follows:
i) Marala Barrage: This barrage is located at 23 km towards the northeast of Sialkot
city. It is 16 km from the foothills of Pir Punjal range, where Chenab river enters the
plains near Akhnoor. The capacity of the weir is 19800 m3/s (700000 cfs) with an
upstream pond level of 246.28 m (808 feet). The weir experienced its highest flood on
record on July 26, 1957 with value of 31150 m3/s (1100000 cfs). The left and right bunds
were allowed to breach resulting in a catastrophic inundation of lands and ‘abadies’
(dwellings).
Marala weir was abandoned in 1968, when it was replaced by a new barrage
under the Indus Basin Replacement Works. The lay out of the new barrage is shown in
Fig. 3.2. The new barrage has been constructed at 344 m (1130 feet) downstream of the
old weir. The discharge capacity of the left pocket was enhanced to accommodate a huge
canal diversion of 1100 m3/s. The new barrage experienced an exceptionally high flood of 23930 m3/s (845090 cfs) on September 10, 1992, when a flood level of 248 m (814
feet) at weir crest was attained.
The pond level of the new barrage was raised by four feet in order to improve the
performance of the off-taking channels. The pond level was thus fixed at 247.5 m (812
feet). This reduced the approach velocity and considerably reduced the erosion of "belas"
58 (sand bars). The old weir crest in the normal bays was at 244 m (800 feet). The same
crest level was kept for the new barrage. The barrage has a left marginal bund, a right
closure bund, guide banks and numerous river training works. The weir comprises three
sections: a left section with 13 bays, a right section with 7 bays and central section with
46 bays, each having width 18.3m (60 feet).
Ga tes 3.96 Х 18.29 14.02 Х 18.29 2.13 Х 18.29
Left Midd le Right Gate Axis
264.26 m 940.69 m 141.27 m
Left and Right Across the Barrage 242.31 m
239.88 m 239.12 m Flow 1:10 1: 3 Flow
243.84 m Middle Parabolic Curves
242.31 m Flow 240.64 m Flow 1:16 1: 3
Figure 3.2: Layout of Marala barrage.
Salient Features of Marala Barrage
• Year of commission 1968
• Maximum designed capacity 31149 m3/s (1100000 cfs)
• Highest flood passing (during 1992) 23930 m3/s (845090 cfs)
• Normal pond level 247.5 m (812 ft)
• Crest level - standard bays 244 m (800 ft)
59 • Crest level - under-sluice bays 242.3m (795 ft)
• No. of left under-sluice bays 13
• No. of right under-sluice bays 7
• No. of standard bays 46
• Width of each bay 18.3 m (60 ft)
• Width of waterway of the left under-sluice 238 m (780ft)
• Width of waterway of the main weir 841 m (2,760 ft)
• Total width of waterway 1028 m (3373 ft)
• Width between abutments (including piers) 1363 m (4472 ft)
• No. of divide walls 2
• Top level of each gate 248 m (814 ft)
• Upstream floor level 242 m (795 ft)
• Downstream floor level 240.6 m (789.5 ft)
Off-taking Canals a) Upper Chenab canal (UCC)
• Design discharge 477 m3/s (16,850 cfs)
• Crest level 244.4 m (802 ft)
• No. of bays 6
• Width of each bay 12.2 m (40 ft)
• Total width of water way 73.2 m (240 ft)
• Width between abutments 80.8 m (265 ft)
60 b) Marala Ravi Link Canal
• Design discharge 622 m3/s (22,000 cfs)
• Crest level 244.9 m (803.5 ft)
• No. of bays 8
• Width of each bay 12.2 m (40 ft)
• Total width of water way 97.6 m (320 ft)
• Width between abutments 108.2 m (355 ft)
ii) Khanki Headworks: This headworks is located on river Chenab at 16 km (10 miles) downstream of Alexandra bridge. It is the oldest (constructed in 1890-92) major irrigation diversion structure in use in Pakistan. The headwork consists of three under-
sluices and six bays controlled by shutters. It has a design discharge capacity of 22654
m3/s (800000 cfs). Commissioning of Qadirabad Barrage, at about 30 km (19 miles)
downstream of the headworks has accreted the river bed between Khanki and Qadirabad
by 0.6 to 1.2m (2 to 4 ft) rendering the former non-modular during high floods. This
raises the stage through the weir resulting in frequent breaching of the marginal bunds.
Salient Features of Khanki Headworks
• Year of commission 1890-92
• Maximum designed capacity 22654 m3/s (800000 cfs)
• Highest flood passing (during 1992) 25768 m3/s (910000 cfs)
• Crest level - standard bays 221.4 & 221.6 m (726.5 & 727 ft)
• Crest level - under-sluice bays 217.9 m (715 ft)
• No. of under-sluice bays 3
61 • No. of standard bays 6
• Total width of waterway 1197.6 m (3,929 ft)
• Width between abutments 1337.2 m (4387 ft)
(including piers)
Off-taking canals a) Lower Chenab Canal
• Design discharge 477 m3/s (16,850 cfs)
• Crest level 221.3 & 219.8 m (726 & 721 ft)
• No. of bays 18
• Width of each bay 7.5 m (24.5 ft)
• Total width of water way 134.4 m (441 ft) iii) Qadirabad Barrage: It is located in Gujranwala district, 16 km west of Alipur- chatha town and 30 km downstream of old Khanki weir. The barrage was commissioned in 1969. It has a design discharge capacity of 25485 m3/s (900000 cfs).
The river Chenab is bounded between marginal bunds between Khanki headworks and Qadirabad barrage. The river while flowing between its alluvial bed meanders widely between the marginal bunds. To cop with the situation, river training works have been provide. Qadirabad-Balloki (QB) link canal off-takes from the left bank and is the central part of the Rasul-Balloki link canal system, which carries Mangla releases to Ravi river for its utilization in the areas which were previously (before IWT in 1960) irrigated by the Ravi and Sutluj rivers. The QB link canal joins river Ravi at about 16 km upstream
62 of Balloki headworks. It supplies irrigation water to about 2.43 million hectares of cultivable land. The layout of Qadirabad barrage is shown in Figure 3.3.
Gates 1.52 m Х 18.29 13.72 m Х 18.29
Left Main
99.97 m 916.84 m
1028.09 m
Left Section 209.25 m 207.26 m 205.44 m 203.0 m 5 1 1 : : 3
Across Qadirabad Barrage
208.64 m Main Section
205.44 m 204.52 m : 9 1 1 : 3
Figure 3.3: Layout of Qadirabad barrage.
Salient Features of Qadirabad barrage
Year of commission 1968
Maximum designed capacity 25485 m3/s (900000 cfs)
Highest flood passing (1992) 26844 m3/s (948000 cfs)
Normal pond level (summer) 212.1 m (696 ft)
Normal pond level (winter) 213.7 m (701 ft)
Crest level - standard bays 208.5 m (684.5 ft)
Crest level - under-sluice bays 207.3 m (680 ft)
No. of under-sluice bays 5
63 No. of standard bays 45
Width of each bay 18.3 m (60 ft)
Width of waterway of the left under-sluice 823.0 m (2,700 ft)
Width of waterway of the main weir 91.4 m (300 ft)
Total width of waterway 914.4 m (3000 ft)
Width between abutments 1028.1 m (3373 ft)
No. of divide wall 1
Thickness of divide wall upstream floor level 2.9 m (9.6 ft)
Downstream floor level (weir portion) 204.5 (671 ft)
Downstream floor level (under-sluices portion) 202.0 m (666 ft)
Off-taking Canals a) Qadirabad-Balloki (QB) Link Canal
Design discharge 623 m3/s (22,000 cfs)
Crest level 210.3 m (690.5 ft)
No. of bays 6
Width of each bay 12.2 m (40 ft)
Total width of water way 73.2 m (240 ft)
iv) Trimmu barrage: This barrage is located about 3 km downstream of the confluence of Jhelum and Chenab rivers. The barrage is located at about 292 km downstream of Marala barrage on Chenab river. It falls in the semi-arid zone with desert on its right and Rechna Doab on the left.
64 Due to synchronization of floods of Jhelum and Chenab rivers, during the year
1929, the barrage was designed for 18265 m3/s (645000 cfs) discharge. Flows approach the barrage at an angle of 55 degrees from the normal and the safe discharge capacity is estimated to be 16, 800 m3/s (592000 cfs). In 1959 an exceptionally high flood peak of
26700 m3/s (943000 cfs) was recorded which was more than its capacity. Three canals named Haveli main line, Trimmu Sidhnai (TS) link canal and Rangpur canal are off- taking from the barrage. Haveli Main canal and TS Link canal off-take from the left flank, whereas the Rangpur canal off-takes from right flank of the barrage.
Salient Features of Trimmu Barrage
Trimmu barrage was constructed in 1938-39 as a part of the Haveli project. The main features of the barrage are given as under and shown in Fig. 3.4.
Gates 2.44 m Х 9.14 m 11.28 m Х 18.29 m 1.83 m Х 9.14 m
Left Main Right
88.09 m 753.47 m 65.53 m
922.02 m
Main Section Across Trimmu Barrage
145.54 m
1 143.26 m : 4 : 1 4 142.65 m
9.14 m 1.83 m 9.14 m
Figure 3.4: Layout of Trimmu barrage.
65 Salient features of Trimmu barrage
• Year of commission 1939
• Maximum design capacity 18265 m3/s (645000 cfs)
• Normal pond level 150.0 m (492 ft)
• Crest level of the main weir 145.5 m (477.5 ft)
• Crest level of left and right under-sluices 143.9 m (472 ft)
• No. of Bays in the main weir 37
• No. of bays in the left under-sluice 8
• No. of bays in the right under-sluice 6
• Width of each under sluice bay 9.1 m (30 ft)
• Width of each standard bay 18.3 m (60 ft)
• Width of waterway of left under-sluice 73.15 m (240 ft)
• Width of waterway of right under-sluice 54.9 m (180 ft)
• Width of waterway of the main weir 676.7 m (2,220 ft)
• Total width of waterway 804.7 m (2,640 ft)
• Length between abutments 922.0 m (3,025 ft)
Off-taking Canals a) Haveli Main Line Canal
• Design discharge 146.4 m3/s (5,170 cfs)
• Crest level 2146.6 m (481 ft)
• No. of bays 5
66 • Width of each bay 7.32 m (24 ft)
• Total width of water way 36.6 m (120 ft) b) Trimmu-Sidhnai Link Canal
• Design discharge 354.0 m3/s (12500 cfs)
• Crest level 146.6 m (481 ft)
• No. of bays 10
• Width of each bay 7.32 m (24 ft)
• Total width of water way 73.15 m (240 ft) c) Rangpur Canal
• Design discharge 76.7 m3/s (2,710 cfs) • Crest level 147.4 m (483.50 ft) • No. of bays 3 • Width of each bay 7.32 m (24 ft) • Total width of water way 22.0 m (72 ft)
C. Bridges
The following bridge crossings exist over Chenab river in Pakistan:
• At Wazirabad, 40 km downstream of Marala, four bridge crossings exist over
Chenab, viz.: Alexandra railway bridge with a total span of 683 m and an old road
bridge about 120 m downstream of railway bridge with 724 m span. Two road
bridge crossings at Grand Trunk road exist about 900 m downstream of the old
road bridge. The distance between these two bridges is 150 m while the bridges
span about 750 m. Alexandra railway bridge was constructed in 1870-76 with 64
spans of 40 m (132 feet) each. Of these, 36 spans were closed in the year 1890.
67 Later on 11 spans were closed during 1918-19, leaving 17 spans. The present
design discharge capacity is 31152 m3/s (1100000 cfs). The approach of the river
to the bridge is oblique. This has reduced the capacity of the bridge to 20135 m3/s
(711000 cfs) which corresponds to a return period of 16 years.
• The new road bridge at Talibwala, located at 149 km downstream of Marala is
part of the Lahore-Islamabad motorway (M2).
• Chiniot bridge is constructed at 182 km downstream of Marala. Near Chiniot the
river traverses a rocky ridge. The river water is conveyed through two channels;
the east and west channels separated by a hill. Two rail-cum-road bridges span
over these channels. The total bridge span at Chiniot is 407 m. Its design capacity
is 22656 m3/s (800000 cfs), which corresponds to return period of 16 years.
• Rivaz bridge is located at 261 km downstream of Marala near Jhang. The width of
the waterway at Rivaz bridge is 671 m. Rivaz bridge consists of 11 spans of 61 m
(200 feet) each. Its design capacity is 20,671 m3/s (730,000 cfs).
• Between the confluences of Ravi and Sutlej rivers with Chenab river and 444 km
downstream of Marala is Shershah rail-cum-road bridge with a waterway width of
1035 m. A new bridge is under construction just downstream of the existing
Shershah bridge.
3.1.6 Travel and Lead Times
The lead time between rainfall in the upper catchment (i.e. the catchment between
Benswar and Marala) and the occurrence of the flood at Marala ranges from 14 to 20 hours, where the shorter period refers to the higher floods. These lead times can be
68 achieved provided that proper rainfall estimates can be obtained through weather radar.
The rainfall is then transformed into runoff at Marala using the rainfall-runoff models developed for the upper reaches. The quality of the forecasts can be improved by making use of data from Salal dam; the travel time from Salal dam to Marala ranges from 9 hours for extreme floods to 17 hours for medium floods. The above lead times can further be increased by making use of rainfall forecasts. These forecasts when put into the rainfall- runoff models will allow lead time for Marala to increase to 38 hours. The lead time for the extremely flashy right bank tributaries entering the Chenab between Marala and
Khanki are very short. Travel times and potential lead times along Chenab river in
Pakistan are indicated in Table 3.1. It is recommended to take steps to receive regular information on flows at the Salal and Baglihar dams, at least four times per day during severe floods (NESPAK, 2007).
Table 3.1: Observed wave travel time and potential lead time in river Chenab. Chainage Travel time from Potential Location Marala barrage lead time (km) (hrs) (days) (days) Marala barrage 0 0 0 1.5
Alexandra bridge 40 8-9 0.35 1.9
Khanki barrage 56 10-13 0.5 2
Qadirabad barrage 85 18-20 0.8 2.5
Talibwala bridge 149 45-48 2 3.5
Chiniot bridge 182 58-62 2.5 4
Rivaz bridge 261 82-84 3.5 5
Jhelum confluence/Trimmu 292 91-96 4 5.5
Ravi confluence 362 119-125 5 6.5
Shershah bridge 443 148-156 6 7.5
Sutlej confluence/Panjnad 533 186-196 8 9.5
Indus confluence 576 202-212 8.5 10
Source: (NSPAK, 2007).
69 3.1.7 Historical Discharges
Table 1 presents mean monthly discharge of river Chenab at Marala for 60 year
period. The table clearly designates that the low flow months are October to March when
the mean monthly flow is only about 9 to 22 percent of the maximum flow during the
months of August. It is also interesting to note that flow during the lean flow period
(October to March) is only 18.4 percent of the total river flow.
Table 3.2: Mean monthly discharge of river Chenab at Marala for 60 year period (1946-47 to 2006-07).
High Flow Period Lean Flow Period Discharge Percent of Discharge Percent of 3 3 Maximum Months (m /sec) Maximum Months (m /sec)
April 717 27.3 October 492 18.7 May 1095 41.7 November 267 10.2 June 1726 65.7 December 235 9 July 2619 99.7 January 346 13.2 August 2628 100 February 378 14.4 September 1435 54.6 March 586 22.3 Total 10220 2304
3.2 SELECTION OF RIVER REACH
The present study was restricted to a 292 km length of river Chenab, starting from
its entrance in Pakistan at Marala (Sialkot) to Trimmu (Jhang), due to limitation of time
and resources. Another reason of this selection was that the maximum intervention of
drains is in the river segment upstream of Trimmu before confluence of river Jehlum.
70 After its entrance in Pakistan, the river Chenab traverses through a number of densely populated and industrial cities (e.g. Sialkot, Gujranwala, Gujrat, Hafizabad,
Faisalabad, Sargodha and Jhang) in the Punjab province of Pakistan (Fig. 3.5). According to a recent estimate, population of these cities is about 23.5 millions (GoP, 2008).
Figure 3.5: Map of selected river reach of river Chenab.
A schematic diagram of the selected river reach is presented in Figure 3.6 for illustration. The figure shows the canals, drains and hydraulic structures at the river that were considered during the monitoring and modeling parts of this study.
71
Figure 3.6: Schematic diagram of selected reach of river Chenab.
3.3 DESIGN METHODOLOGY
To accomplish the design and establishment of the water quality monitoring
network, the concept of monitoring cycle developed by United Nations Economic
Commission for Europe (UN/ECE, 2000) was taken into account. According to this concept the process of monitoring and assessment is a sequence of related activities that
starts with the definition of the information needs, and ends with the use of the
information product (Figure 3.7). These successive activities in the monitoring cycle
should be specified and designed in light with the required information product as well as
the preceding part of the chain. The ultimate goal of a monitoring program is to provide
the information needed to answer specific questions during decision making process, thus
it is important to clearly define and specify the requirements in terms of information.
After the specification of the information needs, assessment strategies are followed by the
72 design and operated in such a way that the required information is obtained according to
UN/ECE monitoring cycle (Ward et al., 2004).
Water Management
(9) Information Utilization (1) Information Need
(2) Monitoring Strategy (8) Reporting
(3) Network Design (7) Data Analysis
(6) Data Handling (4) Sample Collection
(5) Laboratory Analysis
Figure 3.7: Monitoring cycle developed by UN/ECE.
The design and operation of monitoring programs includes many aspects, such as field measurements, sampling (collection, pre-treatment, storage methods and transport), chemical analysis and data compilation and finally reporting the information. Four stages
(from 2 to 5) of the monitoring cycle (Figure 3.7) are explained in context of the present study as follows:
73 3.3.1 Monitoring Strategy
Before the initiation of a monitoring program, the government agencies involved in monitoring of the surface waters of Pakistan were contacted. Their experience was shared by arranging frequent meetings, training sessions and joint field tours. Data requirements of the study were then identified. For secondary data, the concerned departments, agencies and NGOs were contacted. The collection methods and techniques were complied for primariy data collection e.g. water and wastewater sampling, discharge estimation of drains, etc. Constraints of time and resources were also considered during formulation of the monitoring strategy.
Pre-sampling visits were also made to:
• Identify the location of various surface drains (point source pollution).
• Select appropriate sampling stations along the river.
• Take geological co-ordinates of the selected sampling stations.
• Estimate different equipments and material required during the monitoring
program.
• Plan the frequency of sampling.
• Selection of minimum water quality parameters within the available time and
resources, without jeoperdising the core objectives of the study.
3.3.2 Network Design
In the light of the experience learnt from all above activities, a final monitoring network was designed. Seven sampling sites were selected, as shown in Figure 3.8, along river Chenab for the collection of water samples on monthly basis. The geographic co-
74 ordinates of these sampling stations as noted from hand held Geological Positioning
System (GPS) device and their distances from Marala barrage (set as 0 km) are given in
Table 3.3. Apart from these sampling stations at the Chenab river, the wastewater samples were also collected from all major drains contributing to the river (see Figures
3.5 and 3.6).
Figure 3.8: Water quality sampling network at river Chenab.
75 Table 3.3: Water quality monitoring stations selected for the water quality study on river Chenab.
Station Description Distance (km) Northing Easting Code
SS1 Marala Headworks 0 32o 19’ 13’’ 74 o 28’ 51’’ SS2 Khanki Headworks 57.5 32o 28’ 67’’ 73 o 58’ 22’’ SS3 Qadirabad Headworks 84.8 32 o 24’ 20’’ 73 o 58’ 22’’ SS4 Chiniot Bridge 185 31o 45’ 17’ 72 o 57’ 38’’ SS5 5 km upstream of Faqirian o o Sillanwali drain outfall 218.6 31 37’ 34’’ 72 34’ 50’’
SS6 10 km downstream of Faqirian Sillanwali drain outfall 233.2 31o 37’ 35’’ 72 o 34’ 50’’
SS7 Trimmu Headworks 292 31o 38’ 41’’ 72o 31’ 58’’
3.3.3 Sample Collection
A) Sampling Methods: Grab samples of water and wastewater were collected at all
selected sampling sites. In case of sampling stations located at the river, the samples were
collected using different methods according to the situation i.e.
• By immersing the sampler with the help of rope from bridge crossing and
barrages.
• Manually sampling from the boat in the case of absence of any hydraulic
structure at the river.
The samples were collected from different points along the width of river channel and mixed to prepare a representative sample. In case of surface drains all wastewater samples were collected manually.
76 B) Sample Handling and Delievery: Table 3.4 provides guidelines about the material of the container, size, preservation methods and storage time for water and wastewater samples analysed for different water quality constituents.
Table 3.4: Summry of special sampling and handling requirements.
Maximum Min. storage Determination Container Sample Size Preservation Recommended/ (ml) Regulatory♣ Acidity P,G(B) 100 Refrigerate 24h/14d Alkalinity P,G 200 Refrigerate 24h/14d BOD P,G 1000 Refrigerate 6h/48d COD P,G 100 Analyze as soon as possible, or 7d/28d add H2SO4 to pH<2
Conductivity P,G 500 Refrigerate 28d/28d
Hardness P,G 100 Add HNO3 to pH 6months/6months Kjeldhal Nitrogen P,G 500 Refrigerate, Add H2SO4 to pH<2 7d/28d Dissolved Oxygen: Electrode G, BOD 300 Analyze immediately 0.5h/stat Wrinkler Bottle Titration may be delayed after 8h/8h acidification pH P,G - Analyze immediately 2h/stat
Salinity G, wax seal 240 Analyze immediately, or use 6months/N.S wax seal
Solids P,G - Refrigerate 7d/2-7d Temperature P,G Analyze immediately stat/stat -
Metals general P(A), G(A) For dissolved metals filter 6months/6months immediately, add HNO3 to pH<2 Source: (APHA, 1998).
P = Plastic, G = Glass, P(A) or G(A) = rinsed with 1 + 1HNO3, N.S. = Not stated in the reference, stat = no storage allowed; analyze immediately. ♣ Environmental Protection Agency of United States, Rules and Regulations. Field register 49, No. 209, 1984.
77 The guidelines for the material of the sampling containers, as presented in Table
3.4, were followed during the monitoring program. All the collected samples were refrigerated in an ice box designed for containing the sampling bottles. No chemical was added preservation as samples were collected with short holding time (< 1 day).
3.3.4 Laboratory Analysis
The collected samples were analyzed for a variety of cardinal water quality parameters. The selected water quality parameters for the analysis are listed in Table 3.5.
Table 3.5: The selected water quality parameters for laboratory and in-situ analysis.
Water Quality Parameters Unit
Bicarbonates (HCO3) mg/l HCO3 Biochemical Oxygen Demand (BOD) mg/l Calcium (Ca) mg/l Chemical Oxygen Demand (COD) mg/l Coliform Bacteria (fecal) No./100 ml Dissolved Oxygen (DO) mg/l Electric Conductivity (EC) µS/cm Magnesium (Mg) mg/l pH pH units Residual Sodium Carbonate (RSC) meq/l Sodium (Na) mg/l Sodium Adsorption Ratio (SAR) - Temperature 0C Total Coliforms No./100 ml Total Dissolved Solids (TDS) mg/l
Total Hardness mg/l as CaCO3 Total Kjeldhal Nitrogen (TKN) mg/l
78 Both in-situ and laboratory analysis of the collected samples was performed using analytical methods and guidelines published by United Nation Environment Program
(UNEP), Global Environment Monitoring System/Water Program (2004). The details of these analytical techniques are provided in Appendix-I. The calculation of SAR and RSC were made as follows:
Na + SAR = (3.1) Ca +2 + Mg +2 2
Where Na+, Ca+2 and Mg+2 are the concentrations of sodium, calcium and magnesium ions, respectively.
−2 −1 +2 +2 RSC = (CO3 + HCO3 ) − (Ca + Mg ) (3.2)
−2 −1 Where CO3 and HCO3 are the concentration of carbonate and bicarbonate ions respectively.
79 CHAPTER IV
WATER QUALITY INDEXING
4.1 THEORATICAL CONCEPTS
One of the most daunting prospects facing water quality scientist is how to turn
often very complex water quality data into information which is understandable and
usable by nonscientists e.g., managers, planners, and general public. In an attempt to
convey the information content of data more simply, resets have been made to produce
just one or perhaps a few numbers, which have been designed to integrate the data pool in
some way. Such numbers are called indices. All indexing systems require measurements
to be made for a selection of water quality determinants. From these measurements, a
sub-index rating value is obtained for each determinant. These values are then aggregated
to produce the final index score.
Water quality indices can be broadly classified into ‘objective’ and ‘subjective’
types. Objective indices are those which do not make use of any subjective inference and
are based on the expert opinion, questionnaires, etc. These are often called the statistical
indices. Subjective indices, on the other hand, need two important specifications viz weights (i.e. values according to importance value of the water quality parameters) and rating functions. These specifications are entirely subjective and are drawn out of questionnaire analysis inquiring the opinion of the experts. Unlike the objective indices, however, the subjective indices have some causal basis for representing the multivariate
(i.e., consisting of more than one water quality parameter) data. The advantage of
objective index is in terms of its unbiasedness (Ott, 1978).
80 4.1.1 Structure of Various Water Quality Indices
i. Arithmetic Water Quality Index (WQIA)
This water quality index is an index originally proposed by Horton (1965), also called as the arithmetic water quality index (WQIA). Many researchers (Brown et al.,
1970; Prati et al., 1971; Dinius, 1972) have used this index in their research work, which is basically the weighted arithmetic mean in the following form: n WQI = ∑ wiqi (4.1) A i=1
Where n is the number of variables, wi is the relative weight of the ith parameter
n such that ∑ wi = 1 and qi is the quality rating of ith parameter. i=1
ii. Multiplicative Water Quality Index (WQIM)
This water quality index is a multiplicative form of index proposed by Brown et
al., 1972. Later researchers (Landwehr et al., 1974; Walski and Parker, 1974; Bhargava,
1985, Dinius, 1987) have also employed a weighted geometric mean for aggregation. The
multiplicative water quality index (WQIM) is defined as follows:
n wi WQI M = ∏ qi (4.2) i=1
The construction of the above two indices suggests that each parameter may be of different weight based on the importance of water quality situation. The possibility that such weights may be unnecessary in distinguishing between different quality situations was explored by the formulation of two additional indexes (Landwehr, 1976).
81 iii. Un-weighted Arithmetic Water Quality Index (WQIUA)
The un-weighted arithmetic water quality index (WQIUA) is defined as given in the following equation.
n WQIUA = (1/ n)∑ qi (4.3) i=1
iv. Un-weighted Multiplicative Water Quality Index (WQIUM)
Un-weighted multiplicative water quality index (WQIUM) is defined as:
1/ n ⎛ n ⎞ ⎜ ⎟ WQIUM = ⎜∏ qi ⎟ (4.4) ⎝ i=1 ⎠ v. Harkins’ Water Quality Index (WQIH)
This index was proposed by Harkins, 1974 which is based on nonparametric multivariate ranking procedure. Harkins’ water quality index values (WQIH) are calculated as follows:
n 2 (Rin − Ric ) WQI H = ∑ (4.5) i=1 vari
Where ‘n’ is the number of parameters being used, Rin is the rank of the nth water sample according to value of the ith parameter when compared to the values of that parameter among all the p water samples, Ric is the control value of the ith parameter and vari is the rank variance for the ith parameter, which is given by:
82 ki 1 ⎡ 3 3 ⎤ vari = × ⎢( p − p) − ∑(tij − tij )⎥ (4.6) 12 p ⎣⎢ j=1 ⎦⎥
Where p is the total number of water samples in the particular data set under consideration, observations plus the number of control points, tij is the number of elements involved in the jth tie encountered in ordering the measured values of the ith parameter and ki is the total number of ties encountered in ranking the measured values of the ith parameter.
vi. Delphi Approach for Water Quality Index Calculation
Delphi method is based on human evaluation approach. This method is to survey the water quality experts and ask them to give the score for each pollution parameter. The scoring functions can be formed by regression analysis for all the pollution parameters.
The mathematical function, such as: arithmetic weighted, arithmetic unweighted, geometric unweighted, geometric weighted, minimum scoring etc are selected upon the purpose of the water use (House and Ellis, 1980). If the mathematical function needs the weight, the experts are asked to provide the relative weighting factors for all the pollution parameters. The WQI is calculated afterward.
vii. The British Columbia Index
This water quality index was developed by the Environment Protection
Department of British Columbia (Rocchini and Swain, 1995). The British Columbia approach to calculate a water quality index included a factor not considered in any of the other indices:
83
½ WQI = (F1+F2 +F3) (4.7)
Where:
F1 is the percentage of water quality guidelines exceeded.
F2 is the percentage of measurements in which one or more of the guidelines
were exceeded.
F3 is the maximum (normalized to 100) by which any of the guidelines were
exceeded.
The British Columbia index had the longest and most extensive application. As with the other jurisdictions; British Columbia used its index based on a variety of objectives. Each water body is assessed with respect to designated uses. British Columbia has different objectives for drinking, recreation, irrigation, livestock watering, wildlife, and aquatic life. Separate rankings were published based on each use relevant to the water body.
4.1.2 CWQI Model Development i. Conceptual Model
The Water Quality Index Technical Subcommittee of Canada adopted the conceptual model from the British Columbia index to develop the Canadian Water
Quality Index (CWQI) that is used in the whole country. The task group of the Canadian
Council of Ministers of the Environment (CCME) developed a computer model for the calculation of CWQI. There are three factors in the index, each of which has been scaled
84 to range between 0 and 100. Figure 4.1 shows the conceptual model for the index. The values of the three measures of variance from selected objectives for water quality are combined to create a vector in an imaginary ‘objective exceedence’ space. The length of the vector is then scaled to range between zero and 100, and subtracted from 100 to produce an index which is 0 or close to 0 for very poor water quality, and close to 100 for excellent water quality. Since the index is designed to measure water quality, it was felt that the index should produce higher numbers for better water quality.
Figure 4.1: CCME water quality index formulation.
85 The CWQI 1.0 model consists of three measures of variance from selected water quality objectives: Scope (F1), the number of variables not meeting water quality objectives; Frequency (F2), the number of times these objectives are not met; and
Amplitude (F3), the amount by which the objectives are not met. The index produces a number between zero (worst water quality) and 100 (best water quality). These numbers are divided into five descriptive categories to simplify presentation. The model calculates the WQI as follows
a) Scope -How many? - The number of water quality variables that do not meet objectives in at least one sample during the time period under consideration, relative to the total number of variables measured. It represents the extent of water quality guideline non-compliance over the time period of interest.
Number of failed variables F = ×100 (4.8) 1 Total number of variables
Where variables indicate the water quality parameters tested during the time period for index calculation.
b) Frequency – How often? - The number of individual measurements that do not meet objectives, relative to the total number of measurements made in all samples for the time period of interest. It represents the percentage of individual tests that do not meet the objectives i.e. failed tests.
86 Number of failed tests F = ×100 2 Total number of tests (4.9)
c) Amplitude – How much? – The amount by which measurements which do not meet the objectives depart from those objectives. It represents the amount by which failed test values do not meet their objectives. F3 is calculated in three steps as follows:
Step 1: 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 an ‘excursion’ and is expressed as follows. When the test value must not exceed the objective:
⎡ ⎤ Failed test value i (4.10) excursioni = ⎢ ⎥ − 1 ⎣⎢ Objective j ⎦⎥
For the case in which the test value must not fall below the objective:
⎡ Objective j ⎤ excursioni = ⎢ ⎥ − 1 (4.11) ⎣ Failed test valuei ⎦
Where i and j = 1, 2, 3,.………,n
Step 2: The collective amount by which individual tests are out of compliance is calculated by summing the excursions of individual tests from their objectives and is divided by the total number of tests (both those meeting objectives and those not meeting objectives). This variable, referred to as the normalized sum of excursions, or nse, is calculated as:
87 n excursion ∑ i (4.12) nse = i=1 Number of tests
Step 3: Finally F3 is calculated by an asymptotic function that scales the normalized sum of the excursions from objectives (nse) to yield a range between zero and 100.
⎡ nse ⎤ F3 = ⎢ ⎥ ⎣0.01nses + 0.01⎦ (4.13)
The WQI is then calculated as:
⎛ F 2 + F 2 + F 2 ⎞ WQI = 100 − ⎜ 1 2 3 ⎟ (4.14) ⎜ 1.732 ⎟ ⎝ ⎠
In Equation 4.14, the factor 1.732 arises because each of the three individual index factors can range as high as 100. This means that the vector length can reach
1002 +1002 +1002 = 30000 = 173.2 as a maximum. Division by 1.732 brings the vector length down to 100 as a maximum.
The CWQI values are then converted into rankings by using the index categorization schema as presented in Table 4.1. The model provides a mathematical framework for assessing ambient water quality conditions relative to water quality objectives. It is flexible with respect to the type and number of water quality variables to be tested, the period of application, and the type of water body (stream, river reach, lake, etc.) tested. These decisions are left to the user. Therefore, before the index is calculated,
88 the water body, time period, variables, and appropriate objectives need to be defined
(CCME, 2001).
Table 4.1: CWQI categorization schema.
Rank WQI Value Description
Water quality is protected with a virtual absence of threat or impairment; conditions very close to natural or pristine Excellent 95-100 levels. These index values can only be obtained if all
measurements are within objectives virtually all of the time.
Water quality is protected with only a minor degree of threat Good 80-94 or impairment; conditions rarely depart from natural or desirable levels.
Water quality is usually protected but occasionally threatened Fair 65-79 or impaired; conditions sometimes depart from natural or desirable levels.
Water quality is frequently threatened or impaired; Marginal 45-64 conditions often depart from natural or desirable levels.
Poor 0-44 Water quality is almost always threatened or impaired; conditions usually depart from natural or desirable levels. Source: CCME (2001).
4.2 INPUT DATA FOR CWQI 1.0 MODEL
In the present study, WQIs were calculated at seven locations along river Chenab for its three intended uses viz. drinking, irrigation and aquatic life. The selection of water quality parameters (variables) was made keeping in view the local water quality situations. The set of data used in the CWQI 1.0 model for the calculation of WQIs was adopted from the results of monitoring program conducted during low flow months of
89 2006-7 and 2007-8. In this monitoring program surface water quality was monitored using grab sampling with short holding time (<1day) on monthly basis.
An important consideration during present study was to adopt best suited water quality guidelines (objectives) for WQI calculation in context of indigenous water quality conditions. Numerous sets of standards, or guidelines for water quality, have been issued from time to time by various agencies and authorities (e.g. United States Environmental
Protection Agency (EPA), World Health Organization (WHO), European Union (EU), and other countries) intending to define the maximum acceptable limit of water pollution by various pollutants. Standards for ambient water quality are commonly designated according to the intended use of the water resource (e.g. drinking water, fishing water, irrigation etc.).
In Pakistan, a little attention has been given to formulate surface water quality standards with respect to different water uses. National Environmental Quality Standards
(NEQS) established in 1993 are available only for municipal and liquid industrial effluents and do not provide any guideline for the receiving water bodies. PSI (1987) and
PCRWR (2002) have drafted water quality standards for drinking and irrigation waters but their enforcement is still pending. The most recent advancement in the establishment of water quality guidelines was a stakeholder workshop organized by World Wide Fund
(WWF) Pakistan in November 2006 to disseminate the devised guidelines. In this workshop, stakeholders from different government departments (irrigation, agriculture and environment), Industry Water Management Institutes, Non Government
90 Organizations (NGOs) working on water issues were invited to review the criteria and guidelines in detail and voice their comments. Their comments and suggestion were incorporated in the proposed guidelines by WWF (2007). These guidelines were used as a preferred source in the present study for defining water quality criteria in CWQI 1.0 model. Table 4.2 provides a list of selected water quality parameters for WQIs calculation in different water use categories and their corresponding guidelines/standards.
Table 4.2: Water quality standards for different water uses.
Water Uses Water Quality Parameters Unit Drinking Aquatic Irrigation
Total Dissolved Solids (TDS) mg/l 800 1000 1000 Electric Conductivity (EC) µS/cm 1250 1500 1500 pH Minimum 6.5 6.5 6.4 Maximum 8.5 8.5 8.4 Total Kjeldhal Nitrogen (TKN) mg/l 0.5* 1.2* N/A Sodium Adsorption Ratio (SAR) N/A N/A 8 Residual Sodium Carbonate (RSC) meq/l N/A N/A 2.3 Dissolved Oxygen (DO) mg/l >6 >5 >4 Biochemical Oxygen Demand (BOD) mg/l 2 8 8 Chemical Oxygen Demand (COD) mg/l 25* 50* 70* Coliform Bacteria (total) No./100 ml 50 1000 1000 Coliform Bacteria (fecal) No./100 ml 20 200 500
Total Hardness mg/l as CaCO3 300 N/A N/A Chloride (Cl) mg/l 250 N/A 100 Sodium (Na) mg/l 200 N/A N/A
N/A = Not Applicable or Not Available *Official Gazette of the Turkish Government (1991) Unless otherwise specified data has been adopted from WWF (2007).
91
4.3 STEPS FOR WQI CALCULATION BY USING CWQI 1.0 MODEL
The CWQI 1.0 has been coded into a user-friendly Excel-based model.
Navigation throughout the model is done entirely by clicking the buttons along the top of each page. Following is a brief description of different pages as given on the ‘manual page’ of this model.
4.3.1 Start Page
This is start point of the model and the basic navigation homepage. Various buttons are available at this page (Figure 4.2). Each button opens a new page i.e. data, criteria, manual and the WQI pages which are discussed as follows:
Figure 4.2: The start page of CWQI model.
92 4.3.2 Data Page
Actual data is inserted in the model on this page as shown in Figure 4.3. This can be achieved by three methods:
Figure 4.3: The data page of CWQI model.
1) Type data directly into model in the blue area (if present). For reference the first
data value of the table (top left) should be in cell M4, which corresponds to cell
C3 in Table 4.3.
2) Cut and paste data from a file into model in the blue area.
3) Directly insert data from a file into model. This can be done by pressing the ‘open
new data sheet’ button and selecting the data file. The data file must be in a fixed
format with the parameter columns entered in the same order as that of the model
93 (i.e.: Date (1st column); Grouping Index (2nd column); Parameter 1 (3rd column);
Parameter 2 (4th column); etc.). It should be ensured that the first row of the data
file contains the column heading and that the first data entry must be in the third
row of your data file in order to ensure all data is transferred to the model (i.e. if
importing an Excel data file, the top left cell which contains test data should be
cell "C3" in that file).
4) Thus a data file should follow the following basic format:
Table 4.3: Format for the data entry on data page of CWQI model.
1 2 3 4 5…. Date when Location Water quality Water quality Water quality A sample was where sample parameter # 1 parameter # 2 parameter # 3…. collected was collected B Units Units Units… . C MM/DD/YYYY Site ID
In Table 4.3, column 2 is an ‘unrestricted parameter column’ in which any data which suits the computing and reference needs may be entered (e.g. region, site location, season, etc.). It is important to note that each row and column of data must have at least one cell that is not blank in order for the program to properly ‘find’ all the entered data.
Once the data is entered, the water quality indices for the various uses are calculated by pressing the ‘compute’ button available on data page.
The model also provides the facility to choose ‘site specific guidelines’ (based on test data) for selected parameters. This can be done by clicking on the corresponding
‘settings’ button to view the available options.
94 There are also sensitivity analysis, turbidity flagging, contaminant flagging and grouping subset options available on this page. These are extra options and are not compulsory for WQI calculations. The details are available on manual page of the model.
In the present study, data on water quality of river Chenab were collected during low flow months of two seasons (i.e. 2006-7 and 2007-8). This data was entered as input according to the prescribed format (Table 4.2) of CWQI model.
4.3.3 Criteria Page
The criteria page (Figure 4.4) defines the objectives for various water uses for each variable against which compliance is measured.
Figure 4.4: The criteria page of CWQI model.
95 The user can define site-specific objectives. If some variables are not needed to be considered in analyzing a particular water use, they can be excluded from calculations by entering ‘N/A’ or leaving the appropriate cells blank in that column as shown in Figure
4.4 where cells are left blank in the columns with ‘recreational’ heading. There are extra rows added to the criteria table for the inclusion of any variable that is not already in the table but is important to a particular study. By default the program contains more than 50 predefined parameters and provides 40 list boxes for easy selection of these parameters as well as two predefined data formats complete with critical and index values.
In this study, local water quality guidelines proposed by WWF-Pakistan (2007) were used for calculation of WQIs. For these calculations, user defined data format option was used on the criteria page for defining the order of selected parameters for specific water uses (irrigation, drinking and aquatic life). Some new water quality parameters (not listed in the built-in list of the criteria page) were also defined e.g. sodium adsorption ratio (SAR), residual sodium carbonate (RSC), carbonaceous oxygen demand (COD), etc.
The order of the parameters and their respective units must be the same on the data page as defined on this criteria page. The issue of matching units in the criteria page to those of the input data is up to the user as this model has no ability to confirm or refute the problem. The unit field provided in both tables is purely for ease of reference for the user to facilitate visual validation.
96 This page also defines the ‘ranking categories’ for the Water Quality Index. The default categories are those developed by the CCME (i.e. excellent, good, fair, marginal and poor) and are explained in Table 4.1. The user can change these categories to reflect his particular need.
4.3.4 Report Page
The report page (Figure 4.5) outputs the rank of the overall water quality of a water body as well as the rank of the water quality relating to each particular water use when the ‘compute’ button on the data page is pressed.
Figure 4.5: The report page of CWQI model.
97 This page also presents access to various summaries including the following:
• Number of variables.
• Number of variables that failed.
• Variables with most failed tests.
• Variables with highest normalized sum of excursions (nse).
• List of variables that failed.
• Sensitivity analysis options.
• Number of tests per variable.
• Minimum requirement of four tests per variable listing.
• Statistical Summary of data used.
• A sheet which contains the input data highlighting each individual failure for
each index tested.
• A sheet which contains highlighted data outliers from then input data.
4.3.5 The WQI Page
This page provides a mathematical description and technical details of the
Canadian Water Quality Index 1.0. These details have been discussed earlier in this chapter.
4.3.6 The WQI Chart
This page outputs a bar chart of ranks of the overall water quality of a water body as well as the ranks of the water quality relating to each particular water use when the computation is commenced. It is possible to define up to six ranks for WQI results but
98 only those which are defined on the criteria page appear on the graph and have labels in the legend.
99
CHAPTER V
HYDRODYNAMIC AND WATER QUALITY MODELING
5.1 THEORATICAL CONCEPTS AND WATER QUALITY MODELS
Water quality models are usually grouped into categories based on (Cox, 2003):
1. The environment modeled.
2. Purpose of the model.
3. Spatial characteristics.
4. How the processes are described.
5. Type of the data used.
6. Temporal characteristics.
Figure 5.1 shows the classification of common water quality models. The figure also presents subdivision of models on the basis of above categories. The first subdivision deals with the environment which is modeled. The models are categorized on the basis of their ability to simulate water quality of different surface water bodies i.e. lakes, rivers and estuaries.
Under second subdivision, the water quality models are categorized keeping in view their purpose of modeling. For example hydrochemical models are designed to model water chemistry. The hydrochemical models incorporate the chemical and biological processes which affect the quality constituents of interest. A mixing zone model deals with those portions of the modeled system located downstream or adjacent to the source of pollution into the receiving water body. While the time of travel model are
100
designed to deal with the time taken by the water quality constituent to travel downstream
of the water body in response to the added pollution at upstream. Time of travel models are usually simple in-stream model and describe the solute transport conservative in nature.
Lake River Estuaries
(1) Environment Modele d
Hydrochemical Steady-state (6) (2) Temporal Purpose Mixing zone Dynamic characteristics of the model Time of Travel
WATE R QUALITY MODE LS 0D
Stochastic (5) (3) 1D Type of Spatial Deterministic data used characteristics 2D
3D (4) How Processes are described
Empirical Mechanistic
Figure 5.1: Subdivisions of water-quality models in common use.
On the basis of spatial characteristics, the water quality models are divided into
four categories ranging from the simplest zero dimensional to very complex three
dimensional models. The dimensions simulated by a particular model provide
information on both the complexity of a model and its suitability to specific applications.
101
In zero-dimensional (0D) models, diffusive and transport terms are considered equal to zero and the contaminants are assumed to be mixed completely in any direction within the water body. These models provide an approximate description and results are in the form of gross variation in concentration of water quality variables in the entire water body. The water and contaminants are assumed to travel spatially in the downstream direction of the channel in one dimensional (1D) models. The advection and dispersion processes are also assumed to occur in just one direction (along length of the stream). The stream is assumed to be completely (and instantaneously) mixed across its width and depth. The two dimensional (2D) model, however, also include the movement of volume and concentration either along width or depth of the channel while the model which account for the flow and solute transport in all three directions (length, width and depth of the stream) are referred to as three-dimensional (3D) models. Three dimensional models are often used for deep and wide estuaries with complex mixing patterns.
The empirical and mechanistic models are two categories of water quality models on the basis of the description of processes within the model. Empirical models are black- box models. Purely empirical models, such as many statistical models, allow description of the fixed relationships between input data and output results with a minimum of understanding about how the system works. Statistical models estimate the parameters through statistical analysis and then check the adequacy of the model. One of the principal restrictions of the empirical models is that they cannot be implemented to other close systems or for data out of the range used for creation of the model (Martin, 1999).
Typically, empirical models take the form of regression relationships, and are useful for investigating cause-and-effect relationships if they are used in a formal statistical
102
environment. They are particularly useful because such models can cope with a number
of inputs with minimal computation (Kirchner et al., 1993). It is important to understand
that they can only be used with any confidence within the ranges of the data used to
parameterize them. Thus, an empirical model can never be used with confidence to
predict long term changes, while it is assumed that one can obtain some predictive
capabilities if the models are based on physical and chemical principles (Warfvinge,
1995). On the other hand mechanistic water quality models are the models that by means
of mathematical tools express the mechanisms of the process that cause changes on water
quality and lead to establish the relationship (cause and effect). The mechanistic models
have some advantages as they gain insights and increased understanding of the water
quality of a particular stream and provide information on cause and effect relationships as
well as a predictive capability.
Another division of water quality model, with regard to the type of data used by a
particular model, includes stochastic and deterministic models. Deterministic models
have a fixed relationship between input data and output results which may be empirical or
mechanistic so a deterministic model will always produce the same output given the same inputs. Stochastic models contain some random elements. The term stochastic is used to describe several different types of models but more commonly used for ‘Monte Carlo’ models. Stochastic models are commonly defined as the model which generates an output result with varying input conditions (such as boundary conditions) in the form of a frequency distribution of concentration of the contaminant.
103
The last division of water quality models in Figure 5.1 describes the way in which a model handles temporal variability. The models are called steady-state if time does not appear in any of the model equation. Including time in the model equations makes the model much complex and increase the need of calibration and verification. A dynamic model, however, simulates both spatial and temporal variability. Thus, input variables and model parameters will vary with time, and the model output will also be time varying.
With such a range of different nomenclature, it is of no surprise that the terminology used to describe water-quality models can be confused and that different authors may use the same terms to describe different techniques. Further complications arise because the definitions are in no way exclusive and so a model may be both stochastic and deterministic if it is able to run in more than one ‘mode’. Even within these broad terms there may be substantial variability (Cox, 2003).
All the water quality models, with exception to very simple approaches, are computer based. A variety of modeling codes are available for water quality simulation.
These softwares use different calculation approaches and contain varying features and limitations. Table 5.1 provides an overview of the of some important software products for river water quality modeling.
104
Table 5.1: An overview of some important software products for water quality modeling.
Software for Water Quality Modeling Components Features 1 2 3 4 5 6 7 8 9 10 Hydrodynamics Extern. Input Y Y N N Y N N N N Y Simulated N Y Y Y Y Y Y Y Y Y Control structure N N Y Y Y Y Y Y Y Y Transport Advection Y Y Y Y Y Y Y Y Y Y Dispersion Y Y Y Y Y Y Y Y Y Y
Sediment Quality models N Y Y N Y Y N N e Y
e Water quality Temperature Y N Y Y Y Y Y r N Bacteria N N Y Y Y Y Y N DO-BOD Y Y Y Y Y Y Y Y
Nitrogen Y Y Y Y Y Y Y open structur Y Phosphorus Y Y Y Y Y Y Y Y Silicon N N Y N Y Y Y *open structu N Phytoplankton Y Y Y Y Y Y Y Zooplankton N N Y N Y Y N N Benthic algae N N N N Y Y Y N Systems analysis Parameter estimation N Y Y Sensitivity/uncertainty Y Y Y analysis Source: Ambrose et al. (1996) with extensions made by Rauch et al. (1998).
1 = QUAL2 (US EPA; Brown and Barnwell, 1987); 2 = WASP5 (US EPA; Ambrose et al. 1988); 3 = CE- QUAL-ICM (US Army Engineer Waterways Experiment Station; Cerco and Cole, 1995); 4 = HEC5Q (US Army Engineer Hydrologic Engineering Center, HEC 1986); 5 = MIKE11 (Danish Hydraulic Institute; DHI, 2008); 6 = ATV Model (ATV, Germany; ATV, 1996); 7 = Salmon-Q (HR Wallingford, UK; Wallingford Software 1994); 8 = DUFLOW (University of Wageningen, The Netherlands, Aalderink et al., 1995); 9 = AQUASIM (EAWAG, Switzerland; Reichert, 1994); 10 = DESERT (IIASA; Ivanov et al, 1996). * “Open structure” refers that the user can change the model structure.
While making choice between the water quality models, it is important to be very clear about the objectives of the study and any limitations within which it will be undertaken. The World Bank (1998) has reported that “Simple models are easier to calibrate verify and use, and require less site-specific data. The one dimensional model is the preferred model. Three dimensional models, although intrinsically appealing, should be avoided whenever possible because of the large quantities of site-specific data required to ensure the reliability of the model predictions. It is important to keep model as simple as possible, but the model must fulfill the prediction requirement”. According to
105
Palmer (2001) “Multiple runs of steady state models to represent variations in time are
simpler than time-varying models”.
5.2 MIKE-11 MODEL
MIKE 11 model was selected for the simulation of water quality of river Chenab
in the present study. This is a modeling code for rivers, lakes, reservoirs, irrigation
canals, and other inland water systems. The model has been developed by the Danish
Hydraulic Institute (DHI), Denmark from the DHI’s system 11 originally released in
1972. The model contains modules for run-off simulations, hydrodynamics, flood
forecasting, transport and dilution of dissolved substances, sediment transport, and river
morphology as well as various water quality processes. It has an interface to GIS
allowing for preparation of model input and presentation of model output in a GIS
environment.
5.2.1 The Conceptual Model
MIKE 11 model is a 1D model of flow and quality in the rivers and estuaries. The hydrodynamic module simulates dynamic flows and can be applied to branched and looped networks. Because the model is 1D, the scheme assumes that the flow conditions are homogeneous within the channel although flow over structures such as weirs can be simulated. The transport of solutes is simulated by an advection–dispersion module which solves the same 1D equation of conservation of mass and a dynamic solution is provided. Thus, the conceptual model is still that of reaches in series, but here the flow is
106
modeled explicitly using the full-hydrodynamic equations and the advection dispersion
equation (ADE) is solved dynamically (Cox, 2003).
The advection–dispersion (AD) module can simulate first-order decays of
determinants. Predefined components of several water quality processes can be selected
within AD module. The water quality processes include modeling of DO and BOD with
nutrients, COD with nutrients, eutrophication, heavy metals, iron-oxidation, extended
eutrophication and nutrient transport. The component which involves the modeling of DO
and BOD corresponds to different levels of increasing complexity as shown in Figure 5.2.
Phosphorus and coliform components can also be added to any level of complexity.
Level 1 1st order decay Of DO and BOD plus reaeration and temperature effects
As level 1, plus the resuspension of organic Level 2 matters and sedimentation in the BOD calculation, and sediment oxygen demand due to nitrification is also modeled. Level 3 As level 1, plus ammonium, nitrate balance without de-nitrification
and the oxygen consumption from the nitrification process is included at this level. Level 4 This level models all processes from levels 2 and 3, including denitrification. Level 5 BOD modeled as dissolved, suspended and settled fractions and a delayed oxygen demand due to settled BOD is included. Nitrogen components ammonia and nitrates are not modeled at this level. Level 6 As level 5, plus the nitrogen components modeled as in level 4
Figure 5.2: Description of different complexity levels of water quality component of MIKE 11.
107
5.2.2 Processes of the Model
The hydrodynamic module (HD) is the core of the system and solves either the
full hydrodynamic (St. Venant) equations or one of the two simpler versions called
diffusive wave and kinematic wave equation. Writing equations for the conservation of
mass and momentum separately results in the pair of equations called the Saint Venant
equations i.e.
∂A ∂Q x + = q (Conservation of mass) (5.1) ∂t ∂x
The St. Venant equation for momentum and how the simpler forms may be derived by dropping terms which are shown as follows (Chow, et al., 1988):
⎡αQ 2 ⎤ ∂⎢ ⎥ ∂Q A ∂y + ⎣ ⎦ + gA − gA (S − S ) = 0 (5.2) ∂t ∂x ∂x o f
Local Advective Pressure Gravity Friction acceleration acceleration force force force
Kinematic wave Diffusion wave Dynamic wave
and
∂Q ∂y ∂(αQ 2 / A) = gA(S − S ) − gA − (Conservation of momentum) (5.3) ∂t o f ∂x ∂x
Where A is the wetted area (or reach volume per unit length), t is the time, Q is the discharge, x is the distance downstream, q is the lateral inflow per unit length, g is the
108
acceleration due to gravity, y is the depth, a is a momentum coefficient, S0 is the bed slope and Sf is the friction slope.
The friction slope term can be estimated using empirical formulae such as those of Manning or Chezy. Using Manning’s method:
Q Q S = (5.4) f K 2 and
A5 / 3 K = (5.5) nP 2 / 3
Where k is conveyance, n is Manning’s friction coefficient and P is the wetted perimeter. The velocity can then be estimated using the stream cross-section and friction parameters using the method of Manning or Chezy. The simpler versions are useful in steeper rivers where there will be no backwater effects because the St. Venant equations require complex numerical solutions. MIKE-11 uses an efficient implicit finite-difference model to solve the equations, but the solution of full hydrodynamic model can result in long runtimes.
The AD module describes the basic processes of river water quality in areas influenced by human activities, e.g. oxygen depletion and BOD levels as a result of organic matter loads. Concentrations of DO and BOD are calculated in MIKE 11 by taking into consideration advection, dispersion and the most important biological, chemical and physical processes. A predefined set of water quality (WQ) components is
109
available in AD module which deals with the basic aspects of river water quality. The
WQ component is coupled to the AD module, which means that the WQ module deals with the chemical/biological transforming processes of compounds in the river and the
AD module is used to simulate the simultaneous transport process. The relevant water quality components must be defined in the AD editor. The one-dimensional (vertically and laterally integrated) equation for the conservation of mass of a substance in solution, i.e. the one-dimensional advection-dispersion equation, reads as given in equation 5.6:
∂AC ∂QC ∂ ⎛ ∂C ⎞ + − ⎜ AD ⎟ = −AKC + C2 .q (5.6) ∂t ∂x ∂x ⎝ ∂x ⎠
Where Q is the discharge [L3T-1], C is the concentration (arbitrary unit), D is the dispersion coefficient [L2 T-1], A is the cross-sectional area [L2], K is the linear decay
-1 -3 coefficient [T ], C2 is the source/sink concentration [M L ], q is the lateral inflow
[L2 T-1], x is the space co-ordinate [L] and t is the time co-ordinate [T]. The equation reflects two transport mechanisms:
• Advective (or convective) transport with the mean flow.
• Dispersive transport due to concentrations gradients.
The main assumptions underlying the advection-dispersion equation are:
• The considered substance is completely mixed over the cross-section, implying
that a source/sink term is considered to mix instantaneously over the cross-
section.
• The substance is conservative or subject to a first order reaction (linear decay)
110
• Fick's diffusion law applies, i.e. the dispersive transport is proportional to the
concentration gradient.
The dispersion coefficient (D) is described as a function of the mean flow velocity
(V) as shown in equation 5.7.
D = aV b (5.7)
Where ‘a’ is the dispersion factor and ‘b’ is the dispersion exponent. Typical value of D ranges from 1 to 5 m2/s (for small streams) to 5-20 m2/s (for rivers). If the dispersion exponent is zero then the dispersion coefficient D becomes constant (equal to the dispersion factor). By default the dispersion is zero (i.e. there is only advective transport and no dispersion) (DHI, 2008).
5.2.3 Data Requirements of the Model
The MIKE 11 model provides a graphical user interface and requires a river network to be drawn on a grid (mesh) of cells that represent geographical area to a referenced scale and /or coordinates. The river network is drawn using the mouse that can be corrected in a table corresponding to this graphical presentation. Different set of data
(cross-section and hydrodynamic, advection–dispersion and water quality parameters, etc.) are entered in a series of editors which are referenced to the network by the distance along the river. The discharges and abstractions at top points of rivers and tributaries are referred to as boundary conditions and each of these is linked to a constant or time series of flow and quality entered in the boundary editor. Water-quality parameters are either defined initially in a water quality file or a predefined set of water quality parameters is
111
chosen from the water quality component of the model. Reaction rate constants can be set globally or for specific reaches identified by the distances along the river to which they apply. The length of the run and a time-step are also required to be set in simulation editor.
For present study, the HD module required input data comprising of the river flow, water level, cross-sections, bed slope, Manning’s roughness coefficient and the flow added into or withdrawn from the river at different locations. Salinity was defined as a conservative water quality component in AD module and a predefined water quality component of level 1 (including DO, BOD and temperature) was selected for the simulation of DO and BOD (non-conservative WQ parameter). The input data for AD module included salinity, temperature, DO and BOD levels of: (i) the river water at boundaries (Marala headworks) (ii) the canal water (iii) effluent of the surface drains.
The reaction rate coefficients were also provided in the AD module. The HD module was calibrated by comparing simulated and observed discharge and water level while the calibration of AD module was done by comparing observed and simulated salinity, DO and BOD at selected points of the river.
5.2.4 Operation of the Model
The model operates simulation process by completing flow calculations in the HD module. The results of HD module are used for the simulation of solute transport and water quality processes. The results of both HD and AD module are saved in separate files. Once the HD module is run successfully, its results can be used for any run of water
112
quality module to save time. In other words it is not necessary to run HD module all the times while simulating a number of water quality processes. An implicit finite difference scheme is used for the calculations of the hydrodynamic equations in HD module as well as for one dimensional ADE in AD module. The first order decay equation is solved to simulate predefined water quality components within AD module. The AD module uses a rational extrapolation method to calculate mass balances of the water-quality determinants for all reaches at all time-steps in an integrated two step procedure.
5.2.5 Outputs of the Model
MIKE- 11 provides the user with time-series of (or constant) flow, depth and concentrations of each water quality component in all reaches as an output of successful simulations of the model. A ‘save-step’ interval (e.g. daily values are produced from a model with an hourly frequency) can be specified in the model to save time and avoid very large output data. Numerous charting and statistical options are available in MIKE
11 for better presentation and description of the simulated results.
5.3 COEFFICIENTS USED IN MIKE 11 MODEL FORMULATION
5.3.1 Deoxygenation Coefficient
The deoxygenation coefficient K1 can be estimated from the field data considering both stream area and flow as constant. The concentration of a substance which is characterized by a singular reaction is given by
x −K1 U Lt = L0e (5.8)
113
Where
Lt = ultimate BOD concentration at downstream section of stream (mg/l)
Lo= ultimate mixed BOD at the outfall i.e. at X=0 (mg/l)
U= velocity in (km/d)
K1= deoxygenation coefficient (1/day)
X = distance (kilometers)
The field estimation of K1 can be obtained from semi-log plots of observed long term ultimate BOD stream data (five day stream data may also be used) as a function of distance downstream, thus in natural logs equation 1 can be written as:
− K X ln L = 1 + ln L (5.9) t U 0
A semi log plot of field data usually result in a straight line, the slope of which is
K Slope = − (5.10) U
This procedure provides the estimation of deoxygenation coefficient. Equation 5.8 can also be written as
X L K1 0 = e U (5.11) Lt
114
X Let = ∆t = time of travel in days U
L 0 = e K1∆t (5.12) Lt
Taking natural log on both sides of equation 5.11
L0 ln = K1∆t (5.13) Lt
Therefore
1 L0 K1 = ln (5.14) ∆t Lt
Total concentration in the river is simply the arithmetic addition of individual effects plus the background value. The ultimate mixed BOD concentration L0 is computed by taking a mass balance at the outfall so that
Qr Lr + qw Lw Qr Lr + qw Lw Qr Lr +W L0 = = = (5.15) Qr + qw QR QR
Where
Qr = River flow (cusec).
Qw = Wastewater flow rate (cusec).
Lw = BOD ultimate of wastewater (mg/l).
W = Mass rate of waste discharge.
QR = Total river flow after sewage discharge.
115
In the present study, estimates of deoxygenation coefficients in three reaches of the river Chenab were made by using equation 5.13. Computation of L0 is given by equation 5.14 and Lt is the ultimate BOD value in the downstream section of the stream.
The estimated values are given in table 5.2.
Table 5.2: Estimation of deoxygenation coefficient for selected reaches of river Chenab.
Length (km) Time of travel ∆t (days) K (1/day) River Reach 1
Marala-Khanki 56 2 0.25
Khanki-Chiniot 185 4 0.3 Chiniot-Jehlum confluence 292 5.5 0.7
5.3.2 Dispersion Coefficient (D)
Many theoretical and empirical formulations have been proposed to determine the longitudinal dispersion coefficient. Taylor (1954) introduced the concept of dispersion coefficient for longitudinal mixing in turbulent flow in a straight circular pipe. The extension of Taylor's concept of longitudinal dispersion coefficient to open channels has been presented by many researchers (Elder, 1959; McQuivey and Keefer, 1974; Fischer,
1975; Liu, 1977).
Several more studies have been carried out to empirically or experimentally determine the longitudinal dispersion as a function of hydraulic and geometric parameters. The results obtained from these studies fall in the range of observed values in the specific cases studied. However, in general, when applied to many of the observed cases in natural river flows, the deviation of computed values from the observed values often varies manifold (Ahsan, 2008). Attempts are continuing to improve the prediction
116
of longitudinal dispersion as close to the observed data as possible (e.g. Sooky, 1969;
Bansal, 1971; Chatwin and Sullivan, 1982; Iwasa and Aya, 1991; Jobson, 1997;
Sukhodolov, 1997, Swamee et al, 2000).
The dispersion coefficient depends on the velocity distribution as well as the channel geometry changing significantly along the length of river. Therefore field estimation of dispersion coefficient is a cumbersome task and requires lot of resources. In the present study values of dispersion coefficient were adopted from literature (i.e. MIKE
11 reference manual) based on flow and channel characteristics. Moreover MIKE 11 model requires a range of values instead of a single value of dispersion coefficient. This range serves as an input in the advection-dispersion editor. The dispersion coefficient ranging from 5 to 20 m2 /s was used in the model formulation for the present study.
5.4 STEPS FOR MIKE 11 MODEL FORMULATION
The input requirements of the model e.g. river network, cross-sections, boundary data, HD and AD parameters were set up for the model runs, for different months as well under different management scenarios. The modeling process in MIKE 11 was completed by providing related data and information in different editors. Following is the description of different phases of the model formulation and the editors in MIKE 11 modeling system:
117
5.4.1 The Simulation Editor
The simulation editor is the main editor in MIKE 11 and the corresponding file should always be the first file that is created when initiating a new project. The simulation editor serves three purposes:
1. It contains the simulation and computation control parameters.
2. It is used to start the simulation.
3. It provides a link between network editor and the other Mike11 editors. The
linkage requires a file name to be specified for each of the required editors. The
file names are provided on the input tab of the simulation editor. An alternative is
to select a file from File Menu which will recall the appropriate editor. The edit
menu can then be used to edit the objects.
This editor consists of five tabs i.e. model, input, simulation, results and start tabs.
The ‘model’ tab contains a list of all modules supported by MIKE 11 e.g. hydrodynamic, advection-dispersion, sediment transport, etc. The model includes only those module in calculation which are marked on this tab. This tab also ask user to chose the mode of simulation (unsteasy or quasi-steady).
The ‘input’ tab compiles all input files in the simulation editor (Figure 5.3). The input data files are edited in separate editors which are briefly discussed next.
118
Figure 5.3: The input tab of simulation editor in MIKE 11 model.
The ‘Simulation’ tab (Figure 5.4) contains information about the time setup
(adoptive, fixed or tabulated), simulation period and type of initial conditions used for different modules (HD, AD, ST and RR). On the ‘result’ tab, the path for storing result files is given. The frequency of storing results is also mentioned on this tab in suitable units.
119
Figure 5.4: The simulation tab of simulation editor in MIKE 11 model.
If all specified input files exist, the "Start" button on the start tab (Figure 5.5) is pressed and the simulation commences. The simulation takes place as a separate process
(MIKE11.EXE) and the progress of the simulation is reported in a separate window. Any error or warning message from the simulation is saved in a file with the same name as the simulation file and extension of .log. If any errors or warnings are encountered during simulation the user is given the choice of viewing these at the end of the simulation.
Upon completion the simulation results can be viewed using MIKE View.
120
Figure 5.5: The start tab of simulation editor in MIKE 11 model.
5.4.2 The Network Editor
The river network editor is used for editing and viewing network data. This editor gives an overview of the current setup and provides a common link to the other editors of
MIKE 11. Furthermmore, cross sections can also be edited using the river cross section editor , which is accessible from the this editor. The network editor has two main functions:
121
1. River network input and editing.
2. Overview of all model information in the current simulation.
In this editor, entire network of river Chenab was drawn. All the river alignments were digitized according to the chainage (location point along the river) from the referenced plans used in the background of editor screen. The example of this setup is shown in Figure 5.6.
Figure 5.6: The network editor of MIKE 11 model.
122
5.4.3 The Cross-Section Editor
The cross section editor manages, stores and displays all model cross section information. There are two types of cross section data; the raw survey data and the derived processed data. The raw data describes the shape of the cross section and typically comes from a section survey of the river. The processed data is derived from the raw data and contains all information used by the computer model (e.g. level, cross section area, flow width, hydraulic/resistance radius). The processed data can be calculated by the cross section editor or entered manually.
This setup to input the cross-sections of the river was completed using river survey data of the Chenab acquired from WAPDA1 and ISRIP2. Some of the cross- sections used in the model are presented in Appendix-II. The other related inputs e.g. roughness coefficients at different river reaches were also provided in this editor. Figure
5.7 shows the cross-section setting.
1 Water and Power Development Authority 2 International Sediment Research Institute
123
Figure 5.7: The cross-section editor of MIKE 11 model.
5.4.4 The Boundary Editor
This editor is the most important part of the modeling system. The boundary editor is used to specify boundary conditions of MIKE11 model. It is used not only to specify common boundary conditions such as water levels and inflow hydrographs but also for the specification of lateral flows along the river reaches, solute concentrations of the inflow hydrographs, various meteorological data and certain boundary conditions used in connection with structures applied in MIKE 11 model.
The boundary table shown in the first split window (Figure 5.8) gives an overview of the boundaries included in the model set up. The information required is the boundary description, the boundary type and the location of the boundary. In addition a boundary
124
‘id’ can be entered, although this is optional but specifying an ‘id’ can be convenient for identifying the boundary. The specified ‘id’ has no effect on the calculation. The boundary description describes the nature of the boundary. There are six different types of Boundary Description i.e. open, point source, distributed source, global, structures and closed type boundries. The boundary type specifies the kind of data required for the boundary. For each boundary description there are a number of valid boundary types.
An open boundary can be specified at the free upstream and downstream ends of the model domain. It can be specified as inflow, water level, Q-h relation, bottom level, sediment transport and sediment supply. The point source boundary condition is used at locations within the model domain where time-varying or constant lateral inflows (or outflows) occur. Its valid types are inflow and sediment transport. Both open and point source boundaries require a branch name and chainage in order to identify the location of the boundary. The inflow and water level can be time varying or constant.
The distributed source boundary condition is used along river reaches within the model domain where time-varying or constant lateral distributed inflows (or outflows) need to be specified, or where meteorological boundaries apply. This boundary condition needs a branch name and two chainages to be specified. The two chainages represent the up and downstream ends of the river reach along which the distributed boundary applies.
The order in which the chainages are specified is not important. The distributed source boundary description allows the inflow, evaporation, rainfall, heat balance, resistance factor, wind field boundary types. The global boundary condition is applied when a certain boundary conditions are valid over the entire model domain. In such cases it is not
125
necessary to specify any location. The valid boundary types are: evaporation; rainfall; heat balance; resistance factor and wind field. These boundary types are used in the same manner as distributed sources. When the structure boundary condition is selected, four types of boundaries i.e. dam, dam break, regulation structures and dam break piping can be specified. Finally the closed boundary description is used at free ends points of the model domain where a zero flux condition across the boundary is applicable. It can be used for HD, AD and sediment transport (ST) simulations. For the HD module it corresponds to a zero discharge boundary and for AD and ST modules it corresponds to zero transport across the boundary. No additional information is required except the location described by branch name and a chainage.
The valid boundry types used in the present study are Open Boudry: Inflow and
Water Level and Point Source Boundary: Inflow. Open Inflow boundaries are used to specify inflows at free branch ends (boundaries of the model domain) for HD andAD simulations.
There are three check boxes available in the second split window (Figure 5.8):
i. Include HD Calculation. This box must be checked if the discharge time series is to
be included in the water balance in the HD calculation.
ii. Include AD calculation. This box must be checked if the discharge is to be used
with a concentration to compute the mass inflow of a component in an AD
simulation. When checked, the associated concentrations are entered in the third
split window.
126
iii. Mike 12. If this check box is checked the boundary is applied to a two layered
branch.
The type of data used for the inflow boundry type (constant or time series) is also specified in this split window.
If the “include AD calculation” box is checked then the third split window becomes editable and boundaries for different AD components can be entered. The discharge specified in the second split window is used both in the water balance and in the AD calculation. In the AD calculation it is multiplied with the concentrations in order to calculate the mass inflow for different components. If only an AD simulation was to be computed (based on a previous HD simulation), the "include HD Boundaries" would be turned off. However the discharge is still needed to be specified in order to compute the mass inflow of the components to the AD model.
In the modeling setup of the present study, the observed flow data was acquired from the Drainage and Flood Zone, Punjab Irrigation and Power Department, Lahore.
Constant values (mean monthly inflows calculated from the observed data) were used as input in the boundary editor instead of flow time series (hydrograph). Similarly mean monthly values of selected water quality parameters were used in boundary editor in combination with the corresponding flows. For simplification, all incoming and off- taking canals and drains in the river system were dealt as point sources (i.e. lateral inflow or outflow where no cross-sections or boundary condition is needed)
127
Split Window 1
Split Window 2
Split Window 3
Figure 5.8: The boundary editor of MIKE 11 model.
5.4.5 The Hydrodynamic Parameter Editor
The hydrodynamic parameters editor (HD-editor) as shown in Figure 5.9 is used for setting supplementary data for simulation. Most of the parameters in this editor have default values and in most cases these values are sufficient for obtaining satisfactory simulation results. In the formulation of model for the present study, initial conditions of water depth and discharge were provided at different chainage of the river. The default values were not edited in the remaining tabs of this editor.
128
Figure 5.9: The hydrodynamic parameter editor of MIKE 11 model.
5.4.6 The Advection-Dispersion Editor
The advection dispersion editor of MIKE 11 model is presented in Figure 5.10
The editor manages seven tabs i.e.
• Components: the water quality constituents along with their units and type are
defined in this tab. The modeler can also select from a list of different built in sets
of water quality constituents.
• Dispersion: the information about dispersion factor/coefficient at various location
of the river is defined in this tab.
129
• Initial Condition: in this tab the initial conditions of selected or defined water
quality components in the river are prescribed.
• Decay: Information about the decaying or non decaying nature of the water
quality component defined the component tab is provided here. Decay
coefficients for decaying type water quality components are provided at different
location of the river.
• Additional Output: for additional output requirements i.e. mass, mass balance,
mass in branches etc. This tab contains various checkboxes.
• Sediment Layers: This tab is edited when sediment transport module is also run,
which was not the case in the present study. The tab was left unedited.
• Cohesive Sediment Transport: This tab interacts only with the sediment transport
module.
• Non-Cohesive Sediment transport: This tab is also edited only when sediment
transport module is to be run.
This study included the modeling of salinity and first order DO and BOD decay alternately. In case of salinity, it was defined as a parameter on the component tab of the
AD editor. No decay constant was provided in the editor due to conservative nature of salinity, only initial conditions were inserted on the corresponding tab. In case of DO and
BOD, on the other hand, a pre defined combination (level 1) of water quality parameters was chosen from the list available on the component tab. The BOD decay constant used on decay tab of the editor fluctuated between 0.24 to 0.69 per day for different reaches of
130
the river. Initial conditions for water quality parameters at different locations along the river were provided on the corresponding tab.
Figure 5.10: The advection-dispersion editor of MIKE 11 model
5.5 MIKE VIEW
MIKE View is a Windows application for displaying simulation results from the
DHI Software packages MIKE URBAM, MOUSE, MIKE SWMM, MIKE NET and
MIKE11. the important features of MIKE VIEW are:
• Results can be displayed and animated on Horizontal Plans and in longitudinal
profiles.
131
• Animations can be synchronized, showing MOUSE, MIKE SWMM and MIKE
11 results simultaneously.
• Time series of results can be plotted in combination with the time series from
external sources: ASCII files, Windows Clipboard, MOUSE, MIKE SWMM and
MIKE 11 time series databases.
• Q-H relations can be displayed for selected locations.
• Time series and graphs can be copied to the Windows Clipboard for use in other
applications.
• Horizontal plan presentations can be enhanced by background maps (.DXF and
.BMP format)
5.6 MODEL EVALUATION STATISTICS
The hydrodynamic and advection dispersion modules of MIKE 11 model were calibrated and validated by comparing simulated results with the observed values at different points along river Chenab. Following statistical measures were applied to these results to quantify the accuracy of model and to estimate the errors in the simulated results. These statistical measures along with their interpretation criteria are briefly discussed as follows:
5.6.1 Root Mean Square Error (RMSE)
RMSE is one of the commonly used error index statistics (Chu and
Shirmohammadi, 2004; Singh et al., 2004; Vasquez-Amábile and Engel, 2005). Although it is commonly accepted that the lower the RMSE the better the model performance, only
132
Singh et al. (2004) have published a guideline to qualify what is considered a low RMSE based on the standard deviation of the observed data (Moriasi et al., 2007).
Singh et al. (2004) stated that RMSE and MAE values less than half the standard deviation of the measured data may be considered low and either one is appropriate for model evaluation. RMSE is computed as follows:
n 0.5 ⎡1 2 ⎤ RMSE = ⎢ ∑ (X − Y ) ⎥ (5.16) ⎣n i=1 ⎦
Where
X = Observed value of the parameter being evaluated Y= Simulated value of the parameter being evaluated n = total number of observations i = numbers 1, 2, 3, …….n
5.6.2 Relative Mean Absolute Error (MAE)rel
The mean absolute error (MEA) is an error index commonly used in model evaluation. It is computed as:
⎡1 n ⎤ MAE = ⎢ ∑ X − Y ⎥ (5.17) ⎣n i=1 ⎦
The relative mean absolute error is calculated by dividing the MAE by the mean of observed values of the parameter being evaluated as shown in the equation 5.17 The value of (MAE)rel equal to zero is considered optimal.
MAE ()MAE = (5.18) rel X
133
Where
X = Mean observed values of the parameter being evaluated
5.6.3 Percent Bias (PBIAS)
Percent bias (PBIAS) measures the average tendency of the simulated data to be larger or smaller than their observed counterparts. The optimal value of PBIAS is zero, with low-magnitude values indicating accurate model simulation. Positive values indicate model underestimation bias, and negative values indicate model overestimation bias
(Gupta et al., 1999). PBIAS is calculated as:
n ∑(X − Y) PBIAS = 100× i=1 (5.19) n ∑ X i=1
Where
PBIAS is the deviation of data being evaluated, expressed as a percentage.
Moriasi et al. (2007) has suggested the use of this statistical measure on the basis of following reasons: (1) PBIAS is calculated in the similar manner as that of percent deviation which is recommended by ASCE (1993) for model evaluation and (2) PBIAS has the ability to clearly indicate poor model performance (Gupta et al., 1999).
134
5.6.4 Nash-Sutcliffe Efficiency (NSE)
The Nash-Sutcliffe efficiency (NSE) is a normalized statistic that determines the relative magnitude of the residual variance (“noise”) compared to the measured data variance (“information”) (Nash and Sutcliffe, 1970). NSE is computed as:
n ⎡ 2 ⎤ ⎢ ∑ ( X − Y ) ⎥ NSE = 1 − ⎢ i =1 ⎥ (5.20) n ⎢ 2 ⎥ ⎢ ∑ ( X − X ) ⎥ ⎣ i =1 ⎦
Where
NSE = Nash-Sutcliffe efficiency
NSE ranges between −∞ and 1.0 (1 inclusive). The value of NSE equal to 1.0 is the optimal value. Values between zero and 1.0 are generally viewed as acceptable levels of performance, whereas values less than zero indicate that the mean observed value is a better predictor than the simulated value, which depicts unacceptable performance of the model.
According to Moriasi et al. (2007) NSE is recommended for model evaluation for two major reasons: (1) it is recommended for use by ASCE (1993) and Legates and
McCabe (1999), and (2) it is very commonly used, which provides extensive information on reported values.
135
5.6.5 Coefficient of Determination (R2)
Coefficient of determination (R2) describes the degree of co-linearity between simulated and observed data. R2 describes the proportion of the variance in measured data explained by the model. The value of R2 ranges from zero to 1. A high value of R2 indicates less error in variance, and typically values greater than 0.5 are considered acceptable (Santhi et al., 2001, Van Liew et al., 2003). The coefficient of determination is computed as:
2 ⎡ n( XY) − ( X )( Y) ⎤ R2 = ⎢ ∑ ∑∑ ⎥ ⎢ 2 2 2 2 ⎥ (5.21) ⎣ []n∑∑X − ( X ) [n∑∑Y − ( Y) ]⎦
Where
R2 = Coefficient of determination
136 CHAPTER VI
RESULTS AND DISCUSSION
6.1 LOW FLOW ANALYSIS OF RIVER CHENAB
The low flow analysis was made using mean monthly flow data of river Chenab at
Marala. The flow duration curve (FDC) is presented in Figure 6.1.
5000
4500
4000
3500
ec) 3000 m u c (
e 2500 g
ar
sch 2000 i D 1500
Q50 = 685.3 cumec 1000
500
0 0 102030405060708090100 Exceedence Probability %
Figure 6.1: Flow duration curve for river Chenab at Marala using mean monthly flow data of sixty year period (1947-48 to 2006-07).
The FDC was developed using mean monthly flow data of sixty consecutive years i.e.
1947-48 to 2006-07 (Appendix III). The figure shows a distribution of flows ranging from
floods to low flows. The river flow at Marala ranged between 122 m3/s and 4381 m3/s. The
3 median flow or the flow corresponding to 50 % of time (Q50) was found 685.3 m /s. The river discharge below Q50 may be taken as the low flow part of FDC and it was regarded as the upper bound of low flow in the river during the lean flow period.
137 3500 Long term averge (1947-48 to 200-6-07)
Dscharge during 2006-07 3000 Dscharge during 2007-08 2500 discharge below 50 % exceedence prbability in FDC ) c e
m
u 2000 c e ( g r
a 1500
h sc
i Q of FDC =685.3 cumec D 50 1000
500
0 Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar High flow months of the year Low flow months of the year
Figure 6.2: Long term average of mean monthly flow in river Chenab at Marala during different months of the year.
In Figure 6.2, the river flow data at Marala was arranged on monthly basis after averaging the long term (sixty years) flow records during different months of the year. The monthly trend of flows during the monitoring period (i.e. two years; 2006-07 and 2007-08) is also plotted along with the long term average. The median flow (Q50) obtained from FDC is also plotted in the figure showing the upper bound of the low flows. It is clear from the figure that the average river flow remains below Q50 during the months from October to March.
This period may be considered as low flow season of the year. It is evident from the figure that there is a small deviation of flow pattern during the monitoring years from the long term average especially for low flow months. The only considerable deviation of river flow from long term average is in December 2006.
138 Frequency analysis of low flows in river Chenab at Marala was made based on annual
time series of minimum flows and presented in Figure 6.3. Sixty years (1947-48 to 2006-07)
low flow record was used for low flow frequency analysis (Appendix IV). The return period
was plotted on a log-scale while the magnitude of the discharges was plotted on normal scale.
14 Annual Low discharge (1947-48 to 2006-07) Log-Normal Distribution Log pearson Type III distribution 12
10
/s)
3 8 m
( e g 3 ar 30Q10 = 5.20 m /s
sch 6 i D
4
2
0 1.00 10.00 100.00 1000.00 Return Period (Yrs)
Figure 6.3: Low flow frequency curve for river Chenab at Marala.
Two theoretical distribution functions i.e. Log-Normal and Log-Pearson Type III
distribution functions were used to extrapolate the observed data. It is clear from the Figure
6.3 that both of these distribution functions have excellent fit with the low flow frequency
curve (LFFC). A low flow index (30Q10) was also computed from LFFC. This index can be
defined as the stream-flow below which the annual 30-day minimum flow falls in one year
out of 10 years as a long-term average. The 30Q10 flow was calculated as 5.20 m3/s.
139 6.2 WATER QUALITY ASSESSMENT
Table 6.1 summarizes results of water quality analysis of river Chenab at selected
locations during the low flow months of two seasons i.e. 2006-7 and 2007-8. The data is
plotted in Figure 6.4 for more elaboration. The figure shows percentiles and medians plotted
for individual water quality parameters along with water quality guidelines for different uses.
It is evident that the EC values (Figure 6.4-a) remained relatively constant (~225
µS/cm) in upper 85 km of the selected river reach followed by a slight increase of
approximately 100 µS/cm at SS3 and SS4. There is a steep rise in the median EC value at
SS6 followed by an abrupt fall at SS7. The salt concentration increased tremendously in the lower river reach (218 to 233 km) due to intervention of two major drains (i.e. Faqirian
Sillanwali (FS) and Chakbandi drains) in this river segment. EC remained higher than the permissible limits for drinking and irrigation at SS6. The addition of freshwater from river
Jehlum at 284 km (see Figure 3.6) caused dilution resulting in decreased salinity at SS7. An identical trend is observed in case of TDS concentrations in Figure 6.4-e. The ranges between
10th and 90th percentiles for EC, TDS, SAR, BOD and total coliforms were lower upstream of
SS6 than the values at or downstream of this sampling station. The median trend line formed a more or less similar wedge shape in lower river reach, for all water quality parameters except pH and TKN.
The pH values (Figure 6.4-b) remained in the range of 8.1 to 8.3 at all sampling stations. Thus pH of the river water is not much affected by the drain effluents. Median TKN
(Figure 6.4-i) peaked at SS2 (2.8 mg/l) followed by a fall at SS3 (0.8 mg/l).
140 Table 6.1: Summary of water quality analysis of river sampling during low flow seasons of 2006-7 and 2007-8.
Cardinal Water Quality Parameters Total Monitoring pH EC TDS SAR RSC TKN BOD COD Coliform Station Percentile (µS/cm) (mg/l) (meq/l) (mg/l) (mg/l) (mg/l) (cfu/100ml) th 10 7.9 196.0 150.0 1.0 0.0 0.7 3.0 18.0 20.0 th 25 8.0 217.8 171.0 1.3 0.3 1.2 7.5 20.3 21.5 th SS1 50 8.1 235.0 195.0 2.2 0.4 1.4 9.0 20.7 29.0 th 75 8.3 279.3 198.5 3.0 1.1 3.5 9.3 59.5 33.0 th 90 8.3 310.0 200.0 4.6 1.4 3.9 10.0 139.0 42.0 th 10 7.9 188.0 160.0 1.4 0.0 1.5 9.0 24.0 27.0 th 25 8.0 195.5 169.8 1.6 0.4 2.3 10.5 36.0 27.8 th SS2 50 8.1 255.0 200.0 2.0 0.7 2.8 12.0 49.0 37.0 th 75 8.4 280.0 208.5 3.9 0.8 3.7 14.3 110.5 46.0 th 90 8.5 310.0 219.0 5.1 1.0 4.5 18.0 175.0 58.0 th 10 7.9 161.8 103.5 0.8 0.1 0.1 5.5 15.5 72.5 th 25 8.0 203.0 138.0 1.3 0.5 0.3 7.0 23.0 78.0 th SS3 50 8.1 221.0 161.8 1.8 1.5 0.8 13.0 41.0 153.0 th 75 8.3 240.0 184.0 2.6 6.7 2.1 22.0 85.0 240.0 th 90 8.5 276.0 197.0 14.3 7.2 3.7 39.0 187.5 282.0 th 10 8.0 192.2 150.2 1.8 0.1 0.1 9.0 29.0 123.0 th 25 8.0 290.0 207.4 2.5 0.2 0.3 10.0 30.0 147.0 th SS4 50 8.2 348.5 240.9 3.7 0.3 1.0 15.0 43.5 211.0 th 75 8.3 416.0 261.0 10.7 6.6 3.0 28.0 107.0 258.0 th 90 8.7 484.0 293.5 12.8 9.5 4.8 50.0 221.5 434.0 th 10 7.9 85.2 138.9 2.8 0.1 0.2 7.5 25.0 188.0 th 25 8.1 262.0 174.1 4.5 0.3 0.6 9.0 36.0 226.0 th SS5 50 8.3 314.0 222.2 6.0 0.5 1.1 11.0 41.0 275.0 th 75 8.4 350.0 264.0 11.2 6.6 2.2 15.0 54.0 375.0 th 90 8.8 432.0 313.5 19.4 8.1 3.5 49.5 149.5 445.0 th 10 7.9 185.5 265.0 14.0 0.0 0.2 15.0 57.0 245.5 th 25 8.0 392.0 524.2 22.2 0.1 0.8 23.0 69.0 300.0 th SS6 50 8.1 1060.5 729.5 48.0 0.6 1.5 28.5 99.0 369.0 th 75 8.6 2660.0 1702.0 59.0 12.2 2.2 40.0 129.0 652.0 th 90 8.9 3650.0 2337.0 68.9 17.3 6.5 69.0 225.0 828.0 th 10 7.9 128.1 240.4 11.6 0.2 0.0 5.8 28.6 140.0 th 25 8.0 364.8 300.3 14.5 0.3 0.3 9.3 50.5 189.3 th SS7 50 8.1 555.0 373.0 17.6 0.7 1.1 14.0 54.0 225.0 th 75 8.5 1017.0 647.7 38.6 7.0 5.5 32.5 129.0 311.0 th 90 8.6 1084.0 688.0 51.5 8.4 8.6 55.0 158.0 454.0
141
Figure 6.4: Graphical presentation of different water quality parameters of river Chenab during low flow months of 2006-7 and 2007-8.
142 Furthermore a very high range between 10th and 90th percentiles at SS7 showed high variability in TKN concentrations among the sampling months. For the downstream reach (84-292 km), it remained relatively constant (~ 1 mg/l) except a slight increase at SS6
(1.5 mg/l) which showed that the drain effluents caused a decreasing effect on TKN concentration due to unknown chemical processes. Median TKN exceeded the permissible limit for irrigation at SS6, on the other hand it remained higher than the guideline for drinking at all the sampling stations.
SAR and RSC are very important water quality parameters for the determination of irrigation water quality. The application of irrigation water with high EC, SAR and RSC may cause salinity and sodicity problems in the receiving soils that may result in decreased crop yields. Median SAR (Figure 6.4-c) increased longitudinally from 2.2 to 6 between SS1 and
SS5, peaked at SS6 (48 which is 500 percent higher than the allowable limit) and then decreased to 17.6 at SS7. Median SAR was found higher than the guided concentration for irrigation at SS6, SS7 and SS8. Median RSC remained higher than the permissible limits for irrigation at four sampling sites i.e. SS1, SS2, SS6 and SS7. The highest value was noted at
SS6 where median RSC exceeded the limits by 113 percent.
Organic wastes include both dissolved and particulate matter composed principally of proteins, carbohydrates and fats. Biodegradable organics were measured in terms of BOD and COD. For median BOD and COD, small variations were noted in upstream reach of the river (0 to 218 km). The median of both the constituents peaked at SS6 (28.5 and 99 mg/l respectively) followed by a drop at SS7.
143 The median line for total coliforms (Figure 6.4-h) sloped upwards prominently
between SS2 and SS6. This trend is clearly the effect of untreated sewage added into this
river segment from the surrounding cities.
The discharge variations in river Chenab mainly depend on the flow of the canals and
link canals. These canals and natural streams are mainly located in the river reach between
SS1 and SS3. Downstream of SS3, surface drains (municipal as well as industrial) dispose
off their effluents into the river up to SS6 sampling station. Freshwater supply from river
Jehlum offers more dilution of the pollutants in the river at 284 km.
6.3 WATER QUALITY INDEXING
Table 6.2 presents a summary of three measures of variance for selected water uses
i.e. F1 (scope), F2 (frequency) and F3 (amplitude). The table shows that among all the water
uses except drinking, F1 has higher values than F2 and F3 at all the selected river stations. It denotes that there is higher percentage of failed variables than the percentage of individual failed tests and the amount by which they failed. It can also be seen in the table that F1 values
show an increasing trend from SS1 to SS6 followed by a drop at SS7. This trend infers that
more water quality variables failed (do not meet their objectives) in the downstream reach
polluted by the surface drains.
The highest values of F2 are observed for drinking and lowest for irrigation use of the river water showing that the percentage of individual failed tests is highest in case of
drinking and lowest for irrigation. Similarly F3 values are also higher in case of drinking as
compared with the aquatic life and irrigation uses of the river water. The reason is that for the
144 uses of aquatic life and irrigation, the values of failed variables do not exceed as largely from their objectives as for drinking use.
Table 6.2: Scope, Frequency and Amplitude for different water uses of river Chenab.
Water use Sampling Drinking Aquatic Life Irrigation Station F1 F2 F3 F1 F2 F3 F1 F2 F3 SS1 38 32 48 50 25 13 22 11 3 SS2 50 38 60 38 30 23 22 16 9 SS3 50 41 63 50 24 32 44 17 21 SS4 62 49 72 50 28 41 56 26 27 SS5 62 49 69 50 24 34 56 22 22 SS6 88 62 80 75 38 52 78 43 54 SS7 62 47 72 50 22 40 56 30 37
It can be summarized from the results presented in Table 6.2 that the river water quality was assessed on the basis of three measures: i) the number of variables (water quality constituents) which exceeded the safe limits ii) the number individual measurements not met the safe limits during the study period and iii) the amount by which these failed measurements departed from those safe limits for a particular use. The conclusion drawn from these results is that the spatial degradation of river water quality was more prominent in case of drinking than aquatic and irrigation uses on overall basis.
Figure 6.5 presents the water quality condition of river Chenab in terms of WQI and ranking based on it. Spatial variation of WQIs calculated for different water uses is shown in the figure for low flow seasons of 2006-7 and 2007-8 while discussion follows:
145 i) Drinking: The drinking water quality initially remained marginal at three upstream
stations (SS1 to SS3). The reason may be that the river is less affected by anthropogenic
activities as it traversed through the slopes of Kashmir valley before its entrance in Pakistan
at SS1 (Marala barrage). Up to SS3 (Qadirabad headworks) the river is not polluted by any
drain with significant inflow of any effluent. Downstream of SS3, the water quality
deteriorated and ranked poor for all the remaining sampling sites. The worst water quality
condition was at SS6 where WQI was the lowest (22). This was a clear indication of serious
polluting impact of surface drains as SS6 is located about ten km downstream of Faqirian
Sillanwali drain (the largest drain polluting river Chenab). The WQI was improved up to 39
at SS7 (Trimmu headworks) due to dilution caused by fresh water contribution from river
Jehlum upstream of it, but the water quality ranking still remained poor for drinking at this
station.
100 Excellent
90 Good 86 80 83 Fair 70 x 70 e 69 d 67 n 60 63 63 Marginal 61 62 61 y I 60 59 t 58 i
al 50 u
Q 50 48 Poor r 40 e 43 t 39 40 39 fe a 38 i W L
30 g c on n ti ti i k ga n i i
20 qua r 22 r A Dr I 10
0 SS 1 SS 2 SS 3 SS4 SS 5 SS 6 SS 7 Marala HW Khanki HW Qadirabad HW Chiniot bridge u/s of FS drain d/s of FS drain Trimmu HW (0 km) (57.5 km) (84.8 km) (185 km) (218.6 km) (233.2 km) (292 km) Name and location of sampling stations along river Chenab
Figure 6.5: Water Quality Indices according to different water uses at selected stations along river Chenab.
146 ii) Aquatic: The water quality was ranked fair for aquatic life at SS1 and SS2 based on
WQIs. A small variation (59 to 63) in WQIs was noted among all the sampling stations
downstream of SS2 with an exception of SS6. At SS6 the water quality was ranked poor for
aquatic life with WQI score of 43.
iii) Irrigation: At most of the sampling sites, the water quality for irrigation ranged in
good, fair and marginal categories. The only station with poor water quality was SS6
(WQI = 40) where the river was intensely polluted by the drains. It is important to note that
the water quality of the river was ranked poor for its all uses i.e. drinking, aquatic life and
irrigation, at this sampling site. Fortunately, no canal off-takes in the lower river reach (84-
292 km). On the bases of these results it is not advised to plan any irrigation canal from the river especially between SS5 and SS6.
6.4 MODELING OF RIVER WATER SALINITY
6.4.1 Model Calibration
The steady state hydrodynamic module of MIKE 11 model was calibrated using data collected during November 2006 and subsequently validated for lean flow months up to
December 2007. The calibration was done at various points along the river reach. Figures 6.6 and 6.7 show simulated results for discharge and water levels respectively at four locations along the river i.e. at Khanki headworks, Qadirabad headworks, Chiniot bridge and Trimmu headworks.
147 1400 1200 1200 Observed Modeled 1000 ) ) /s 1000 /s 3 3 800 m m ( ( 800 ge
ge 600 600 har har c c
s 400 s i 400 Di D 200 200 0 0 Nov - Dec- Jan- Feb- Mar - Oct- Nov- Dec - Nov- Dec - Jan- Feb- Mar - Oct- Nov- Dec- 06 06 07 07 07 07 07 07 06 06 07 07 07 07 07 07 Months Months b). Qadirabad HW 84 km a). Khanki HW 57.50
600 1000 900 500 800 ) ) /s /s 700 3 3 400 m m 600 ge ( 300 ge ( 500 har har 400 c 200 c s s i 300 Di D 100 200 100 0 0 Nov- Dec- Jan- Feb- Mar - Oct- Nov- Dec- Nov- Dec- Jan- Feb- Mar - Oct- Nov- Dec- 06 06 07 07 07 07 07 07 06 06 07 07 07 07 07 07 Months Months c). Chiniot Bridge 185 km d). Trimmu HW 292 km
Figure 6.6: Comparison of observed and simulated results of MIKE 11 hydrodynamic module for discharge during lean flow months of 2006-07.
It is evident from Figure 6.6 that simulated results for discharge matches very well with the observed values during all the months at all the calibration points. Various statistical measures e.g. root mean square error (RMSE), relative means absolute error (MAE)rel, percent bias (PBIAS), Nash-Satcliff efficiency and coefficient of determination (R2) were used for error estimation and to assess calibration and validation of the model.
Table 6.3 shows that there exists an excellent match of observed and modeled discharges at different calibration points during calibration and validation months. The values of RMSE are much lower than the half of standard deviation (SD) of the observed data
148 indicating that the error is considerably small. The values of PBIAS are negative for first
three calibration points showing the model underestimation bias (modeled values are less than the observed values) and at fourth point (Trimmu headworks) the value is positive showing overestimation bias. However at all four points the magnitude of bias is very small
suggesting accuracy of the model results. The model accuracy is further confirmed from high
values (close to 1) of NSE and R2 at the calibration points.
Table 6.3: Model evaluation statistics for calibration and verification of discharge during different months.
RMSE (MAE) PBIAS NSE R2 SD/2 Calibration Points rel Khanki headworks 0.344 0.001 -0.283 0.981 0.994 205.200 Qadirabad headworks 1.318 0.003 -0.283 0.990 0.992 188.586 Chiniot bridge 1.867 0.010 -0.782 0.982 0.984 91.220 Trimmu headworks 0.789 0.002 0.036 0.995 1.000 170.399
Figure 6.6 also depicts that the river flow has small variations during the period of
November 2006 to February 2007 whereas a considerable increase in river flow can be noted
in March 2007. Except at Trimmu, the lowest flow is observed in November for both the
years (2006 and 2007) which is supported from the historic flow data presented in Table 3.2
designating December and November as the lowest flow months, respectively. A different
flow trend from upstream calibration points is observed at the last point i.e. Trimmu HW
(Figure 6.6-d). The reason of this difference is that varying flow is contributed from river
Jehlum at its confluence with river Chenab above Trimmu headworks.
Comparison of observed and simulated water levels slightly disagrees (< 0.5m) as
shown in Figure 6.7. The results were also statistically evaluated as presented in Table 6.4.
The table indicates RMSE values are less than the half of SD except at Khanki. The values of
149 NSE and R2 show approximately 65 percent efficiency of the model simulation for water levels at Khanki headworks. This variation may be due to the reason that the river cross- sections used as model input were not very up to date. In the HD module, water level was used as downstream boundary condition at Trimmu headworks causing the exact match of simulated and observed water levels as depicted from all the statistical measures.
222.0 210.0
209.0 ) 221.0 ) m Observed m
( Modeled
( el el 208.0 220.0 Lev Lev r r 207.0 te e t a a W W 219.0 206.0
218.0 205.0 Nov- Dec - Jan- Feb- Mar - Oct- Nov- Dec- Nov - Dec - Jan- Feb- Mar - Oct- Nov - Dec - 06 06 07 07 07 07 07 07 06 06 07 07 07 07 07 07 Months Months
a). Khanki HW 57.50 b). Qadirabad HW 84 km
180.0 150.0
179.0 149.0 ) ) m ( (m l
el 148.0
e 178.0 v e Lev r r L
e 147.0 t e 177.0 t a a W W 176.0 146.0
175.0 145.0 Nov - Dec - Jan- Feb- Mar - Oct- Nov - Dec - Nov- Dec - Jan- Feb- Mar - Oct- Nov- Dec - 06 06 07 07 07 07 07 07 06 06 07 07 07 07 07 07 Months Months
c). Chiniot Bridge 185 km d). Trimmu HW 292 km
Figure 6.7: Comparison of observed and simulated results of MIKE 11 hydrodynamic module for water levels during lean flow months of 2006-07.
150 Table 6.4: Model evaluation statistics for calibration and verification of water levels during different months.
2 Calibration Points RMSE (MAE)rel PBIAS NSE R SD/2
Khanki headworks 0.260 0.024 0.122 0.620 0.651 0.232
Qadirabad headworks 0.366 0.007 0.024 0.690 0.828 0.388 Chiniot bridge 0.300 0.005 0.024 0.694 0.832 0.330 Trimmu headworks 0.000 0.000 0.000 1.000 1.000 0.298
6.4.2 Simulation of Salinity during Low Flow Months
Figure 6.8 shows comparison of observed and simulated profiles of EC (µS/cm) at the same locations where discharge and water levels were calibrated, as shown in Figures 6.6 and
6.7. Two additional calibration points were also selected in the last reach (Chiniot-Trimmu) due to major intervention of surface drains at 224 km and 228 km in this segment.
It is interesting to note that the river salinity does not follow a regular trend during the study period at all the calibration points except in the 1st reach (i.e. Khanki-Qadirabad) in which no surface drain joins the river (Figure 6.8). This situation also infers that the point source pollution (surface drains) mainly cause the river salinity and other pollution sources
(non-point) play a very little role. Figure 6.8 shows a good agreement between observed and simulated data on river salinity at different calibration points during calibration and validation months. The errors between observed and simulated data and the model efficiency were calculated with the help of statistical measures as presented in Table 6.5.
151
350 350 300 Modeled Observed 300
) 250 ) 250 m m c 200 c 200 S/ S/ µ µ (
150 ( 150 EC 100 EC 100 50 50 0 0 Nov- Dec - Jan- Feb- Mar - Oct- Nov - Dec- Nov - Dec - Jan- Feb- Mar - Oct- Nov - Dec - 06 06 07 07 07 07 07 07 06 06 07 07 07 07 07 07 Months Months
a). Khanki HW 57.50 b). Qadirabad HW 84 km
500 600 400 500 ) ) m
m 400 c 300 c S/ S/ µ
µ 300 ( 200 ( EC
EC 200 100 100
0 0 Nov - Dec - Jan- Feb- Mar - Oct- Nov - Dec- Nov- Dec - Jan- Feb- Mar - Oct- Nov- Dec - 06 06 07 07 07 07 07 07 06 06 07 07 07 07 07 07 Months Months d). At 218 km c). Chiniot Bridge 185 km
2500 1400 2000 1200 )
) 1000 m m c
/ 1500 c
S 800 S/ µ µ (
1000 ( 600 EC EC 400 500 200 0 0 Nov - Dec - Jan- Feb- Mar - Oct- Nov - Dec - Nov - Dec - Jan- Feb- Mar - Oct- Nov - Dec - 06 06 07 07 07 07 07 07 06 06 07 07 07 07 07 07 Months Months
e). At 233 km f). Trimmu HW 292 km
Figure 6.8: Comparison of observed and simulated results of MIKE 11 AD module for salinity during flow months of 2006-07.
152 Table 6.5: Model evaluation statistics for calibration and verification of salinity during different months.
2 Calibration points RMSE (MAE)rel PBIAS NSE R SD/2 Khanki headworks 2.518 0.001 -0.150 0.999 0.999 35.192 Qadirabad headworks 14.766 0.024 -2.410 0.978 0.963 30.818 Chiniot bridge 17.710 0.004 0.233 0.968 0.968 53.184 u/s FS drain 18.871 0.028 -2.765 0.956 0.971 48.320 d/s FS drain 21.535 0.005 0.146 0.998 0.998 273.215 Trimmu headworks 0.014 0.001 -0.001 0.999 0.999 146.223
Table 6.5 shows that NSE ranges between 0.956 to 0.999 and R2 between 0.963 and
0.999 showing a very good match between observed and simulated salinity. The values of
RMSE are less than half of SD at all the points. The magnitude of (MAE)rel is also very
small. PBIAS at Chiniot bridge and downstream FS drain have positive values showing
overestimation and at the remaining points it has negative values showing underestimation of
salinity by the model. The magnitude of PBIAS at all the points is reasonably low showing
accuracy of the model.
Figure 6.9 shows simulated profiles of the river water salinity in all the study months.
According to FAO (1994) maximum permissible salinity level for irrigation is 1500 µS/cm.
The highest salinity was observed in the last reach (Chiniot–Trimmu) of the river where there
is a steep rise in the salinity profiles (Figure 6.9). This steep rise starts at 224 km along the
river length where ‘Chakbandi’ drain joins the river from left side followed by another drain
(‘Faqirian Silanwali’, FS drain) joining at approximately 4 km downstream from right side of
the river (see Figure 3.6). These two drains are the main contributors of wastewater in the
entire river segment and collectively contribute 42 percent of total effluents disposed by all
the drains.
153 2500
Nov-07
2000
Dec-07 ) m
c Maximum limit for irrigation use / 1500 S
µ Nov-06 ( C
E Oct-07 d e ll e d 1000 o M
Khanki-Chiniot segment (98 km) Feb-07 Jan-07 500 Dec-06 Mar-07
0 83.0 108.0 133.0 158.0 183.0 208.0 233.0 258.0 283.0 Chiniot-Trimmu segment (107 km)
Distance downstream of Marala headworks (Km)
Figure 6.9: Simulated salinity profiles for lean flow months of 2006-07.
Regular monitoring of water quality of all surface drains was also undertaken during the study period. Table 6.6 presents the variation of salinity in two major drains (Chakbandi and FS drain) in all study months. The table indicates that the second drain (FS) that carries the effluents from surrounding industrial towns, is more contaminated in terms of salinity
(7100 µS/cm to 10450 µS/cm) than the first drain (Chakbandi).
Table 6.6: Variation of salinity in Chakbandi and Faqirian Sillanwali drains during lean flow months.
EC (µS/cm) Name of Drain Nov-06 Dec-06 Jan-07 Feb-07 Mar-07 Oct-07 Nov-07 Dec-07
Chakbandi 4470 3040 3205 3150 3310 4640 4820 4790
Faqirian Sillanwali (FS) 10160 8630 10410 7060 7100 10450 10280 9670
154 The simulated salinity profiles in Figure 6.9 show that salinity is very high during
Nov-06, Nov-07, Dec-07, and Oct-07 in the last reach of the river (Chiniot-Trimmu). A
closer look of Table 6.6 reveals that in these months, effluents from both the drains were more saline. Another important factor causing the rise in salinity is the less dilution of salts in this reach where varying amount of freshwater is contributed from river Jehlum at 284 km.
Figure 6.6(c) shows the amount of freshwater available in this river segment before joining of river Jehlum, in the study months. A clear relationship exists between salinity and freshwater availability in the river as the highest salinity profiles in Figure 6.9 are observed in the lowest flow months (Figure 6.6(c)). The similar research findings are presented by the
US Department of Interior (2003) as given in the review of literature. A rapid fall in the
salinity profile is also visible at downstream end due to contribution of freshwater from river
Jehlum. The mixing causes abrupt and massive decrease in salt concentration below the
confluence of the rivers.
6.5 MODELING OF DISSOLVED OXYGEN AND BIOCHEMICAL OXYGEN DEMAND
6.5.1 Model Calibration
a) Hydrodynamic (HD) Module
The hydrodynamic module (HD) was calibrated using the data of October 2007.
Figure 6.10 shows a comparison of simulated discharge and water levels with the observed
data. The results presented in the figure show that the discharge and water levels were
precisely simulated by the model and matched well with the observed data at different
calibration points along river Chenab. The figure also provides a trend of discharge variations
155 in the river stream caused by the canals, drains and river Jehlum at different locations along river Chenab.
Marala Khanki Qadirabad Chiniot Trimmu Marala Khanki Qadirabad Chiniot Trimmu 250 300 Observed
) Observed
) 250
/s 200 Simulated 3
Simulated m ( 200 m
( 150 e
evel 150 g L
ar 100 r
e 100 t a sch i 50
W 50 D 0 0 0 58 87 185 292 0 58 87 185 292 Distance along Chenab River (km) Distance along Chenab River (km)
Figure 6.10: Comparison of observed and simulated discharge and water level for October 2007.
b) Advection- Dispersion (AD) Module
The DO and BOD profiles were simulated using MIKE 11 AD module and compared with the observed data at different calibration points (Figure 6.11). The simulated values of
DO and BOD are in good agreement with the observed values.
The DO level in the lower river reach (200-275 km) decreased to alarmingly low level (< 3 mg/l) due to intervention of the surface drains in this river segment. In the following downstream reach, the DO level raised due to addition of fresh water from river
Jehlum at 284 km along river Chenab (see Figure 3.6). The river seems to assimilate the
BOD loads in the upper river segment of 100 to 200 km. An abrupt increase in the BOD concentration is observed in the reach downstream of 200 km followed by a gradual decrease.
156 7 60 SimulatedBOD 50
6 ) Observed BOD l / /l)
g 40 g 5 m m 30 (
4 D ( O 20 D
Simulated DO BO 3 10 Observed DO 2 0 0 0 0 0 0 0 0 50 50 100 150 200 250 300 10 15 20 25 30 Distance along Chenab River (km) Distance along Chenab River (km)
Figure 6.11: Calibration of MIKE 11 AD module for October 2007.
This variation in BOD profile can be justified from Table 6.7 showing BOD loading and locations of various point sources of pollution (i.e. surface drains). The table indicates
Faqirian Sillanwali (FS) drain as the largest flowing drain but interestingly it does not contribute the largest share of BOD load as well. The reason is that the BOD concentration in
FS drain is lowest (70 mg/l) among all the drains. This may be due to the fact that the FS drain carries the industrial effluent (rich in chemical substances) in which microbiological life is restricted due to unknown chemical processes. The most polluted of all the drains is
Chakbandi which is largely a municipal drain. This single drain imparts 57 percent of the total BOD pollution added in river Chenab. The BOD concentration in Chakbandi drain was highest (317 mg/l) among all the drains
157 Table 6.7: Characteristics of various point sources of pollutions along river Chenab.
Joins river at Flow DO BOD BOD Load Sr. No. Name of surface drain (km) (m3/sec) (mg/l) (mg/l) (tonnes/day)
1 Vagh Drain 116.22 (L)* 2.49 3.5 135 29.04 (8.80)♣ 2 Ahmadpur Kot Nikka Drain 154.96 (L) 0.4 0.1 103 3.56 (1.08) 3 Chakbandi Drain 223.94 (L) 6.84 0.13 317 187.34 (57.07) 4 Faqirian Sillanwali Drain 227.74 (R) 17 0.2 70 102.82 (31.32) 5 Shahjewna Drain 243.38 (R) 0.5 1.1 128 5.53 (1.68) Total 328.29 * Direction of entry into the river, L = left, R = right. ♣ Percent of total BOD load from all the drains.
c) Model Evaluation Statistics
The model was calibrated for discharge, water level, DO and BOD with the data of October 2007. Different statistical measures were applied to evaluate the accuracy of model calibration, as presented in Table 6.8.
Table 6.8: Model evaluation statistics for calibration of discharge, water level, DO and BOD during October 2007.
2 Parameters RMSE (MAE)rel PBIAS NSE R SD/2
Discharge 0.251 0.002 -0.240 0.991 0.999 43.810 Water Level 0.347 0.001 -0.093 0.986 0.989 18.323 DO 0.186 0.009 -0.892 0.958 0.961 0.487 BOD 1.691 0.025 2.508 0.985 0.986 7.395
The results shown in Table 6.8 indicate that RMSE is less than half of SD for all the calibrated parameters. Moreover the (MAE)rel is also very small for all the parameters.
PBIAS values show that the model slightly underestimated the discharge, water level and DO
158 while it overestimated BOD than the observed values during calibration. The observed and
modeled values show a very reasonable match as depicted from NSE and R2 values close
to 1.
6.5.2 Model Validation and Testing
The calibrated model was validated using data of December 07. The simulated DO
and BOD profiles are presented in Figure 6.12. A comparison of the simulated and observed
data show that the model simulated slightly low BOD than the actual conditions in the river
reach below 85 km. The deviation of simulated results from the observed data were within
the range of zero to 0.75 mg/l for DO and zero to 3.75 for BOD.
8 25 SimulatedBOD 7 Observed BOD
) 20 l l) / / 6 g g 15 m m 5 ( D ( O 10
D 4 3 Simulated DO BO 5 Observed DO 2 0 0 0 0 0 0 0 0 0 0 0 0 0 50 50 10 15 20 25 30 10 15 20 25 30 Distance along Chenab River (km) Distance along Chenab River (km)
Figure 6.12: Validation of MIKE 11 AD module for December 2007.
The evaluation statistics were also applied to evaluate the reliability of model validation for DO and BOD. The results (Table 6.9) confirmed the accuracy of model simulations as high NSE and R2 values were obtained for both the parameters. Also the
RMSE remained well within the permissible limits (< half of SD). Small values of PBIAS
159 and (MAE)rel also showed minimum errors in the simulated values of DO and BOD for validation.
Table 6.9: Model evaluation statistics for validation of DO and BOD during December 2007.
2 Parameters RMSE (MAE)rel PBIAS NSE R SD/2
DO 0.186 0.009 -0.892 0.958 0.956 0.681 BOD 1.691 0.025 2.508 0.985 0.941 3.095
The model was subsequently run for the remaining low flow months (i.e. November
07, January 08 and February 08). Figure 6.13 shows a comparison of simulated DO profiles
for all low flow months of 2007-08. The results indicate that during October 07, DO dropped
to minimum level (< 3 mg/l) in the lower river reach while it remained within acceptable
limits (> 6 mg/l) for the entire river during January 08. The downstream 60 km river segment
(from 210 to 270 km) was affected the most in terms of dissolved oxygen. The reasons are
the maximum addition of effluents and lowest flow condition in this segment of the river.
8
7
6
5 /l) g m
4 r ( Jan-08 o f O
D es
l Feb-08 i
3 of r Dec-07 ed p t a 2 l Nov-07 u m i
S Oct-07 1
0 0 50 100 150 200 250 300 Distance along Chenab River (km)
Figure 6.13: Simulated DO profiles for different months of low flow season 2007-08.
160 The simulated BOD profiles for all low flow months of 2007-08 are presented in
Figure 6.14. The comparison shows that in October 07, extremely high BOD level (51 mg/l)
prevailed in the river reach below 200 km. High BOD concentration (> 10 mg/l) were simulated even in the upper 160 km reach of the river in the same month. In the following months, the BOD level decreased due to the availability of more fresh water in the river and improvement in BOD assimilative capacity of the river. During February 08, BOD exceeded
10 mg/l only in a very small segment of the river (i.e. from 220 to 270 km).
60
50 Oct-07 for
es
l Nov-07 ofi 40 Dec-07 ) ted pr l a / Jan-08 g mul i
30 S Feb-08 D (m BO 20
10
0 0 50 100 150 200 250 300 Distance along Chenab River (km)
Figure 6.14: Simulated BOD profiles for different months of low flow season 2007-08.
6.6 MODEL APPLICATION AND MANAGEMENT SCENARIOS
The calibrated models for salinity and BOD were applied to test different management scenarios as follows:
161 6.6.1 Application of Model for Salinity
For the calibration of AD module, average monthly discharge of all the contributing surface drains was used. In fact, their discharges vary during the low flow months and generally remain within 7 to 15 percent of their respective designed capacities. Two scenarios (Table 6.10) were tested to study the sensitivity of river water salinity to the discharge variation of the contributing drains:
Table 6.10: Management scenarios tested for salinity in river Chenab.
Scenarios Description
Base condition Calibrated model for Nov 2006 with measured discharge of all the surface drains. Scenario 1 All the Surface drains with discharge fixed to 10 percent of their designed capacity. Scenario 2 All the Surface drains with discharge fixed to 20 percent of their designed capacity.
The comparison of salinity profiles for all the lean flow months (Figure 6.9) clearly indicates that the worst situation is found in November and December 2007 when surface water salinity exceeds the maximum limit for irrigation. Among the modeled months of
2006, the highest salinity values were also observed in November 2006. In this context, different scenarios were tested using calibrated model of November 2006 as base flow condition.
Figure 6.15 presents the simulated profiles of river water salinity under different scenarios. The concentration of salts approaches to a maximum level of 2650 µS/cm under scenario 1 that is much higher than the maximum limit for irrigation. At the same time a longer segment of the river has moved in high salinity zone than that for the base condition.
162 Scenario 2 shows the worst condition in terms of river salinity where the maximum salinity has reached up to 4100 µS/cm and the river stretch from 185 km to 270 km is converted into highly saline reach. It is clear from the comparison of scenarios with the base condition that the salinity of river water is highly sensitive to the amount of effluents being added into it.
4500 Calibration points ▲ 4000 Scenario 2
3500
) 3000 m c Scenario 1
(uS/ 2500 d EC e
t 2000 a
mul 1500 Si Base Condition 1000 Maximum EC for irrigation
500
0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 14 15 15 16 16 17 17 18 18 19 19 20 20 21 21 22 22 23 23 24 24 25 25 26 26 27 27 28 28 29 29 Distance from Marala headworks ( km)
Figure 6.15: Salinity profiles in the polluted river reach under different scenarios.
6.6.2 Application of Model for BOD
As BOD levels were above the permissible limits in the downstream reach of river
Chenab, alternative scenarios were tested to find how BOD within guided limits can be
achieved. Five scenarios were simulated using Mike 11 AD module as explained in Table
6.11.
163 Table 6.11: Management scenarios tested for BOD in river Chenab.
Scenario Description Base condition The calibrated model with data of October 2007. Scenario 1 Model with average flow and water quality data for all low flow months of 2007-8. Scenario 2 10 percent increase in discharge of all the drains. Scenario 3 No flow diversion from the river. Scenario 4 All drains with 60 % reduced BOD through treatment facility. Scenario 5 Combination of scenario 3 and 4.
Figure 6.16 shows a comparison of simulated BOD profiles under base condition and alternative scenarios. Among the simulated scenarios, behavior of the river in response to varying flow and water quality conditions was studied. The river flow varies among the low flow months as depicted in Table 3.2. Moreover the surface drains also contribute effluents of varying quantity and quality during the low flow period. Average values of flow and water quality were used for simulation under scenario 1. The results show that the BOD level in the entire river was lower under average condition than the base condition. In scenario 2, the
BOD in the upper 200 km river segment remained almost unaffected while in the following downstream reach it increased by 4 mg/l from that of the base condition.
Management options were tested under scenario 3, 4 and 5 to improve BOD condition of the river. In fact the flow of river Chenab is diverted for irrigation as well as to the eastern rivers of the country creating low flow conditions downstream of the diversion points. In scenario 3, the entire outflow was assumed to be available in the river. Considerable decrease in the maximum BOD level was noted in the lower reach of the river under scenario 3. The maximum BOD was decreased to 12.4 mg/l in contrast to 51 mg/l under the base conditions in the most affected river segment of 210-270 km.
164
60 Base condition Scenario 1 Scenario 2 Scenario 3 50 Scenario 4 Scenario 5
) 40 l / g m
( 30 D
BO 20
10
0 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 Distance along the river (km)
Figure 6.16: Simulated BOD profiles under different management scenarios.
Scenario 4, with 60 percent BOD treatment of drain effluents, proved to attain a drop in maximum BOD up to 20 mg/l in the most affected river segment. In scenario 5 both management options (i.e. scenario 3 and 4) were used simultaneously. The results showed that the BOD remained less than 15 mg/l in almost entire river length (85 percent of total).
The BOD dropped below 10 mg/l in the river segment (200-270 km) with the worst BOD level under based condition.
165 CHAPTER VII
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
7.1 SUMMARY
In this study water quality of river Chenab was monitored over a length of 292 km during low flow months of two seasons (2006-7 and 2007-8). The monitoring program included collection of samples from seven locations along the river and all the drains discharging into it. These samples were analyzed for a number of water quality parameters. The results of water quality analysis showed significant variation in spatial and temporal water quality of the river. As a next step, the water quality monitoring data were translated into an easily understandable and effectively interpreted form. Water quality indices (WQIs) were calculated using CCME WQI 1.0 model. These indices were divided into five descriptive categories (poor to excellent) to simplify presentation. Three intended uses of the river were incorporated for WQIs calculations i.e. irrigation, drinking and propagation of aquatic life.
Mathematical modeling was employed to assess the impacts of pollutants on river water quality. MIKE 11 model was used for the evaluation of temporal and spatial variations in surface water quality.
Salinity of the river water (conservative water quality parameter) was simulated with MIKE 11 model. Results of the study depicted high salinity levels in the lower river reaches that receive polluted effluents from two major drains namely Faqirian Sillanwali
(FS) and Chakbandi drains, respectively. The model simulated the highest salinity levels
166 for the months with lowest flow during both the sampled years (2006 and 2007).
Different scenarios were also tested to predict change in river salinity assuming varying discharges from the drains. The salinity was found highly sensitive to the quantity of the effluents added in the river.
The non-conservative water quality parameters i.e. dissolved oxygen (DO) and biochemical oxygen demand (BOD), were simulated in the second phase of modeling.
The simulated results of the model indicated low DO and a high BOD levels in the downstream river reaches particularly in the segment between 200 and 270 km. The study of management scenarios showed that maximum water quality improvement during the low flow months can be achieved if no flow is diverted from the river coupled with 60 percent reduction in BOD of the drain effluents through treatment.
7.2 CONCLUSIONS
Following conclusions were derived based on the results of the study:
1. The water quality degradation was worst in last reach of the river i.e. from 218 to
292 km. The values of almost all the water quality parameters abruptly changed in
this reach. The main reason of this water quality degradation is addition of
effluents from the surface drains. Another reason is low flow availability for
dilution as most of the river flow is diverted into canals at the barrages in the
upper river reaches. The contribution of fresh water from river Jehlum at 284 km
compensated the situation to some extent in the last river segment up to Trimmu
headworks.
167 2. The WQIs were mostly ranked as ‘marginal’ or ‘fair’ for aquatic life and
irrigation uses. However the river water was of poor quality for drinking at most
of the lowere sampling sites. Sampling site at 233.3 km (SS6) was the most
polluted of all, where the WQIs were in ‘poor’ category for all selected uses of the
river water (irrigation, aquatic and drinking).
3. Simulated results of MIKE 11 AD module indicated that the salinity of river
water increased as the river traversed through the municipalities and industrial
towns where effluents of deteriorated quality were added into the river. The
salinity in the upper river reach remained within the FAO limits for irrigation
(< 1500 µS/cm) whereas it exceeded the permissible limits in the lower reach of
233 to 284 km during low flow months of November and December 2007.
4. Simulations under different scenarios indicated that the salinity of the river was
highly sensitive to the discharge of the surface drains. The model simulations with
discharge of all the drains set to 10 and 20 percent of their respective designed
capacities showed a tremendous increase in salinity (up to 2500 µS/cm) in the
downstream river reach.
5. The results of the model simulations for non-conservative water quality
parameters indicated poor DO and BOD condition in the downstream river reach
particularly from 200 to 270 km.
6. The simulated results under different scenarios showed that an integrated
approach would be needed for the management of river water quality. The
treatment of drain effluent or reduction in flow diversion from the river did not
worked alone to get desired BOD level in the river. Assuming no flow diversions
168 from the river combined with 60 percent reduction in BOD of the surface drains,
proved to be the best option for pollution abatement in the river.
7.3 RECOMMENDATIONS
Following are the recommendations based on the results of this study:
1. Water quality monitoring of the river provides basic data to be used for detailed
analysis and modeling. The sampling of water and wastewater involves intensive
traveling and financial resources. Water quality of the rivers and surface drains
should be regularly monitored with proper planning and as a joint effort of
government agencies and research institutions.
2. In many cases it is more feasible to establish permanent monitoring stations at
highly polluted sites by involving local community in the adjacent vicinities of the
river. This can also be helpful in creating a public awareness about the water
quality issue.
3. Unfortunately, no surface water quality standards have been established in
Pakistan. National standards are available only for wastewater (industrial and
municipal effluents) but these are rarely enforced. National standards for surface
water quality should be established and enforced for sustainable use of water
resources in the country.
4. Treatment plants (at least primary treatment) should be designed and installed
particularly on larger and highly polluted surface drains e.g. Chakbandi drain
joining at 224 km from left and Faqirian Sillanwali drain at 228 km from right
side of the rive Chenab.
169 5. Flow diversions into the canals at upstream barrages should be operated by
promising minimum flow availability in the river for the dilution of added
pollution in the downstream reaches. On the basis of results of the present study it
is unadvisable to plan any irrigation diversion from the lower river segment
especially from 218 km to 270 km.
6. Constraints of data and resources restricted the modeling of many cardinal water
quality parameters. More water quality parameters should be modeled and their
impact on the adjacent aquifers should also be incorporated in the future research
studies.
170 REFERENCES
• Aalderink, R.H., Klaver, N.J. and Noorman, R. (1995). DUFLOW V 2.0: Micro- computer package for the simulation of 1-dimensional flow and water quality in a network of open water courses. Modeling water quality and flow in river Vecht using DUFLOW. In: Heatwole C. (ed.), Proceedings of the International Conference on Water Quality Modeling, ASAE, Orlando, FL, USA.
• Ahmad, K. (1988). GEMS/Water quality monitoring program Phase-2 Report, Institute of Environmental Engineering and Research, University of Engineering and Technology, Lahore, Pakistan.
• Ahmad, K. and Ali, W. (2000). Evaluation of Ravi River water quality. Journal of Drainage and Water Management, 4 (1&2): 27-40.
• Ahsan, N. (2008). Estimating the Coefficient of Dispersion for a Natural Stream. Proceedings of World Academy of Science, Engineering and Technology.
• Ambrose, R.B. Jr., Wool, T.A., Connolly J.P. and Schanz R.W. (1988). WASP4, A hydrodynamic and water quality model-Model theory, user's manual and programmer's guide, Report EPA/600/3-87/039, U.S. EPA, Athens, GA, USA.
• Ambrose, R.B., Barnwell, T.O., McCutcheon, S.C. and Williams, J.R. (1996). Chapter 14: Computer models for water quality analysis. In: Water Resources Handbook, L.W. Mays (ed.), McGraw-Hill, New York.
• Amusja, A.Z., Gutnichenko, V.G. and Shelutko, V.A. (1988). On the Estimation of Mean Annual Value of Minimum Winter Flow of Ungauged Rivers. Report, Trans. State Hydrol. Inst., Leningrad, USSR, 335: 43–53.
• Anderson, A., Nagar, R. and Sarkar, S. (2007). Spatial and temporal trends in surface water quality in a segment of the San Antonio River, Texas. Developments in Environmental Sciences, 5: 591-608.
• APHA (1998). Standard Methods for the Examination of Water and Wastewater 20th edition. American Public Health Association. Baltimore, Maryland, USA.
• Armentrout, G.W. and Wilson, J.F. (1987). Assessment of Low Flows in Streams in Northeastern Wyoming. USGS Water Resources Investigations Report, 85- 4246: 30.
• Aron, G. and Emmanuel, G.B. (1982). Evaluation of Drought Potentials and Damages in the Northeastern United States. Completion Report, Institute for Research on Land and Water Resources. USA.
• ASCE. (1993). Criteria for evaluation of watershed models. Journal of Irrigation and Drainage Engineering. 119(3): 429-442.
171 • Atkins, J.B. and Pearman, J.L. (1995). Low Flow and Flow-Duration Characteristics of Alabama Streams. USGS Water-Resources Investigations Report, 93-4186: 269.
• ATV (1996). Allgemein verfuegbares Gewaesserguetemodell, Projektabschlussbericht 02 WA9104/4. ATV Hennef, Germany.
• Bahzad, A. (2002). Application of QUAL 2E Model to the Contaminent Transport in River Ravi. M.Sc. (WRM) Thesis, Centre of Excellence in Water Resources Engineering, University of Engineering and Technology, Lahore, Pakistan.
• Balco, M. (1977). Multi-parameter regression analysis of low flows. Vodohospodarsky Casopis, 26 (4): 378–385.
• Ball, R.O. and Church, R.L. (1980). Water quality indexing and scoring. Journal of Environmental Engineering Division, ASCE, 106 (4): 757–771.
• Bansal M.K. (1971). Dispersion in natural streams. Journal of Hydraulic Division, ASCE, 97(11): 1867 – 1886.
• Bengraine, K. and Marhaba, T.F. (2003). Using principal component analysis to monitor spatial and temporal changes in water quality. Journal of Hazardous Materials, 100: 179–195.
• Bhargava, D.S. (1983). Use of a water quality index for river classification and zoning of Ganga River. Environmental Pollution (Series B), 6: 51–67.
• Bhargava, D.S. (1985). Expression for drinking water supply standards. Journal of Environmental Engineering Division, ASCE, 113 (3): 304–316.
• Biswas, H. and Bell, B.A. (1984). A method for establishing site-specific stream design flows for waste load allocations. Journal of Water Pollution Control Federation, 56 (10): 1123–1130.
• Brown, L.C. and Barnwell, T.O. (1987). The enhanced stream water quality models QUAL2E and QUAL2E-UNCAS: Documentation and User Manual, Report EPA/600/3-87/007, U.S. EPA, Athens, GA, USA.
• Brown, R.M., McClelland, N.I., Deininger, R.A. and O’Connor, M.F. (1972). Water Quality Index-Crashing, the Psychological Barrier, Proceedings of 6th Annual Conference, Advance in Water Pollution Research, pp 787–794.
• Brown, R.M., McClelland, N.I., Deininger, R.A. and Tozer, R.G. (1970). A water quality index do we dare? Water and Sewage Works, 117 (10): 339–343.
• CCME (2001). Canadian Water Quality Guidelines for the Protection of Aquatic Life: CCME Water Quality Index 1.0, Technical Report. In: Canadian
172 Environmental Quality Guidelines, 1999. Canadian Council of Ministers of the Environment, Winnipeg.
• Cerco C.F. and Cole T. (1995). User's guide to the CE-QUAL-ICM three dimensional eutrophication model, release version 1.0,Technical Report EL-95- 15, US Army Eng. Waterways Experiment Station, Vicksburg, MS, USA.
• Cervione, M.A., Richardson, A.R. and Weiss, L.A. (1993). Low-Flow Characteristics of Selected Streams in Rhode Island. USGS Water-Resources Investigations Report 93-4046, 16.
• Chang, H. (2005). Spatial and temporal variations of water quality in the han river and its tributaries, Seoul, Korea, 1993–2002. Water, Air, and Soil Pollution, 161(1-4): 267-284.
• Characteristics of low flows. (1980). Journal of Hydraulic Engineering. ASCE, 106 (HY5): 717–737.
• Chatwin P.C. & Sullivan P.J. (1982). The effect of aspect ratio on longitudinal diffusivity in rectangular channel. Journal of Fluid Mechanics, (120): 347 – 358.
• Chiang, S.L. and Johnson, F.W. (1976). Low flow criteria for diversions and impoundments. Journal of Water Resoures Plannig and Management, ASCE, 102 (WR2): 227–238.
• Chilundo, M., Kelderman, P. and O´keeffe, J.H. (2008). Design of a water quality monitoring network for the Limpopo River Basin in Mozambique. Physics and Chemistry of the Earth, (In press).
• Chow, V.T., Maidment, D.R. and Mays, L.W. (1988). Applied Hydrology. McGraw-Hill Inc., New York.
• Chu, T. W. and Shirmohammadi. A. (2004). Evaluation of the SWAT model’s hydrology component in the piedmont physiographic region of Maryland. Transactions of the ASAE. 47(4): 1057-1073.
• Churchill, M.A., Elmore, H.L. and Buckingham, R.A. (1962). Prediction of stream reaeration rates. International Journal of Air and Water Pollution, 6: 467- 504.
• Couillard, D. and Lefebvre, Y. (1985). Analysis of Water Quality Indices. Journal of Environmental Management, 21: 161-179.
• Cox, B.A. (2003). A review of currently available in-stream water-quality models and their applicability for simulating dissolved oxygen in lowland rivers. The Science of the Total Environment, 314-316: 335–377.
173 • Crabtree, B., Earp, W. and Whalley, P. (1996). A demonstration of the benefits of integrated wastewater planning for controlling transient pollution. Water Science and Technology, 33(2):209 –218.
• Crosa, G., Froebrich, J., Nikolayenko, V., Stefani, F., Galli, P. and Calamari, D. (2006). Spatial and seasonal variations in the water quality of the Amu Darya River (Central Asia). Water Research, 40: 2237- 2245.
• Cruz, R.T. (1997). The Pasig River, Philippines: case study. In: Water Pollution Control - A Guide to the Use of Water Quality Management Principles. R. Helmer and I. Hespanhol (eds.) on behalf of UNEP and WHO, ISBN 0 419 22910 8.
• Cumming Cockburn Ltd. (1990). Assessment of the Biologically Based Low Flow Analysis Technique. Ontario Ministry of the Environment, Canada, 10.
• de Azevedo, L.G., Gates, T.K., Fontane, D.G., Labadie, J.W. and Porto, R.L. (2000). Integration of water quantity and quality in strategic river planning. Journal of Water Resources Planning and Management, 126(2): 85-87.
• DHI (2008). Reference Manual of MIKE 11: a microcomputer based modeling system for rivers and channels. Danish Hydraulic Institute, Hoersholm, Denmark.
• Dinius, S.H. (1972). Social accounting system for evaluating water resources. Water Resources Research, 8 (5): 1159–1177.
• Dinius, S.H. (1987). Design of an index of water quality. Water Resources Bulletin, 23(5): 833-843.
• Dunnette, D.A. (1979). A geographically variable water quality index used in Oregon. Journal of Water Pollution Control Federation, 51(1): 53–61.
• Duvail, S., and Hamerlynck, O. (2003). Mitigation of negative ecological and socio-economic impacts of the Diama dam on the Senegal River Delta wetland (Mauritania), using a model based decision support system. Hydrology and Earth System Sciences, 7(1): 133–146.
• en.wikipedia.org/wiki/Chenab_River, downloaded on August 14, 2008.
• Ensink, J.H., Mahmood, T., van der Hoek, W., Raschid-Sally, L., Amerasinghe, F.P. (2004). A nation-wide assessment of wastewater use in Pakistan: an obscure activity or a vitally important one? Water Policy, 6: 1-10.
• Even, S., Mouchel, J.M., Servais, P., Flipo, N., Poulin, M., Blanc, S., Chabanel, M. and Paffoni, C. (2007). Modelling the impacts of Combined Sewer Overflows on the river Seine water quality. Science of the Total Environment, 375: 140–151.
174 • Eyre, B.D. and Pepperell, P. (1999). A spatially intensive approach to water quality monitoring in the Rous River catchment, NSW, Australia. Journal of Environmental Management. 56(2): 97-118.
• FAO (1994). Water Quality for Agriculture. Irrigation and Drainage Paper No.29, Rev. 1, Food and Agriculture Organization, United Nations, Rome.
• Fischer H.B. (1975). Simple method for predicting dispersion in streams Discussion by R.S. McQuivey and T.N. Keefer. Journal of Environmental Engineering Division, ASCE, 101(3): 453 – 455.
• Foundation for Water Research (1994). A planning guide for the management of urban wastewater discharges during wet weather. Urban Pollution Management (UPM), United Kingdom.
• FREND (1989). Flow Regimes from Experimental and Network Data, I: Hydrological Studies; II: Hydrological Data, Wallingford, UK.
• Giese, G.L. and Mason, R.R. (1993). Low-Flow Characteristics of Streams in North Carolina. USGS Water-Supply Paper 2403: 29.
• GoP (2005). Expert determination regarding the design of the Baglihar Hydro- Electric Plant. Ministry of Irrigation and Power, Government of Pakistan.
• GoP (2007). Surface water quality monitoring in Punjab during February 2007. Directorate of Land Reclamation, Government of Punjab, Lahore.
• GoP (2008). Pakistan Statistical Year Book 2008. Federal Bureau of Statistics, Ministry of Economic Affairs and Statistics, Islamabad, Pakistan.
• Gupta, A.K., Gupta, S.K. and Patil, R.S. (2003). A Comparison of Water Quality Indices for Coastal Water. Journal of Environmental Science and Health, Part A, 38(11: 2711-2725.
• Gupta, H. V. Sorooshian, S. and Yapo. P. O. (1999). Status of automatic calibration for hydrologic models: Comparison with multilevel expert calibration. Journal of Hydrologic Engineering. 4(2): 135-143. • Harkins, R.D. (1974). An objective water quality index. Journal of Water Pollution Control Federation, 46 (3), 588–591.
• Harris, J. and Middleton, B.J. (1993). Calculations of low flows for selected South African rivers and implications for water quality. Proceedings of the Sixth South African National Hydrology Symposium, Pietermaritzburg, South Africa, vol. 2: 655–664.
175 • HEC (1986). HEC-5 Simulation of flood control and conservation systems, Appendix on water quality analysis, Report CPD-5Q, Hydrologic Engineering Center, U.S. Army Corps of Engineers, Davis, CA, USA.
• Heejun, C. (2005). Spatial and temporal variations of water quality in the Han river and its tributaries, Seoul, Korea, 1993 to 2002. Water, Air and Soil Pollution, 161(18): 267-284.
• Horton, R.K. (1965). An index number for rating water quality. Journal of Water Pollution Control Federation, 37 (3): 300-306.
• House, M. and Ellis, J.B. (1980). Water quality indices: an additional management tool? Water Technology, 13: 413-423.
• Hughes, G.H. (1981). Low-Flow Frequency Data for Selected Stream-Gauging Stations in Florida. USGS Open-File Report No. 81-69, 10 pp.
• Hussain, S.A. (1995). Pollution of Surface Waters in Punjab. M.Sc. (Environmental Engineering) Thesis, Institute of Environmental Engineering and Research, University of Engineering and Technology, Lahore, Pakistan.
• Hutson, S.S. (1988). Instream flow analysis and monitoring procedure for little river in eastern Tennessee. Proceedings of the AWRA Symposium on Water-Use Data for Water Research Management, Bethesda; Maryland, USA, 67–75.
• Institute of Hydrology. (1987). Low flow estimation in Scotland. Institute of Hydrology, Report No. 101, Wallingford, UK.
• Ivanov, P., Masliev, I., De Marchi, C. and Somlyody, L. (1996). DESERT- Decision Support System for Evaluating River Basin Strategies, User's Manual. International Institute for Applied Systems Analysis, Laxenburg, Austria.
• Iwasa Y. & Aya S. 1991. Predicting Longitudinal Dispersion Coefficient in Open- Channel Flows. Proceedings of International Symposium on Environmental Hydraulics¸ Hong Kong, 505 – 510.
• Jamil, M. and Latif, M. (2006). Prediction of water quality of Indus river downstream of Ghazi using QUAL 2K model. Engineering News, Pakistan. 42 (6 &7): 5-22.
• Jobson H.E. (1997). Predicting travel time and dispersion in rivers and streams. Journal of Hydraulic Engineerin, 123(11): 971 – 978.
• Kamal, M.M., Malmgren-Hansen, A., and Badruzzaman, A.B.M. (1999). Assessment of pollution of the river Buriganga, Bangladesh, using a water quality model. Water Science and Technology, 40 (2): 129–136.
176 • Kaurish, F.W. and Younos, T. (2007). Development of a standardized water quality index for evaluating surface water quality. Journal of American Water Resources Association, 43: 533-545.
• Kazmi, A.A. and Hansen, I.S. (1997). Numerical models in water quality management: a case study for the Yamane River (India). Water Science and Technology, 36(5): 193-200.
• Khan, A.T., and Vongvisessomjai, W. (2002). Mike 11 Application for Salinity Intrusion in Southwest Bangladesh. Proceedings of 1st Asia-Pacific DHI Software Conference, 17-18 June. 2002. Bangkok, Thailand.
• Khan, F., Hussain, T., and Lumb, A. (2003). Water quality evaluation and trend analysis in selected watersheds of the Atlantic regions of Canada. Environmental Monitoring and Assessment, 88: 221-248.
• Kirchner, J.W., Dillon, P.J. and LaZerte, B.D. (1993). Separating hydrological and geochemical influences on acidification in spatially heterogeneous catchments. Water Resources Research, 29: 3903-3916.
• Krusche, A.V., de Carvalho, F.P., de Moraes, J.M., de Camargo, P.B., Ballester, M.V.R., Hornink, S., Martinelli, L.A. and Victoria, R.L. (1997). Spatial and temporal water quality variability in the Piracicaba River Basin, Brazil. Water Resources Bulletin, 33(5): 1117-1123.
• Kulkarni, G.J. (1997). In: Water Supply and Sanitary Engineering. 10th edition, Ahmadabad, India. pp 497.
• Landwehr, J.M. (1979). A statistical view of a class of water quality indices. Water Resources Research. 15(2): 460–468.
• Landwehr, J.M. and Deininger, R.A. (1976). A comparison of several water quality indices. Journal of Water Pollution Control Federation, 48 (5): 954–958.
• Landwehr, J.M., Deininger, R.A., McClelland, N.L. and Brown, R.M. (1974). An objective water quality index. Discussion by Ralph D. Harkins. Journal of Water Pollution Control Federation, 46 (7): 1804–1809.
• Legates, D. R. and McCabe. G. J. (1999). Evaluating the use of “goodness-of-fit” measures in hydrologic and hydroclimatic model validation. Water Resources Research. 35(1): 233-241. • Liu H. (1977). Predicting dispersion coefficient of stream. Journal of Environmental Engineering Division, ASCE, 103 (1): 59 – 69.
• Lumb, A., Halliwell, D. and Sharma, T. (2006). Application of CCME water quality index to monitor water quality: a case study of the Ackenzie River basin, Canada. Environmental Monitoring and Assessment, 113: 411–429.
177 • Ma, J., Ding, Z., Wei, G., Zhao, H. and Huang, T. (2008). Sources of water pollution and evolution of water quality in the Wuwei basin of Shiyang River, Northwest China. Journal of Environmental Management,( In press).
• Male, J.W. and Ogawa, H. (1982). Low Flows of Massachusetts Streams. Water Resource Research Center, Amherst Publication No. 125: 160.
• Martin, J. and McCutcheon. (1999). Hydrodynamics and Transport for Water Quality Modeling. CRC Press, New York.
• McBride, G.B. (2002). Calculating stream reaeration coefficients from oxygen profiles. Journal of Environmental Engineering, 128(4), 384-386.
• McClelland, N.I. (1974). Water Quality Index Application in the Kansas River Basin, Report, US Environmental Protection Agency, Kansas City, MO, EPA- 907/9-74-001.
• McQuivey R.S. and Keefer T.N. (1974). Simple method for predicting dispersion in streams. Journal of Environmental Engineering Division ASCE, 100(4): 997 – 1011.
• Midgley, D.C., Pitman, W.V. and Middleton, B.J. (1994). Surface Water Resources of South Africa 1990. Water Research Commission Report No. 298/5.1/94, Pretoria, South Africa.
• Milovanovic, M. (2005). Water quality assessment and determination of pollution sources along the Axios/Vardar River, Southeastern Europe. Proceedings of International Conference on New Water Culture of South East European Countries-AQUA, 21–23 October 2005, Athens, Greece.
• Mitchell, M.K. and Stapp, W.B. (1996). Field Manual for Water Quality Monitoring: An Environmental Education Program for Schools, Thomson-Shore Inc. pp 176.
• Mohammed, M., Stednick, J.D. and Smith, F.M. (2002). Comparison of field measurements to predict reaeration coefficients, K2, in the application of water quality model, QUAL-2E, to a tropical river. Water Science and Technology, 46(9): 47-54.
• Moriasi, D.N. Arnold, J.G., Van Liew, M.W., Bingner, R.L., Harmel, R.D. and Vieth, T.L. (2007). Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Transactions of the ASABE. 50:(3), 885–900. • Morishita, T. (1988). Environmental hazards of sewage and industrial effluent on irrigated farmlands in Japan. In: Treatment and use of sewage effluents for Irrigation, M.B. Pescod and A. Arar. Sevenoaks (eds.), United Kingdom, Butterworths.
178 • Musiake, K., Takahasi, Y. and Ando, Y. (1984). Statistical analysis on effects of basin geology on river flow regime in mountainous areas of Japan. Proceedings of Fourth Cong. Asian & Pacific Reg. Div. Int. Assoc. Hydraul. Res., Bangkok, APD-IAHR/Asian Institute of Technology, 2: 1141–1150.
• Nash, J. E. and Sutcliffe. J. V. (1970). River flow forecasting through conceptual models part I-A discussion of principles. Journal of Hydrology. 10(3): 282-290. • NESPAK (2007). Strengthening flood forecasting and warning system: Flood forecasting model for Chenab River. Report No. 216/063/15/07. National Engineering Services, Lahore, Pakistan.
• Nhan, N.T.T. (2005). Assessment of Urban Water Quality in Nhieu Loc–Thi Nghe Basin, Vietnam. Thesis (M.Eng.) No. WM-04-01. Asian Institute of Technology, Thailand.
• O’Connor, D.J. and Dobbins, W.E. (1958). Mechanism of reaeration in natural streams. Transaction of the American Society of Civil Engineering, 123: 641-667.
• Official Gazette of the Turkish Government (1991). Article on technical methods for water quality control regulations. No. 20748. Ankara, Turkey.
• Old, G.H., Leeks, G.J.L., Packman, J.C., Smith, B.P.G., Lewis. S. and Hewitt, E. J. (2006). River flow and associated transport of sediments and solutes through a highly urbanized catchment, Bradford, West Yorkshire. Science of the Total Environment, 360: 98–108.
• Ott, W.R. (1978). In: Environmental Quality Indices Theory and Practice. Ann Arbor Science Publishers, Michigan.
• Palmer, M.D. (2001). Water Quality Modeling: A Guide to Effective Practice. The World Bank. Washington, D.C.
• Paulson, C.L. and Sanders, T.G. (1987). Evaluation of Design Flow Criteria for Effluent Discharge Permits in Colorado. Colorado Water Resources Research Institute, Fort Collins, Completion Report No. 147, 312 pp.
• PCRWR (2002). Irrigation water quality manual. Pakistan Council of Research in Water Resources. Ministry of Science and Technology, Government of Pakistan, Islamabad.
• Perona, E., Bonilla, I. and Mateo, P. (1999). Spatial and temporal changes in water quality in a Spanish River. The Science of the Total Environment, 241: 75- 90.
• Pesce, S.F. and Wunderlin, D.A. (2000). Use of water quality indices to verify the impact of Córdoba City (Argentina) on Suquía River. Water Research, 34(11): 2915-2926.
179 • Post, D.A., Kinsey-Henderson, A.E., Stewart, L.K., Roth, C.H., and Reghenzani, J. (2003). Optimizing drainage from sugar cane fields using a one-dimensional flow routing model: a case study from Ripple Creek, North Queensland. Environmental Modeling and Software, 18 (8–9): 713–720.
• Prairie, J.R., Rajagopalan, B. (2007) A basin wide stochastic salinity model Journal of Hydrology, 344: 43– 54.
• Prati, L., Pavenello, R. and Pesarin, F. (1971). Assessment of surface water quality by single index of pollution. Journal of Water Research, 5: 741–751.
• PSI (1987). Pakistan standards (1932-1987) specification for drinking water. Pakistan Standards Institution, Karachi.
• Radwan, M. and El-Sadek, A. (2005). Control Structures Impact on River Protection and Development. Proceedings of ICID 21st European Regional Conference, 15-19 May 2005 - Frankfurt (Oder) and Slubice - Germany and Poland.
• Rauch, W., Henze, M., Koncsos, L., Reichert, P., Shanahan, P., Somlyódy, L. and Vanrolleghem. P. (1998). River water quality modeling: State of the art. Proceedings of the IAWQ Biennial International Conference. 21-26 June 1998. Vancouver, British Columbia, Canada.
• Refsgaard, J.C. and Hansen, E. (1976). Economic value of low flow data. Nordic Hydrol, 7 (1): 57–72.
• Reichert P. (1994). AQUASIM-A tool for simulation and data analysis of aquatic systems. Water Science and Technology, 30(2), 21-30.
• Riggs, H.C. (1985). Stream Flow Characteristics. Elsevier, Amsterdam.
• Rocchini, R. and L.G. Swain. (1995). The British Columbia Water Quality Index. Water Quality Branch, Environmental Protection Department, British Columbia Ministry of Environment, Land and Parks. pp 13.
• Royal Commission on Sewage Disposal (1912). Eighth report, HMSO, London.
• Said, A., Stevens, D.K. and Sehlke, G. (2004). An innovative index for evaluating water quality in streams. Environmental Management, 34(3): 406–414.
• Sakovich, V.M. (1990). The Representation of Minimum Runoff Series in Karelia and North-Western Region. Trans. Hydro-meteorological Center USSR 308: 129– 136.
• Santhi, C. Arnold, J. G., Williams, J. R., Dugas, W. A., Srinivasan, R. and Hauck, L. M. (2001). Validation of the SWAT model on a large river basin with
180 point and nonpoint sources. Journal of American Water Resources Association. 37(5): 1169-1188. • Searcy, J.C. (1959). Flow Duration Curves. United States Geological Survey, Washington, DC, Water Supply Paper 1542A.
• Shrestha, S. and Kazama, F. (2007). Assessment of surface water quality using multivariate statistical techniques: A case study of the Fuji river basin, Japan. Environmental Modelling and Software, 22: 464-475.
• Silvia, F.P., Daniel, A.W. (2000). Use of water quality indices to verify the impact of Córdoba city (Argentina) on Suquia River. Water Research, 34(11): 2915-2926.
• Singh, J. Knapp, H. V. and Demissie, M. (2004). Hydrologic modeling of the Iroquois River watershed using HSPF and SWAT. ISWS CR 2004-08. Champaign, Ill.: Illinois State Water Survey. Available at: www.sws.uiuc.edu/pubdoc/CR/ISWSCR2004-08.pdf.
• Singh, K.P., Malik, A. and Sinha, S. (2005). Water quality assessment and apportionment of pollution sources of Gomti river (India) using multivariate statistical technique- a case study. Analytica Chimica Acta, 538: 355–374.
• Smakhtin, V.U. (2001) Low flow hydrology: a review. Journal of Hydrology, 240: 147–186.
• Smith, D.G. (1987). Water Quality Indexes for Use in New Zealand’s Rivers and Streams, Water Quality Centre Publication No. 12, Water Quality Centre, Ministry of Works and Development, Hamilton, New Zealand.
• Smith, D.G. (1989). A new form of water quality index for rivers and streams. Water Science and Technology, 21(2): 123–127.
• Smith, D.G. (1990). A Better Water Quality Indexing System for Rivers and Streams. Journal of Water Research, 24(10): 1237–1244.
• Somlyódy, L., Henze, M., Koncsos, L., Rauch, W., Reichert, P., Shanahan, P., and Vanrolleghem P. (1998). River water quality modeling III, Future of the Art. Water Science and Technology, 38(11): 253-260.
• Sooky A.A. (1969). Longitudinal dispersion in open channels. Journal of Hydraulic Division, ASCE, 95(4): 1327 – 1346.
• Streeter, H. W. and Phelps, E.B. (1925). A study of the pollution and natural purification of the Ohio river. Public Health Bulletin No. 146. U.S. Public Health Service, Washington D.C.
181 • Sukhodolov A.N., Nikoa V.I., Rowinski P.M., & Czernuszenko W. (1997). A case study of longitudinal dispersion in small lowland rivers. Water Environment Research, 69(7): 1246 – 1333.
• Swamee P.K., Pathak S.K. & Sohrab M. (2000). Empirical relations for longitudinal dispersion in streams. Journal of Environmental Engineering, 126(11): 1056 – 1062.
• Swamee, P.K. and Tyagi, A. (2000). Describing water quality with aggregate index. Journal of Environmental Engineering, ASCE, 126(5): 451–455.
• Szolgay Jr., J. (1977). On the study of monthly low flows. Vodohospodarsky Casopis, 25 (2), 135–144.
• Tariq, M.N. and Ziai, K.H. (1980). Waste Assimilative Capacity of Ravi River. Report No. 045, Institute of Public Health Engineering (IPHE) and Research, University of Engineering and Technology, Lahore, Pakistan.
• Thompson, J.R., Sorenson, H.R., Gavin, H., and Refsgaard, A. (2004). Application of the coupled MIKE SHE/MIKE 11 modelling system to a lowland wet grassland in southeast England. Journal of Hydrology, 293(1–4): 151–179.
• Thyssen, N., Erlandsen, M., Jeppesen, E. and Ursin, C. (1987). Reaeration of oxygen in shallow, macrophyte rich streams: relationship between the reaeration rate coefficient and hydraulic properties. Int. Rev. Ges. Hydrobiol, 72(5): 575- 597.
• Train, R.E. (1972). The quest for environmental indices. Science, 178: 121.
• U.S. Department of the Interior (2003). Quality of Water Colorado River Basin Progress. Report 21, Bureau of Reclamation, Salt Lake City, UT, USA.
• UN/ECE. (2000). Guidelines on Monitoring and Assessment of Transboundary Rivers. Task Force on Monitoring and Assessment, RIZA.
• United Nations Environment Program (UNEP) (2004). Analytical Methods for Environmental Water Quality, Global Environment Monitoring System (GEMS)/Water Program, Burlington, Ontario, L7R 4A6, Canada.
• Van Liew, M. W., Arnold, J. G. and Garbrecht, J. D. (2003). Hydrologic simulation on agricultural watersheds: Choosing between two models. Transactions of the ASAE. 46(6): 1539-1551.
• Vazquez-Amábile, G. G. and Engel. B. A. (2005). Use of SWAT to compute groundwater table depth and stream flow in the Muscatatuck River watershed. Transactions of the ASAE. 48(3): 991-1003.
182 • Vega, M., Pardo, R., Barrado, E. and Debân, L. (1998). Assessment of seasonal and polluting effects on the quality of river water by exploratory data analysis. Wat. Res, 32(12): 3581-592.
• Vladimirov, A.M. (1990). Gidrologicheskije Raschety (Hydrological Calculations). Gidrometeoizdat, Leningrad, USSR.
• Walkowicz, J. (1978). Conversion of minimum summer flows of 50% probability into other probabilities. Arch. Hydrotechol. 25 (2): 277–287.
• Wallingford Software (1994). Salmon Q - User documentation version 1.01, Wallingford, Oxfordshire, UK.
• Walski, T.M., and Parker, F.L. (1974). Consumer’s water quality index, Journal of Environmental Engineering, ASCE, 100(3): 593–611.
• Ward, R.C., Timmerman, J.G., Peters, C.A., Adriaanse, M. (2004). In search of a common water quality monitoring and terminology. In: Timmerman, J.G., Willem, H., Behrens, A., Bernardini, F., Daler, D., Ross, P., van Ruiten, K.J.M., Ward, R.C. (Eds.), Proceedings of Information to Support Sustainable Water Management: From Local to Global Levels, Monitoring Tailor-Made IV, conference. St. Michielsgestel, Netherlands. pp. 195–206.
• Warfvinge, P. (1995). Basic principles of frequently used models. In: Trudgill S.T. (ed.). Solute modelling in catchment systems. Wiley.
• Western, A. W. (1994). The Behaviour of Salinity and Density Stratified Flows in the Wimmera River, Australia. PhD thesis, Department of Civil and Environmental Engineering, University of Melbourne.
• WHO/UNICEF (2006). "Joint Monitoring Programme for Water Supply and Sanitation. World Health Organization, Pakistan.
• WMO (1974). International Glossary of Hydrology, World Meteorological Organization, Geneva.
• World Bank (1998). Pollution Prevention and Abatement Handbook. World Bank, Washington, D.C.
• World Bank (2005). Pakistan Country Water Resources Assistance Strategy, Water Economy: Running Dry. Report No. 34081-PK. Agriculture and Rural Development Unit, South Asia Region.
• WWF (2007). National water classification criteria and irrigation water quality guidelines for Pakistan. World Wide Fund (WWF)-Pakistan, Lahore.
183 • Xie, H., Tong, X., Qiu, Y., Zhang, H. and Zhao, J. (2006). Remote sensing based water quality monitoring and spatial-temporal analysis in Huangpu River, Shanghai. Proceedings of IEEE International Geoscience and Remote Sensing Symposium, Denvar, USA. July 31- August 4, 2006.
• Yevstigneev, V.M. (1990). Rechnoj Stok and Gidrologicheskie Raschety (River Runoff and Hydrological Calculations). Moscow University, Moscow.
• Zalants, M.G. (1992). Low-Flow Frequency and Flow Duration of Selected South Carolina Streams through 1987. USGS Water-Resources Investigations Report 91-4170: 87.
• Zandbergen, P.A. and Hall, K.J. (1998). Analysis of the British Columbia water quality index for watershed managers: A case study of two small watersheds. Water Quality Research Journal of Canada. 33(4): 519–549.
184 APPENDIX-I
ANALYTICAL TECHNIQUES FOR THE ANALYSIS OF WATER AND WASTEWATER SAMPLES.
Code Method Description Name Units Deci- mals 02055 SALINITY SALINITY ppt 0 TDS-Salinity-Conductivity Meter at 25°C Specific Conductance of a solution is the ability of the solution to carry electric current and has some relationship to the Total Ionic Concentration of the solution. The salinity is measured through a calibrated TDS-salinity-conductivity meter (Orion 105 or 115, or equivalent) and corrected to 25oC, where the salinity is compared to a salinity table with a precision of 0.5% and reported in parts per thousand. The range is 0 to 80 ppt. The salinity-conductivity meter is calibrated on a per use basis.
02041 ELECTRICAL CONDUCTANCE Conductivity Meter ELEC. µS/cm 0 Specific Conductance of a solution is the ability of the COND. solution to carry electric current and has some relationship to the Total Ionic Concentration of the solution. The specific conductance is measured by a conductivity meter with Pt electrodes and is equilibrated to 25oC before the sample measurement is made. The conductivity meter is calibrated on a per use basis. Note: This parameter was formerly measured in ^mho/cm. As a result of the change to the metric system, the unit now is µS/cm (microsiemens/cm). 1 µmho/cm = 1 µS/cm. The method detection limit is 2 µS/cm.
185 02055 SALINITY SALINITY ppt 0 TDS-Salinity-Conductivity Meter at 25°C Specific Conductance of a solution is the ability of the solution to carry electric current and has some relationship to the Total Ionic Concentration of the solution. The salinity is measured through a calibrated TDS-salinity-conductivity meter and corrected to 25oC, where the salinity is compared to a salinity table with a precision of 0.5% and reported in parts per thousand.The range is 0 to 80 ppt. The salinity- conductivity meter is calibrated on a per use basis. 02050 TOTAL DISSOLVED SOLIDS TDS mg/L 0 Calibrated Conductivity Meter at 25°C The Total Dissolved Solids (TDS) is measured through a calibrated conductivity meter (Orion 105 or 115, or equivalent) and corrected to 25oC, where the values are compared to the Critical Table Values, with a relative standard deviation (RSD) of 0.87% and an accuracy of ± 0.5%. The resolution is 3 significant digits or 1 mg/L and the range varies between 0 and 19900 mg/L.
02061 TEMPERATURE Mercury Thermometer TEMP oC 1 Both atmospheric and water temperature are measured upon sample collection. The atmospheric temperature is measured in a well-ventilated area and in the shade, at 1.2 to 1.5 m above the ground, using a 50oC calibrated (liquid in glass) thermometer. The water temperature is measured by immersing a calibrated thermometer into the water or by measuring the temperature immediately after collection using a calibrated thermometer. The Hg-filled thermometer has a precision of ± 0.1oC.
186 02062 TEMPERATURE TEMP 1 Water Temperature is measured with a battery-operated YSI thermistor (or equivalent) calibrated against a certified thermometer. Precision is 0.1oC and accuracy is 0.2 oC. The method detection limit is 0.1oC.
10301 pH pH pH 1 pH Meter Electrometry Units The pH meter, with a glass combination electrode and automatic temperature compensation probe, is calibrated with buffers at pH 4.7 and 10 at 25oC. The pH and temperature values of the sample aliquot are recorded upon reading. The precision is within 0.1 pH units. Reference: APHA 1998, 4500-H B Electrometric Method.
08101 DISSOLVED OXYGEN Winkler Method DISS O2 mg/L 1
A sample is collected and analyzed in the field or preserved O2 at 4oC and analyzed as soon as possible. A sample aliquot is
treated with manganous sulphate (MnSO4) and a strong
alkaline iodide reagent (NaN3, NaI and NaOH). The manganous hydroxide formed reacts with the dissolved
oxygen to form a brown precipitate (MnO(OH)2 (a KF solution is added if ferrous ions are present). Upon acidification, in the presence of iodide, the iodine liberated is equivalent to the dissolved oxygen originally present in the sample. The iodide is titrated with a standardized sodium
thiosulphate solution (Na2S2O3), starch is used as an indicator. Interferences: ferrous ion at 1 mg/L (if KF is added the
interference level for ferrous ion is 100-200 mg/L), SO3 ion,
S2O3 ion, polythionate ions, free Cl2, OCl ion, oxidizing and reducing agents, and turbidity. The method detection limit is 0.01 mg/L.
187 08102 DISSOLVED OXYGEN, oxygen meter DISS O2 mg/L 1
Measurements are made in the field using a calibrated O2 dissolved oxygen meter. The electronic cell, containing a gold cathode and a silver anode, is covered with an Oxygen permeable membrane to prevent interferences. Upon entering the cell, the Oxygen is reduced and the current is directly proportional to the oxygen concentration at a specific temperature. The DO ranges are usually automatically temperature corrected (between -50C and + 40oC). Regular calibration against the Winkler Titration Method is recommended or by exactly following the manufacturer's procedure. (DO meter is calibrated in air saturated with moisture and the reading is taken when steady condition is obtained). The method detection limit is 0.1 mg/L. Reference: APHA 1998, 4500-0 G Membrane Electrode Method.
08202 BIOCHEMICAL OXYGEN DEMAND (BOD) Five days BOD BOD 0 incubation at 20oC mg/L
02 BOD is a measure of the oxygen demand produce by carbonaceous and nitrogenous materials in a sample. It is measured by determining the decrease of oxygen content, using a dissolved oxygen meter, after incubation at 20oC for five days. A sample is preserved in the field at 4oC and the analysis started within four hours. The sample aliquot is incubated, at 20oC for five days under proper conditions. The procedure depends on the nature of the sample. After aeration of the samples to bring the dissolved oxygen content to saturation, one of the following three variations can be used, depending on the type of samples to be analysed: 1 - The direct method: If the BOD does not exceed 7 mg/L,
188 then the BOD is determined directly by measuring the dissolved oxygen content of the water, before and after a five days incubation period at 20oC. 2 - Unseeded dilution method: For waters having BOD values greater than 7 mg/L: appropriate sample aliquots are diluted using dilution water, saturated with oxygen, and the oxygen content is determined before and after the incubation period. A minimum of three dilutions per sample, with a final content between 40% and 70% of the original oxygen concentration, will give best results. 3 - Seeded dilution method: It is extremely important that the conditions be appropriate for the living organisms to function unhindered during the incubation period. Toxic substances should be absent, and necessary nutrients, such as nitrogen and phosphorus, should be present. It is important that a mixed group of organisms (called "seed") should be present during the test. The dilution water is seeded with the proper kind and number of organisms and saturated with oxygen (overnight) before the BOD test. Siphon the diluted sample to fill three BOD bottles; one for incubation (five days), one for the determination of the dissolved oxygen content (measured and record as "initial DO") and the other for the determination of the immediate dissolved oxygen demand (IDOD), after a 15 minutes incubation period. A minimum of three dilutions per sample, with a final content between 40% and 70% of the original oxygen concentration, will give best results. Use a calibrated oxygen meter to measure the oxygen concentrations before and after incubation. Interference: Many synthetic organic components from industrial wastewaters are not biodegradable without adding seeding water due to the toxic effect or the absence or deficiency of appropriate microorganisms. Sample containing residual Cl2, that is acidic or alkaline, must be
189 neutralized to pH=7, and sometimes titrated with a Na2S2O3 solution to liberate the chloride from solution. A sample, containing sulphide, sulphite and/or ferrous ions, creates an immediate demand, corrected by the IDOD. The method detection limit is 1 mg/L. Reference: APHA 1975, 422F and 507. 08203
08301 CHEMICAL OXYGEN DEMAND K2Cr2O7 digestion COD mg/L 0
Most organic compounds are oxidised by potassium O2 dichromate under acid condition. A sample is preserved in the field at 4oC. The sample aliquot
is refluxed for two hours in concentrated H2SO4 with a
known amount of K2Cr2O7, containing sulphamic acid against
the interference of nitrites, HgSO4 against the interferences of chlorides and Ag2SO4, as a catalyst for organic compounds. The sample is cooled and the excess dichromate is titrated with standardised ferrous ammonium sulphate
(Fe(NH4)2(SO4)2), using ferroin (a complex of ferrous ion and 1,10 phenanthroline) as an indicator. The amount of oxidisable organic matter is proportional to the dichromate consumed. A reagent blank is identically analysed. The concentration of COD is calculated from the difference between sample and blank aliquots. The method detection limit is 1 mg/L. Reference: Environment Canada 1974. 10603 HARDNESS TOTAL EDTA Titration HARDNESS mg/L 1 TOTAL If turbid, the sample aliquot is filtered through a 0.45 µm CaCO3 membrane filter. The titration method depends on the ability of the ethylenediamine tetraacetic acid (EDTA) and its sodium salts to form stable unionised complexes with calcium and magnesium ions. A buffer solution (NH4Cl, NH4OH, and Mg salt of EDTA) is added to a sample aliquot to adjust the pH between 10.1 - 10.2, followed by an
190 indicator (Eriochrome Black T*) forming a pink complex. Upon titration, the EDTA removes the calcium and magnesium from the complex dye and changes the solution to its original blue colour as an end point. Interference: Total heavy metal ion concentration of 0.5 mg/L. The method detection limit is 1 mg/L. Reference: Environment Canada 1974. 20003 CALCIUM – TOTAL Ca TOTAL mg/L 3 Atomic Absorption Spectrometry Ca A sample is preserved in the field at 4oC. A LaCl3 solution is added to the sample aliquot, mixed and aspirated. The absorbance is measured spectrometrically at 422.7 nm, and compared to identically-prepared standard and blank solutions. The method detection limit is 0.002 mg/L. Reference: Environment Canada 1974.
20101 CALCIUM - DISSOLVED Ca DISS mg/L 3 EDTA Titration Ca A sample is filtered in the field through a 0.45 Mm membrane filter and preserved at 4oC. The pH of the sample aliquot is adjusted between 12 and 13, with a 1 N NaOH solution, to precipitate the magnesium to its hydroxide form. Add Calver II) indicator and titrate the aliquot with a standardised EDTA (disodium dihydrogen ethylenediamine tetraacetate) solution. The colour changes from pink to purple when the calcium is removed. The samples are compared to identically-prepared standard and blank solutions. Interferences: Total heavy metal ion concentration of 0.5 mg/L. The method detection limit is 0.5 mg/L. Reference: Environment Canada 1974.
191 12103 MAGNESIUM - DISSOLVED Mg DISS mg/L 0 EDTA Titration Mg A sample is filtered in the field through a 0.45 |im membrane filter and preserved at 4oC. The pH of the sample aliquot is
adjusted to 10.0 ± 0.1 with a buffer (NH4Cl, NH4OH and Mg EDTA salt) solution; an indicator (Eriochrome Black T) is added and the aliquot then is slowly titrated with EDTA within five minutes to avoid precipitation. The colour changed from a wine red to a blue colour (a fluorescent light is highly recommended to see the complete disappearance of the red). Interference: Ca ion concentration of 1 mg/L, total heavy metal ion concentrations of 0.5 mg/L. The method detection limit is 1 mg/L. Reference: Environment Canada 1974. 11002 SODIUM - TOTAL Flame Photometry Na TOTAL mg/L 0 The sample is collected and preserved in a polyethylene Na bottle at 40oC. A sample aliquot is mixed with lithium nitrate and passed into the burner of a flame photometer equipped with filters to isolate the spectral lines of sodium. The intensity of light produced is proportional to the concentration of sodium in the sample and compared to identically-prepared standard and blank solutions, using propane and air flame. The method detection limit is 0.1 mg/L. Reference: Environment Canada 1979. 11201 SODIUM ADSORPTION RATIO (SAR): Difference SAR Relativ 3 Calculation e Excess Sodium in irrigation water, relative to calcium and Units magnesium or to total salt content, can affect soil structure, soil aeration, flow rate, permeability, infiltration, etc. The ratio can be calculated as follows:
192
Method 1:
Na + SAR = Ca 2 + + Mg 2 + 2 Method 2: SAR = 1.41*0.04350*Na / SQRT(A) Where A = 0.01988*TH , if Total Hardness (TH) is present or A = 0.04990*Ca + 0.08226*Mg , if TH is not present. If Na is not present, or A cannot be calculated because of lack of sufficient parameters, SAR is not calculated. Caution: These calculated results are computed from measured analytical values according to the formula indicated. The computations may be in error if the parameters used in the calculation are subsequently edited or changed. Reference : Environment Canada 1988. 07401 NITROGEN ORGANIC DISSOLVED, Kjeldahl with ORG. NIT.- mg/L 1
removal of NH3 DISS A sample is collected in the field and preserved at 4oC. The shaken sample aliquot is neutralized, if necessary, to pH=7. A phosphate buffer solution (pH=7.4) is added. If Ca ion exceeds 250 mg/L, more buffer solution is added, and the solution is titrated to pH=7.4. Approximately one third of the sample is distilled to remove free NH3. The residual solution is digested with concentrated H2SO4, in the
presence of HgSO4 (as a catalyst) and K2SO4 to give
NH4HSO4. The solution is made alkaline and the NH3 is
distilled and collected in a H3BO3 solution. The distillate is
then titrated with 0.02N H2SO4, using an 'N Point' indicator and compared to identically-prepared standard and blank solutions. The method detection limit is 0.5 mg/L.
193 Environment Canada 1974.
36002 COLIFORMS - TOTAL Membrane Filtration COLIFORM No/100 0 Minor variations in analytical techniques can cause change in TOTAL mL results; therefore microbiological methods and sterilizing procedures must be standardized to obtain uniform results from different laboratories. The Membrane Filter technique is usually more rapid and more reproducible than the multiple-tube technique in monitoring drinking water and a variety of natural waters. MF has limitations: i.e.: testing samples with high turbidity or large background (non- coliforms) bacteria. Apply sufficient medium (lauryl tryptose broth) in fermentation tubes, incubate at 20oC overnight before use and discard tubes with growth or bubbles. The coliform group is defined as all bacteria that produce a red colony with a metallic (golden) sheen within 24 hours at 35oC on an Endo- type medium containing lactose (production of aldehydes). The medium is stable for a maximum of three weeks and the broth for four days at 4oC. Filter 100 to 1000 mL of water, place filter paper on saturated lactose pad for two hours at 35°C, remove from incubator and transfer to M-endo medium pad, incubate for 20 to 22 hours at 35oC ± 0.5oC, then count the colonies on membrane filters using a 10 to 15 times magnifying binocular wide field dissecting microscope or equivalent, with a cool white fluorescent light source directed to provide maximum viewing of the sheen. APHA 1995. 36012 FAECAL COLIFORM BACTERIA, Membrane FAEC COL No/100 0 Filtration Elevated temperature distinguishes faecal coliforms from total coliforms. Minor variations in analytical techniques can cause change in analytical results; therefore microbiological
194 methods and sterilising procedures must be standardized to obtain uniform results from different laboratories. Method 1: Filter a volume of sample (to yield counts of 20 to 80 faecal coliform colonies), rinse with sterile water between filtration; analyse a blank membrane filter and a duplicate sample after every 10 samples. Place a sterile absorbing pad in each culture dish and saturate with M-FC medium. Place the prepared filter on medium pad, insert in waterproof container and incubate by placing in a plastic bag and immerge in a water bath at 44.5oC ± 0.2oC for 24 hours. Colonies produced by faecal coliform bacteria on M-FC medium are various shades of blue (nonfaecal coliform colonies are grey to cream coloured.). Count colonies on membrane filters using a 10 to 15 times magnifying binocular wide field microscope, with a cool white fluorescent light. Count coliforms based on 100 mL of sample. Method 2: A measured volume of water sample is filtered through a sterile cellulose ester membrane where the pore size is small enough to retain the organisms to be enumerated. The membrane is placed on an absorbent pad saturated with membrane lauryl sulphate broth (containing lactose and phenol red as indicator of acidity) and incubated 4 hours at 30oC then 14 hours at 44oC. The colonies of organism with characteristic colour and morphology are counted with subsequent confirmation of the ability to produce acid and gas from the lactose broth and indole formation from tryptophan broth. The results are expressed in number per 100 mL of sample.
Source: United Nations Environment Program (UNEP) (2004). Global Environment Monitoring System (GEMS)/Water Program, .Analytical Methods for Environmental Water Quality, Burlington, Ontario, L7R 4A6, Canada.
195 *Eriochrome Black T (Indicator Preparation; calcium analysis) Sodium salt of 1-(1-hydroxy-2-naphthylazo)-5-nitro-2-naphthol-4- ulfonic. Dissolve 0.5 g dye in 100 g triethanolamine or ethylene glycol monomethyl ether. Add 2 drops per 50 mL solution to be titrated. Adjust volume if necessary. If the end point colour change of this indicator is not clear and sharp, it usually means that an appropriate complexing agent is required. If NaCN inhibitor does not sharpen the end point, the indicator probably is at fault. Stable for one year.
References:
• American Public Health Association. 1975. Standard Methods for the Examination of Water and Wastewater, 14th edition, APHA, New York.
• American Public Health Association. 1995. Standard Methods for the Examination of Water and Wastewater. 19th edition, APHA, New York.
• American Public Health Association. 1998. Standard Methods for the Examination of Water and Wastewater. 20th edition, APHA, New York.
• Environment Canada. 1974. Analytical Methods Manual. Water Quality Branch, Environment Canada, Ottawa.
• Environment Canada. 1988. NAQUADAT Dictionary of Parameter Codes, Water Quality Branch, Environment Canada, Ottawa.
196 APPENDIX-II
SELECTED CROSS-SECTIONS OF CHENAB RIVER USED IN THE FORMULATION OF MIKE 11 (HD) MODEL
242
) 240 m ( l ms 238 om r f n o i
at 236 ev l E
234
232 0 3 5 8 2 89 46 04 61 66 23 80 38 95 52 49 21 33 48 73 11 16 21 25 28 33 37 42 46 51 55 Width of channel (m)
Cross-section at 11076 m downstream Marala.
233
232
231 ) m ( l 230 om ms
r 229 on f
ati 228 ev l E 227
226
225 0 457 793 854 915 976 1037 1098 1220 1677 2287 2896 3506 4116 4726 5305 5457 6159 6402 Width of channel (m)
Cross-section at 33160 m downstream Marala.
197
225
224
223
) 222 m ( l s 221 om m
r 220 on f i 219 at ev l
E 218
217
216
215 0 457 915 1220 1311 1402 1829 2287 2744 3201 3659 4116 4268 4726 5183 5335 Width of channel (m)
Cross-section at 64132 m downstream Marala.
221
220
219 )
m 218 l ( s m 217 om r 216 ion f at ev
l 215 E 214
213
212 0 244 305 610 854 915 1067 1372 1677 1982 2287 2591 2896 3201 3506 3811 4116 4421 4726 5030 5335 5640 5915 Width of channel (m)
Cross-section at 73979 m downstream Marala.
205
204
) 203 m ( l s
m 202 m o r n f o
i 201 at ev l
E 200
199
198 0 457 610 762 220 677 134 591 896 049 140 232 354 811 994 085 573 840 924 030 488 945 402 1 1 2 2 2 3 3 3 3 3 3 4 4 4 4 5 5 5 6 Width of channel (m)
Cross-section at 106295 m downstream Marala.
198 187
186
185 )
m 184 l ( s
m 183 m o r f
n 182 o i at ev
l 181 E 180
179
178 0 7 0 7 0 4 6 4 8 9 4 9 6 3 1 3 3 1 42 61 106 125 152 161 213 234 243 274 304 350 396 442 457 518 567 Width of channel (m)
Cross-section at 1174995 m downstream Marala.
168
167 ) m 166 ( l s m om r 165 on f i at ev l 164 E
163
162 0 457 762 220 677 134 591 896 049 201 659 116 573 817 183 640 970 1 1 2 2 2 3 3 3 4 4 4 5 5 5 Width of channel (m)
Cross-section at 233138 m downstream Marala.
163
162
161
) 160 m ( l s 159 m m o r 158 on f i
at 157 v e l E 156
155
154
153 0 305 610 915 1220 1524 1829 2134 2439 2683 2744 2866 2973 3049 3201 3506 3811 Width of channel (m)
Cross-section at 265113 m downstream Marala.
199 159
158
157 ) m ( l
s 156 m
om
r 155 n f o i
at 154 v e l E 153
152
151 0 305 762 1220 1677 2134 2378 2561 2683 3049 3384 3567 3720 3963 4421 4878 5335 5793
Width of Channel (m)
Cros-ssection at 271068 m downstream Marala
156
154 )
m 152 ( l s m om
r 150 on f i at
ev 148 l E
146
144 0 305 610 915 1220 1524 1677 1829 1982 2195 2317 2591 2744 2896 3201 3506 3811 4116 4421 4726 5030 5335 5640 5945
Width of channel (m)
Cross-section at 289841 m downstream Marala.
200 APPENDIX III
CALCULATION OF FLOW DURATION CURVE FOR RIVER CHENAB AT MARALA
Discharge data in Discharge data in descending order of chronologic order magnitude Exceedence Rank Probability % Sr. No. (M) (P=100* DATE Discharge DATE Discharge M/(N+1)) (Month, Year) ((m3/s)) (Month, Year) ((m3/s))
1 Apr-1947 532.36 Jan-1950 4381.75 1 0.14 2 May-1947 1127.86 Aug-1996 3999.27 2 0.28 3 Jun-1947 1813.13 Aug-1976 3910.84 3 0.42 4 Jul-1947 2159.73 Aug-1948 3667.31 4 0.55 5 Aug-1947 2495.85 Jul-1959 3657.12 5 0.69 6 Sep-1947 2064.30 Jul-1988 3652.31 6 0.83 7 Oct-1947 557.28 Jul-1948 3645.79 7 0.97 8 Nov-1947 261.36 Aug-1973 3565.94 8 1.11 9 Dec-1947 244.37 Aug-1975 3536.21 9 1.25 10 Jan-1948 249.75 Aug-1957 3512.14 10 1.39 11 Feb-1948 380.01 Jul-1950 3509.02 11 1.53 12 Mar-1948 1426.60 Jul-1956 3497.13 12 1.66 13 Apr-1948 885.47 Jul-1978 3477.03 13 1.80 14 May-1948 1398.57 Jul-1994 3451.54 14 1.94 15 Jun-1948 1482.95 Jul-1995 3445.31 15 2.08 16 Jul-1948 3652.31 Aug-1955 3417.28 16 2.22 17 Aug-1948 3910.84 Jul-1975 3383.58 17 2.36 18 Sep-1948 1457.47 Aug-1983 3378.77 18 2.50 19 Oct-1948 537.17 Aug-1978 3355.26 19 2.64 20 Nov-1948 235.03 Jul-1958 3341.39 20 2.77 21 Dec-1948 205.58 Aug-1959 3301.74 21 2.91 22 Jan-1949 221.72 Aug-1997 3277.11 22 3.05 23 Feb-1949 667.99 Jul-1957 3269.75 23 3.19 24 Mar-1949 654.97 Sep-1950 3269.75 24 3.33 25 Apr-1949 829.68 Aug-1994 3200.37 25 3.47 26 May-1949 1577.81 Jul-1977 3171.77 26 3.61 27 Jun-1949 1531.09 Aug-1982 3059.64 27 3.74 28 Jul-1949 2556.16 Aug-1956 3048.59 28 3.88 29 Aug-1949 2368.99 Jul-1960 3016.03 29 4.02 30 Sep-1949 1323.25 Jul-1998 2986.86 30 4.16 31 Oct-1949 395.02 Jul-1961 2976.67 31 4.30 32 Nov-1949 194.54 Jul-1981 2958.26 32 4.44 33 Dec-1949 164.80 Aug-1950 2950.62 33 4.58 34 Jan-1950 5038.42 Jul-2005 2932.21 34 4.72 35 Feb-1950 656.67 Jul-1993 2931.36 35 4.85 36 Mar-1950 590.41 Aug-1995 2901.06 36 4.99 37 Apr-1950 705.09 Jul-1986 2856.89 37 5.13
201 38 May-1950 1454.35 Aug-1964 2856.04 38 5.27 39 Jun-1950 2445.44 Aug-1988 2845.84 39 5.41 40 Jul-1950 3512.14 Aug-1961 2819.51 40 5.55 41 Aug-1950 2958.26 Aug-1960 2813.28 41 5.69 42 Sep-1950 3269.75 Jul-1980 2796.85 42 5.83 43 Oct-1950 636.00 Sep-1959 2773.92 43 5.96 44 Nov-1950 310.64 Aug-1981 2773.07 44 6.10 45 Dec-1950 240.69 Aug-1986 2760.04 45 6.24 46 Jan-1951 219.74 Jul-1985 2750.42 46 6.38 47 Feb-1951 296.19 Aug-1984 2729.74 47 6.52 48 Mar-1951 376.61 Jul-1964 2726.63 48 6.66 49 Apr-1951 470.91 Jul-1989 2718.42 49 6.80 50 May-1951 1040.93 Aug-2006 2707.94 50 6.93 51 Jun-1951 1384.69 Aug-1977 2681.89 51 7.07 52 Jul-1951 1876.84 Jul-1973 2673.11 52 7.21 53 Aug-1951 2380.31 Jul-1968 2656.40 53 7.35 54 Sep-1951 928.79 Jul-1976 2613.36 54 7.49 55 Oct-1951 419.94 Jul-1967 2593.26 55 7.63 56 Nov-1951 220.02 Jul-1979 2580.80 56 7.77 57 Dec-1951 170.75 Jul-1991 2570.32 57 7.91 58 Jan-1952 211.24 Aug-1992 2558.43 58 8.04 59 Feb-1952 276.37 Aug-1966 2556.16 59 8.18 60 Mar-1952 547.93 Jul-1949 2545.97 60 8.32 61 Apr-1952 665.45 Jul-1969 2520.20 61 8.46 62 May-1952 1079.16 Jul-1996 2516.52 62 8.60 63 Jun-1952 1973.68 Jun-1996 2498.40 63 8.74 64 Jul-1952 2498.40 Jul-1952 2495.85 64 8.88 65 Aug-1952 2412.03 Aug-1947 2493.86 65 9.02 66 Sep-1952 1242.26 Aug-1985 2491.03 66 9.15 67 Oct-1952 349.43 Jul-1953 2472.06 67 9.29 68 Nov-1952 182.08 Jul-1965 2453.65 68 9.43 69 Dec-1952 152.06 Jul-1990 2445.44 69 9.57 70 Jan-1953 220.02 Jun-1950 2442.89 70 9.71 71 Feb-1953 219.46 Jun-1973 2436.66 71 9.85 72 Mar-1953 338.67 Aug-1958 2433.55 72 9.99 73 Apr-1953 643.64 Aug-1953 2419.39 73 10.12 74 May-1953 1066.98 Aug-1967 2413.73 74 10.26 75 Jun-1953 1832.95 Aug-1990 2412.03 75 10.40 76 Jul-1953 2491.03 Aug-1952 2407.78 76 10.54 77 Aug-1953 2433.55 Sep-1988 2380.31 77 10.68 78 Sep-1953 1518.63 Aug-1951 2376.92 78 10.82 79 Oct-1953 363.31 Jul-1997 2368.99 79 10.96 80 Nov-1953 230.22 Aug-1949 2357.94 80 11.10 81 Dec-1953 171.88 Aug-1968 2353.13 81 11.23 82 Jan-1954 325.64 Aug-1998 2348.32 82 11.37 83 Feb-1954 877.82 Jun-1991 2320.28 83 11.51 84 Mar-1954 618.72 Aug-1969 2319.43 84 11.65 85 Apr-1954 792.02 Jul-1987 2319.15 85 11.79
202 86 May-1954 1105.77 Jul-1983 2314.90 86 11.93 87 Jun-1954 1821.62 Jul-2006 2302.44 87 12.07 88 Jul-1954 2224.29 Jul-1982 2300.74 88 12.21 89 Aug-1954 2224.01 Aug-1963 2297.91 89 12.34 90 Sep-1954 2072.79 Aug-1980 2279.22 90 12.48 91 Oct-1954 685.27 Aug-1979 2276.67 91 12.62 92 Nov-1954 276.66 Jul-2003 2272.14 92 12.76 93 Dec-1954 197.65 Aug-2003 2269.88 93 12.90 94 Jan-1955 160.27 Aug-1971 2239.86 94 13.04 95 Feb-1955 178.96 Jun-1956 2229.39 95 13.18 96 Mar-1955 292.80 Jul-1992 2224.29 96 13.31 97 Apr-1955 482.24 Jul-1954 2224.01 97 13.45 98 May-1955 792.02 Aug-1954 2220.61 98 13.59 99 Jun-1955 1656.54 Aug-2005 2220.04 99 13.73 100 Jul-1955 2146.42 Sep-1961 2217.21 100 13.87 101 Aug-1955 3445.31 Aug-1987 2207.30 101 14.01 102 Sep-1955 1687.68 Jul-1984 2204.47 102 14.15 103 Oct-1955 1497.68 Jul-1963 2180.40 103 14.29 104 Nov-1955 402.95 Sep-1992 2168.22 104 14.42 105 Dec-1955 280.62 Jun-1984 2167.65 105 14.56 106 Jan-1956 259.95 Aug-1999 2159.73 106 14.70 107 Feb-1956 207.00 Jul-1947 2146.42 107 14.84 108 Mar-1956 618.72 Jul-1955 2131.41 108 14.98 109 Apr-1956 841.86 Jul-2001 2129.43 109 15.12 110 May-1956 1789.62 Jun-1978 2124.05 110 15.26 111 Jun-1956 2239.86 Aug-2002 2123.48 111 15.40 112 Jul-1956 3509.02 Aug-1989 2120.37 112 15.53 113 Aug-1956 3059.64 Jul-1972 2110.74 113 15.67 114 Sep-1956 1386.68 Jul-1999 2102.53 114 15.81 115 Oct-1956 853.75 Jul-2000 2093.46 115 15.95 116 Nov-1956 299.31 Jun-2003 2083.27 116 16.09 117 Dec-1956 194.25 Sep-1983 2072.79 117 16.23 118 Jan-1957 399.55 Sep-1954 2064.30 118 16.37 119 Feb-1957 490.73 Sep-1947 2055.80 119 16.50 120 Mar-1957 597.77 Jun-1966 2054.95 120 16.64 121 Apr-1957 1156.18 Sep-1958 2039.66 121 16.78 122 May-1957 1195.82 Aug-1970 2039.66 122 16.92 123 Jun-1957 1919.03 Jun-1995 2025.50 123 17.06 124 Jul-1957 3277.11 Sep-1966 2016.16 124 17.20 125 Aug-1957 3536.21 Jun-1980 2015.03 125 17.34 126 Sep-1957 1445.01 Aug-1962 1984.16 126 17.48 127 Oct-1957 626.65 Jun-1959 1977.37 127 17.61 128 Nov-1957 483.37 Aug-2001 1973.68 128 17.75 129 Dec-1957 538.30 Jun-1952 1973.68 129 17.89 130 Jan-1958 284.02 Jun-1975 1971.42 130 18.03 131 Feb-1958 279.20 Jul-1966 1959.53 131 18.17 132 Mar-1958 408.90 Jun-1990 1949.33 132 18.31 133 Apr-1958 807.03 Aug-1991 1936.02 133 18.45
203 134 May-1958 811.56 Sep-1975 1926.96 134 18.59 135 Jun-1958 1401.68 Aug-2000 1925.26 135 18.72 136 Jul-1958 3355.26 Jul-1974 1919.03 136 18.86 137 Aug-1958 2436.66 Jun-1957 1912.24 137 19.00 138 Sep-1958 2054.95 Jun-1968 1911.67 138 19.14 139 Oct-1958 968.44 Jul-1962 1900.63 139 19.28 140 Nov-1958 348.30 Aug-1974 1885.90 140 19.42 141 Dec-1958 496.11 Jun-1961 1883.35 141 19.56 142 Jan-1959 473.46 Jul-1971 1876.84 142 19.69 143 Feb-1959 967.59 Jul-1951 1873.16 143 19.83 144 Mar-1959 715.00 Aug-1965 1860.42 144 19.97 145 Apr-1959 909.82 Sep-1995 1846.54 145 20.11 146 May-1959 975.52 Jul-2002 1843.43 146 20.25 147 Jun-1959 1984.16 Jun-1963 1832.95 147 20.39 148 Jul-1959 3667.31 Jun-1953 1830.12 148 20.53 149 Aug-1959 3341.39 Jun-1988 1821.62 149 20.67 150 Sep-1959 2796.85 Jun-1954 1817.94 150 20.80 151 Oct-1959 931.06 Jun-1982 1817.94 151 20.94 152 Nov-1959 685.27 Jun-1989 1813.13 152 21.08 153 Dec-1959 242.11 Jun-1947 1798.12 153 21.22 154 Jan-1960 220.02 Jun-1994 1789.62 154 21.36 155 Feb-1960 220.31 May-1956 1769.80 155 21.50 156 Mar-1960 377.46 Jun-1969 1764.99 156 21.64 157 Apr-1960 400.12 Jun-1986 1757.34 157 21.78 158 May-1960 839.03 Aug-1972 1748.00 158 21.91 159 Jun-1960 1407.35 Sep-2006 1735.54 159 22.05 160 Jul-1960 3048.59 Jul-1970 1730.16 160 22.19 161 Aug-1960 2819.51 May-1990 1726.48 161 22.33 162 Sep-1960 1246.79 Jun-1987 1711.19 162 22.47 163 Oct-1960 423.34 Jun-1979 1700.14 163 22.61 164 Nov-1960 235.88 Aug-2004 1697.03 164 22.75 165 Dec-1960 185.76 Jun-2002 1695.33 165 22.88 166 Jan-1961 299.59 Jun-2005 1687.68 166 23.02 167 Feb-1961 362.74 Sep-1955 1678.34 167 23.16 168 Mar-1961 360.76 Jun-1998 1660.22 168 23.30 169 Apr-1961 663.46 Jun-1965 1656.54 169 23.44 170 May-1961 859.70 Jun-1955 1655.69 170 23.58 171 Jun-1961 1885.90 Sep-1973 1648.89 171 23.72 172 Jul-1961 2986.86 Jun-1972 1647.19 172 23.86 173 Aug-1961 2845.84 Jun-1992 1633.32 173 23.99 174 Sep-1961 2220.04 Aug-1993 1628.50 174 24.13 175 Oct-1961 594.09 May-1981 1617.74 175 24.27 176 Nov-1961 300.16 Jun-1993 1601.03 176 24.41 177 Dec-1961 247.49 Jul-2004 1578.10 177 24.55 178 Jan-1962 203.60 Jun-1981 1577.81 178 24.69 179 Feb-1962 278.07 May-1949 1577.25 179 24.83 180 Mar-1962 401.25 Jun-1964 1565.92 180 24.97 181 Apr-1962 649.31 Sep-1997 1563.09 181 25.10
204 182 May-1962 737.65 Jun-1967 1559.41 182 25.24 183 Jun-1962 1512.97 Sep-1976 1551.76 183 25.38 184 Jul-1962 1911.67 Jun-1976 1551.76 184 25.52 185 Aug-1962 2015.03 Jun-1985 1544.12 185 25.66 186 Sep-1962 1497.11 Jun-1971 1539.59 186 25.80 187 Oct-1962 470.34 Sep-1990 1533.92 187 25.94 188 Nov-1962 277.51 Sep-1977 1531.94 188 26.07 189 Dec-1962 255.13 Sep-1970 1531.09 189 26.21 190 Jan-1963 198.50 Jun-1949 1518.63 190 26.35 191 Feb-1963 192.27 Sep-1953 1512.97 191 26.49 192 Mar-1963 701.41 Jun-1962 1502.78 192 26.63 193 Apr-1963 729.73 Sep-1964 1497.68 193 26.77 194 May-1963 956.83 Oct-1955 1497.68 194 26.91 195 Jun-1963 1843.43 May-1973 1497.11 195 27.05 196 Jul-1963 2204.47 Sep-1962 1483.80 196 27.18 197 Aug-1963 2300.74 Sep-1996 1482.95 197 27.32 198 Sep-1963 997.60 Jun-1948 1460.30 198 27.46 199 Oct-1963 368.97 Sep-1994 1457.47 199 27.60 200 Nov-1963 259.67 Sep-1948 1454.35 200 27.74 201 Dec-1963 234.18 May-1950 1454.07 201 27.88 202 Jan-1964 425.32 May-1983 1453.22 202 28.02 203 Feb-1964 322.53 May-1978 1445.01 203 28.16 204 Mar-1964 413.14 Sep-1957 1429.15 204 28.29 205 Apr-1964 612.49 Apr-1983 1426.60 205 28.43 206 May-1964 816.37 Mar-1948 1415.84 206 28.57 207 Jun-1964 1577.25 May-2006 1411.03 207 28.71 208 Jul-1964 2729.74 Jun-1977 1407.35 208 28.85 209 Aug-1964 2856.89 Jun-1960 1403.67 209 28.99 210 Sep-1964 1502.78 Sep-1991 1402.25 210 29.13 211 Oct-1964 484.22 May-1982 1401.68 211 29.26 212 Nov-1964 261.36 Jun-1958 1398.57 212 29.40 213 Dec-1964 237.86 May-1948 1386.68 213 29.54 214 Jan-1965 292.51 Sep-1956 1384.69 214 29.68 215 Feb-1965 429.57 Jun-1951 1366.85 215 29.82 216 Mar-1965 388.79 Sep-1999 1343.07 216 29.96 217 Apr-1965 817.51 May-1998 1323.25 217 30.10 218 May-1965 1101.24 Sep-1949 1318.72 218 30.24 219 Jun-1965 1660.22 Jun-2001 1316.73 219 30.37 220 Jul-1965 2472.06 Sep-1987 1316.73 220 30.51 221 Aug-1965 1873.16 Jun-1997 1311.92 221 30.65 222 Sep-1965 787.21 Sep-1998 1311.07 222 30.79 223 Oct-1965 329.61 Apr-1991 1284.45 223 30.93 224 Nov-1965 231.35 May-1995 1281.90 224 31.07 225 Dec-1965 162.26 May-1991 1281.90 225 31.21 226 Jan-1966 141.30 Sep-2003 1280.49 226 31.35 227 Feb-1966 451.09 May-2003 1273.41 227 31.48 228 Mar-1966 553.03 Sep-1978 1273.41 228 31.62 229 Apr-1966 559.82 Jun-1999 1272.28 229 31.76
205 230 May-1966 888.02 Sep-1984 1256.42 230 31.90 231 Jun-1966 2055.80 Sep-1967 1248.77 231 32.04 232 Jul-1966 1971.42 Jun-1983 1246.79 232 32.18 233 Aug-1966 2558.43 Sep-1960 1242.26 233 32.32 234 Sep-1966 2025.50 Sep-1952 1241.13 234 32.45 235 Oct-1966 565.77 Jun-2000 1230.08 235 32.59 236 Nov-1966 282.32 May-1976 1225.27 236 32.73 237 Dec-1966 195.39 Jun-1970 1223.57 237 32.87 238 Jan-1967 174.43 May-1975 1222.44 238 33.01 239 Feb-1967 227.95 Sep-2005 1216.77 239 33.15 240 Mar-1967 602.30 Jun-2004 1209.98 240 33.29 241 Apr-1967 755.21 Jun-2006 1201.77 241 33.43 242 May-1967 762.29 May-1987 1195.82 242 33.56 243 Jun-1967 1563.09 May-1957 1194.69 243 33.70 244 Jul-1967 2613.36 May-1988 1184.78 244 33.84 245 Aug-1967 2419.39 May-2002 1169.49 245 33.98 246 Sep-1967 1256.42 Mar-2007 1156.18 246 34.12 247 Oct-1967 363.59 Apr-1957 1136.36 247 34.26 248 Nov-1967 222.85 May-2000 1128.99 248 34.40 249 Dec-1967 434.95 May-1992 1127.86 249 34.54 250 Jan-1968 402.95 May-1947 1127.29 250 34.67 251 Feb-1968 428.15 May-1993 1117.38 251 34.81 252 Mar-1968 581.06 Mar-1990 1112.57 252 34.95 253 Apr-1968 659.78 May-1968 1105.77 253 35.09 254 May-1968 1112.57 May-1954 1105.21 254 35.23 255 Jun-1968 1912.24 May-1996 1101.24 255 35.37 256 Jul-1968 2673.11 May-1965 1095.30 256 35.51 257 Aug-1968 2357.94 May-1999 1095.01 257 35.64 258 Sep-1968 1079.72 Apr-1998 1079.72 258 35.78 259 Oct-1968 355.66 Sep-1968 1079.72 259 35.92 260 Nov-1968 199.07 Sep-1993 1079.16 260 36.06 261 Dec-1968 168.77 May-1952 1066.98 261 36.20 262 Jan-1969 172.73 May-1953 1064.15 262 36.34 263 Feb-1969 306.39 May-1986 1063.86 263 36.48 264 Mar-1969 449.39 Sep-2000 1061.03 264 36.62 265 Apr-1969 628.63 Jun-1974 1053.39 265 36.75 266 May-1969 1030.45 Sep-1972 1047.72 266 36.89 267 Jun-1969 1769.80 Apr-1982 1045.74 267 37.03 268 Jul-1969 2545.97 Sep-2002 1043.48 268 37.17 269 Aug-1969 2320.28 May-1989 1042.06 269 37.31 270 Sep-1969 984.58 Apr-1981 1040.93 270 37.45 271 Oct-1969 348.30 May-1951 1030.45 271 37.59 272 Nov-1969 195.39 May-1969 1007.23 272 37.73 273 Dec-1969 146.11 May-2004 1006.10 273 37.86 274 Jan-1970 152.06 Sep-1986 999.87 274 38.00 275 Feb-1970 142.72 Mar-2005 997.60 275 38.14 276 Mar-1970 250.89 Sep-1963 990.24 276 38.28 277 Apr-1970 433.25 Sep-1985 985.71 277 38.42
206 278 May-1970 773.90 May-1984 984.58 278 38.56 279 Jun-1970 1225.27 Sep-1969 983.73 279 38.70 280 Jul-1970 1735.54 Apr-1979 975.52 280 38.83 281 Aug-1970 2039.66 May-1959 974.10 281 38.97 282 Sep-1970 1531.94 Sep-1979 968.44 282 39.11 283 Oct-1970 396.15 Oct-1958 967.59 283 39.25 284 Nov-1970 201.05 Feb-1959 967.59 284 39.39 285 Dec-1970 149.23 Sep-1980 961.92 285 39.53 286 Jan-1971 121.76 Sep-1989 956.83 286 39.67 287 Feb-1971 163.11 May-1963 956.54 287 39.81 288 Mar-1971 242.11 May-1994 946.63 288 39.94 289 Apr-1971 343.48 May-2005 942.10 289 40.08 290 May-1971 552.18 Mar-1983 937.29 290 40.22 291 Jun-1971 1544.12 Apr-1973 936.72 291 40.36 292 Jul-1971 1883.35 May-1979 931.06 292 40.50 293 Aug-1971 2269.88 Oct-1959 928.79 293 40.64 294 Sep-1971 781.54 Sep-1951 927.38 294 40.78 295 Oct-1971 310.92 Mar-1979 924.26 295 40.92 296 Nov-1971 172.73 Mar-1982 917.18 296 41.05 297 Dec-1971 133.09 Mar-1998 915.48 297 41.19 298 Jan-1972 130.26 Sep-1982 912.65 298 41.33 299 Feb-1972 305.54 Sep-1981 910.10 299 41.47 300 Mar-1972 363.59 Mar-1978 909.82 300 41.61 301 Apr-1972 445.42 Apr-1959 906.42 301 41.75 302 May-1972 818.92 Feb-2005 906.14 302 41.89 303 Jun-1972 1648.89 Sep-2004 903.59 303 42.02 304 Jul-1972 2120.37 Mar-1993 895.38 304 42.16 305 Aug-1972 1757.34 May-1980 893.68 305 42.30 306 Sep-1972 1053.39 Mar-1992 888.02 306 42.44 307 Oct-1972 296.19 May-1966 885.47 307 42.58 308 Nov-1972 197.37 Apr-1948 877.82 308 42.72 309 Dec-1972 208.13 Feb-1954 862.81 309 42.86 310 Jan-1973 356.79 Apr-1988 862.81 310 43.00 311 Feb-1973 420.22 Mar-2003 859.70 311 43.13 312 Mar-1973 743.60 May-1961 853.75 312 43.27 313 Apr-1973 937.29 Oct-1956 850.64 313 43.41 314 May-1973 1497.68 May-1985 849.79 314 43.55 315 Jun-1973 2442.89 May-2001 846.67 315 43.69 316 Jul-1973 2681.89 Apr-1976 841.86 316 43.83 317 Aug-1973 3645.79 Apr-1956 839.03 317 43.97 318 Sep-1973 1655.69 May-1960 834.50 318 44.11 319 Oct-1973 423.34 Apr-2003 829.68 319 44.24 320 Nov-1973 218.04 Apr-1949 829.12 320 44.38 321 Dec-1973 186.32 Mar-1988 826.85 321 44.52 322 Jan-1974 207.56 Apr-1990 824.87 322 44.66 323 Feb-1974 245.22 Apr-1986 822.04 323 44.80 324 Mar-1974 310.07 Sep-2001 819.21 324 44.94 325 Apr-1974 453.92 Apr-1975 818.92 325 45.08
207 326 May-1974 577.66 May-1972 817.51 326 45.21 327 Jun-1974 1061.03 Apr-1965 816.37 327 45.35 328 Jul-1974 1925.26 May-1964 811.56 328 45.49 329 Aug-1974 1900.63 May-1958 807.88 329 45.63 330 Sep-1974 787.21 Apr-2005 807.03 330 45.77 331 Oct-1974 244.37 Apr-1958 805.33 331 45.91 332 Nov-1974 128.28 Oct-1988 804.20 332 46.05 333 Dec-1974 191.99 Apr-1996 801.37 333 46.19 334 Jan-1975 193.97 Apr-1992 798.82 334 46.32 335 Feb-1975 404.65 Feb-2003 795.14 335 46.46 336 Mar-1975 644.21 Mar-1991 794.85 336 46.60 337 Apr-1975 819.21 Apr-1978 792.02 337 46.74 338 May-1975 1223.57 Apr-1954 792.02 338 46.88 339 Jun-1975 1973.68 May-1955 791.46 339 47.02 340 Jul-1975 3417.28 May-1997 787.21 340 47.16 341 Aug-1975 3565.94 Sep-1965 787.21 341 47.30 342 Sep-1975 1936.02 Sep-1974 781.54 342 47.43 343 Oct-1975 596.92 Sep-1971 777.86 343 47.57 344 Nov-1975 267.03 Apr-1997 773.90 344 47.71 345 Dec-1975 173.30 May-1970 762.29 345 47.85 346 Jan-1976 222.29 May-1967 762.29 346 47.99 347 Feb-1976 497.53 Mar-1981 755.21 347 48.13 348 Mar-1976 624.39 Apr-1967 748.13 348 48.27 349 Apr-1976 846.67 May-1977 743.60 349 48.40 350 May-1976 1230.08 Mar-1973 741.90 350 48.54 351 Jun-1976 1551.76 Apr-1995 738.22 351 48.68 352 Jul-1976 2656.40 Apr-1987 737.65 352 48.82 353 Aug-1976 3999.27 May-1962 730.29 353 48.96 354 Sep-1976 1559.41 Mar-1996 729.73 354 49.10 355 Oct-1976 479.97 Apr-1963 715.00 355 49.24 356 Nov-1976 227.38 Mar-1959 713.58 356 49.38 357 Dec-1976 183.78 Apr-1993 707.07 357 49.51 358 Jan-1977 369.82 Apr-1999 705.09 358 49.65 359 Feb-1977 249.47 Apr-1950 701.41 359 49.79 360 Mar-1977 244.94 Mar-1963 695.74 360 49.93 361 Apr-1977 463.55 Nov-1959 685.27 361 50.07 362 May-1977 748.13 Oct-1954 667.99 362 50.21 363 Jun-1977 1411.03 Feb-1949 665.45 363 50.35 364 Jul-1977 3200.37 Apr-1952 663.46 364 50.49 365 Aug-1977 2707.94 Apr-1961 659.78 365 50.62 366 Sep-1977 1533.92 Apr-1968 656.67 366 50.76 367 Oct-1977 494.13 Feb-1950 654.97 367 50.90 368 Nov-1977 255.70 Mar-1949 649.31 368 51.04 369 Dec-1977 210.39 Apr-1962 648.17 369 51.18 370 Jan-1978 191.42 Mar-1995 644.21 370 51.32 371 Feb-1978 263.06 Mar-1975 643.64 371 51.46 372 Mar-1978 910.10 Apr-1953 642.51 372 51.60 373 Apr-1978 794.85 Feb-1995 641.66 373 51.73
208 374 May-1978 1453.22 Feb-1981 636.00 374 51.87 375 Jun-1978 2129.43 Oct-1950 628.63 375 52.01 376 Jul-1978 3497.13 Apr-1969 626.65 376 52.15 377 Aug-1978 3378.77 Oct-1957 624.39 377 52.29 378 Sep-1978 1273.41 Mar-1976 618.72 378 52.43 379 Oct-1978 497.81 Mar-1954 618.72 379 52.57 380 Nov-1978 317.15 Mar-1956 618.72 380 52.70 381 Dec-1978 182.93 Oct-1998 612.49 381 52.84 382 Jan-1979 225.97 Apr-1964 611.93 382 52.98 383 Feb-1979 365.57 Feb-1991 609.94 383 53.12 384 Mar-1979 927.38 Feb-1998 603.15 384 53.26 385 Apr-1979 983.73 Oct-1996 602.30 385 53.40 386 May-1979 936.72 Mar-1967 597.77 386 53.54 387 Jun-1979 1711.19 Mar-1957 596.92 387 53.68 388 Jul-1979 2593.26 Oct-1975 594.09 388 53.81 389 Aug-1979 2279.22 Oct-1961 590.41 389 53.95 390 Sep-1979 974.10 Mar-1950 581.06 390 54.09 391 Oct-1979 341.50 Mar-1968 577.66 391 54.23 392 Nov-1979 235.88 May-1974 569.17 392 54.37 393 Dec-1979 190.86 Oct-1997 566.90 393 54.51 394 Jan-1980 225.40 Mar-1986 565.77 394 54.65 395 Feb-1980 356.51 Oct-1966 564.64 395 54.79 396 Mar-1980 506.02 Mar-1987 559.82 396 54.92 397 Apr-1980 554.16 Apr-1966 558.69 397 55.06 398 May-1980 895.38 Oct-1995 557.28 398 55.20 399 Jun-1980 2016.16 Oct-1947 554.16 399 55.34 400 Jul-1980 2813.28 Apr-1980 553.03 400 55.48 401 Aug-1980 2297.91 Mar-1966 553.03 401 55.62 402 Sep-1980 967.59 Apr-2006 552.18 402 55.76 403 Oct-1980 383.41 May-1971 547.93 403 55.89 404 Nov-1980 225.69 Mar-1952 544.53 404 56.03 405 Dec-1980 215.21 Apr-1984 543.68 405 56.17 406 Jan-1981 471.19 Apr-2000 542.55 406 56.31 407 Feb-1981 641.66 Oct-1985 538.30 407 56.45 408 Mar-1981 762.29 Dec-1957 537.17 408 56.59 409 Apr-1981 1042.06 Oct-1948 535.47 409 56.73 410 May-1981 1628.50 Oct-2006 532.36 410 56.87 411 Jun-1981 1578.10 Apr-1947 525.84 411 57.00 412 Jul-1981 2976.67 Apr-1989 520.46 412 57.14 413 Aug-1981 2773.92 Oct-1983 514.52 413 57.28 414 Sep-1981 912.65 Apr-2004 506.87 414 57.42 415 Oct-1981 361.04 Apr-2002 506.02 415 57.56 416 Nov-1981 245.51 Mar-1980 505.74 416 57.70 417 Dec-1981 193.69 Feb-1992 503.84 417 57.84 418 Jan-1982 204.73 Oct-1978 497.81 418 57.98 419 Feb-1982 237.58 Feb-1976 497.53 419 58.11 420 Mar-1982 924.26 Dec-1958 496.11 420 58.25 421 Apr-1982 1047.72 Oct-1977 494.13 421 58.39
209 422 May-1982 1402.25 Oct-1986 493.56 422 58.53 423 Jun-1982 1817.94 Feb-1957 490.73 423 58.67 424 Jul-1982 2302.44 Oct-1990 490.73 424 58.81 425 Aug-1982 3171.77 Dec-1990 489.32 425 58.95 426 Sep-1982 915.48 Oct-1987 487.33 426 59.08 427 Oct-1982 291.10 Dec-1986 485.35 427 59.22 428 Nov-1982 245.51 Oct-1964 484.22 428 59.36 429 Dec-1982 227.67 Nov-1957 483.37 429 59.50 430 Jan-1983 267.03 Apr-1955 482.24 430 59.64 431 Feb-1983 330.46 Oct-1976 479.97 431 59.78 432 Mar-1983 942.10 Jan-1959 473.46 432 59.92 433 Apr-1983 1429.15 Jan-1981 471.19 433 60.06 434 May-1983 1454.07 Apr-1951 470.91 434 60.19 435 Jun-1983 1248.77 Oct-1962 470.34 435 60.33 436 Jul-1983 2319.15 Apr-1994 466.38 436 60.47 437 Aug-1983 3383.58 Apr-1977 463.55 437 60.61 438 Sep-1983 2083.27 Oct-1992 455.05 438 60.75 439 Oct-1983 520.46 Apr-1974 453.92 439 60.89 440 Nov-1983 261.36 Feb-1966 451.09 440 61.03 441 Dec-1983 188.31 Jan-1991 450.24 441 61.17 442 Jan-1984 166.22 Mar-1969 449.39 442 61.30 443 Feb-1984 194.25 Apr-1972 445.42 443 61.44 444 Mar-1984 358.21 Feb-1990 443.16 444 61.58 445 Apr-1984 544.53 Dec-1967 434.95 445 61.72 446 May-1984 985.71 Apr-1970 433.25 446 61.86 447 Jun-1984 2168.22 Feb-1965 429.57 447 62.00 448 Jul-1984 2207.30 Oct-2005 429.00 448 62.14 449 Aug-1984 2750.42 Oct-1999 428.43 449 62.27 450 Sep-1984 1272.28 Feb-1968 428.15 450 62.41 451 Oct-1984 303.84 Jan-1964 425.32 451 62.55 452 Nov-1984 210.39 Oct-1960 423.34 452 62.69 453 Dec-1984 173.02 Oct-1973 423.34 453 62.83 454 Jan-1985 218.04 Oct-1994 421.64 454 62.97 455 Feb-1985 201.33 Feb-1973 420.22 455 63.11 456 Mar-1985 210.68 Oct-1951 419.94 456 63.25 457 Apr-1985 368.12 Feb-2004 419.09 457 63.38 458 May-1985 850.64 Oct-2003 418.24 458 63.52 459 Jun-1985 1551.76 Feb-1996 415.69 459 63.66 460 Jul-1985 2760.04 Nov-1997 413.43 460 63.80 461 Aug-1985 2493.86 Mar-1964 413.14 461 63.94 462 Sep-1985 990.24 Mar-1997 411.44 462 64.08 463 Oct-1985 542.55 Mar-1958 408.90 463 64.22 464 Nov-1985 213.23 Dec-2006 408.90 464 64.36 465 Dec-1985 365.00 Oct-1991 405.50 465 64.49 466 Jan-1986 255.70 Feb-1975 404.65 466 64.63 467 Feb-1986 331.59 Nov-1955 402.95 467 64.77 468 Mar-1986 566.90 Jan-1968 402.95 468 64.91 469 Apr-1986 824.87 Mar-1962 401.25 469 65.05
210 470 May-1986 1064.15 Mar-2006 400.40 470 65.19 471 Jun-1986 1764.99 Apr-1960 400.12 471 65.33 472 Jul-1986 2901.06 Jan-1957 399.55 472 65.46 473 Aug-1986 2773.07 Mar-2004 397.85 473 65.60 474 Sep-1986 1006.10 Oct-1970 396.15 474 65.74 475 Oct-1986 493.56 Oct-1989 395.59 475 65.88 476 Nov-1986 366.14 Oct-1949 395.02 476 66.02 477 Dec-1986 485.35 Mar-1999 394.17 477 66.16 478 Jan-1987 287.98 Feb-1987 390.49 478 66.30 479 Feb-1987 390.49 Mar-1965 388.79 479 66.44 480 Mar-1987 564.64 Oct-1980 383.41 480 66.57 481 Apr-1987 738.22 Feb-2000 382.84 481 66.71 482 May-1987 1201.77 Feb-1948 380.01 482 66.85 483 Jun-1987 1726.48 Mar-1989 379.45 483 66.99 484 Jul-1987 2319.43 Jan-1992 379.45 484 67.13 485 Aug-1987 2217.21 Oct-2004 377.75 485 67.27 486 Sep-1987 1316.73 Mar-1960 377.46 486 67.41 487 Oct-1987 487.33 Mar-1951 376.61 487 67.55 488 Nov-1987 285.15 Jan-1977 369.82 488 67.68 489 Dec-1987 201.33 Jan-2000 369.25 489 67.82 490 Jan-1988 211.53 Oct-1963 368.97 490 67.96 491 Feb-1988 219.17 Apr-1985 368.12 491 68.10 492 Mar-1988 829.12 Jan-1995 368.12 492 68.24 493 Apr-1988 862.81 Oct-2000 367.84 493 68.38 494 May-1988 1194.69 Oct-1993 366.99 494 68.52 495 Jun-1988 1830.12 Oct-2002 366.42 495 68.65 496 Jul-1988 3657.12 Nov-1986 366.14 496 68.79 497 Aug-1988 2856.04 Feb-1979 365.57 497 68.93 498 Sep-1988 2407.78 Dec-1985 365.00 498 69.07 499 Oct-1988 805.33 Dec-1994 364.15 499 69.21 500 Nov-1988 346.32 Oct-1967 363.59 500 69.35 501 Dec-1988 298.18 Mar-1972 363.59 501 69.49 502 Jan-1989 357.08 Oct-1953 363.31 502 69.63 503 Feb-1989 260.23 Feb-1961 362.74 503 69.76 504 Mar-1989 379.45 Jan-2005 362.74 504 69.90 505 Apr-1989 525.84 Oct-1981 361.04 505 70.04 506 May-1989 1043.48 Mar-1961 360.76 506 70.18 507 Jun-1989 1817.94 Mar-1984 358.21 507 70.32 508 Jul-1989 2726.63 Jan-1989 357.08 508 70.46 509 Aug-1989 2123.48 Jan-1973 356.79 509 70.60 510 Sep-1989 961.92 Feb-1980 356.51 510 70.74 511 Oct-1989 395.59 Oct-1968 355.66 511 70.87 512 Nov-1989 227.38 Mar-2000 353.39 512 71.01 513 Dec-1989 199.35 Jan-1999 352.26 513 71.15 514 Jan-1990 279.49 Oct-1952 349.43 514 71.29 515 Feb-1990 443.16 Apr-2001 349.15 515 71.43 516 Mar-1990 1117.38 Nov-1958 348.30 516 71.57 517 Apr-1990 826.85 Oct-1969 348.30 517 71.71
211 518 May-1990 1730.16 Nov-1988 346.32 518 71.84 519 Jun-1990 1959.53 Apr-1971 343.48 519 71.98 520 Jul-1990 2453.65 Oct-1979 341.50 520 72.12 521 Aug-1990 2413.73 Jan-2006 339.24 521 72.26 522 Sep-1990 1539.59 Mar-1953 338.67 522 72.40 523 Oct-1990 490.73 Feb-1986 331.59 523 72.54 524 Nov-1990 252.87 Feb-1983 330.46 524 72.68 525 Dec-1990 489.32 Oct-1965 329.61 525 72.82 526 Jan-1991 450.24 Oct-2001 329.61 526 72.95 527 Feb-1991 611.93 Jan-1954 325.64 527 73.09 528 Mar-1991 795.14 Feb-1964 322.53 528 73.23 529 Apr-1991 1311.07 Feb-1999 317.72 529 73.37 530 May-1991 1281.90 Nov-1978 317.15 530 73.51 531 Jun-1991 2348.32 Jan-2004 317.15 531 73.65 532 Jul-1991 2580.80 Nov-2006 316.30 532 73.79 533 Aug-1991 1949.33 Mar-1994 314.60 533 73.93 534 Sep-1991 1403.67 Mar-2002 312.62 534 74.06 535 Oct-1991 405.50 Dec-1997 312.33 535 74.20 536 Nov-1991 231.35 Nov-1998 312.33 536 74.34 537 Dec-1991 178.40 Nov-1999 312.33 537 74.48 538 Jan-1992 379.45 Oct-1971 310.92 538 74.62 539 Feb-1992 505.74 Nov-1950 310.64 539 74.76 540 Mar-1992 893.68 Mar-1974 310.07 540 74.90 541 Apr-1992 801.37 Jan-1996 307.52 541 75.03 542 May-1992 1128.99 Feb-1993 306.95 542 75.17 543 Jun-1992 1647.19 Feb-2007 306.67 543 75.31 544 Jul-1992 2229.39 Feb-1969 306.39 544 75.45 545 Aug-1992 2570.32 Feb-1972 305.54 545 75.59 546 Sep-1992 2180.40 Oct-1984 303.84 546 75.73 547 Oct-1992 455.05 Nov-1996 303.84 547 75.87 548 Nov-1992 267.03 Nov-1961 300.16 548 76.01 549 Dec-1992 203.03 Jan-1961 299.59 549 76.14 550 Jan-1993 289.68 Nov-1956 299.31 550 76.28 551 Feb-1993 306.95 Dec-1988 298.18 551 76.42 552 Mar-1993 903.59 Feb-1951 296.19 552 76.56 553 Apr-1993 713.58 Oct-1972 296.19 553 76.70 554 May-1993 1127.29 Mar-1955 292.80 554 76.84 555 Jun-1993 1617.74 Jan-1965 292.51 555 76.98 556 Jul-1993 2932.21 Oct-1982 291.10 556 77.12 557 Aug-1993 1633.32 Nov-1995 290.81 557 77.25 558 Sep-1993 1079.72 Jan-1993 289.68 558 77.39 559 Oct-1993 366.99 Jan-1987 287.98 559 77.53 560 Nov-1993 245.51 Nov-1987 285.15 560 77.67 561 Dec-1993 174.15 Jan-1958 284.02 561 77.81 562 Jan-1994 276.37 Nov-1966 282.32 562 77.95 563 Feb-1994 240.41 Dec-1955 280.62 563 78.09 564 Mar-1994 314.60 Jan-1990 279.49 564 78.22 565 Apr-1994 466.38 Feb-1958 279.20 565 78.36
212 566 May-1994 956.54 Feb-1962 278.07 566 78.50 567 Jun-1994 1798.12 Nov-1962 277.51 567 78.64 568 Jul-1994 3477.03 Nov-1954 276.66 568 78.78 569 Aug-1994 3269.75 Feb-1952 276.37 569 78.92 570 Sep-1994 1460.30 Jan-1994 276.37 570 79.06 571 Oct-1994 421.64 Feb-2006 272.12 571 79.20 572 Nov-1994 224.55 Jan-1998 267.88 572 79.33 573 Dec-1994 364.15 Nov-1975 267.03 573 79.47 574 Jan-1995 368.12 Jan-1983 267.03 574 79.61 575 Feb-1995 642.51 Nov-1992 267.03 575 79.75 576 Mar-1995 648.17 Jan-1997 264.48 576 79.89 577 Apr-1995 741.90 Feb-1978 263.06 577 80.03 578 May-1995 1284.45 Nov-1947 261.36 578 80.17 579 Jun-1995 2039.66 Nov-1964 261.36 579 80.31 580 Jul-1995 3451.54 Nov-1983 261.36 580 80.44 581 Aug-1995 2931.36 Feb-1989 260.23 581 80.58 582 Sep-1995 1860.42 Jan-1956 259.95 582 80.72 583 Oct-1995 558.69 Nov-1963 259.67 583 80.86 584 Nov-1995 290.81 Nov-2003 258.53 584 81.00 585 Dec-1995 252.59 Nov-2005 257.68 585 81.14 586 Jan-1996 307.52 Nov-1977 255.70 586 81.28 587 Feb-1996 415.69 Jan-1986 255.70 587 81.41 588 Mar-1996 730.29 Dec-1962 255.13 588 81.55 589 Apr-1996 804.20 Nov-1990 252.87 589 81.69 590 May-1996 1105.21 Dec-1995 252.59 590 81.83 591 Jun-1996 2516.52 Mar-1970 250.89 591 81.97 592 Jul-1996 2520.20 Jan-1948 249.75 592 82.11 593 Aug-1996 4381.75 Feb-1977 249.47 593 82.25 594 Sep-1996 1483.80 Dec-1998 247.77 594 82.39 595 Oct-1996 603.15 Dec-1961 247.49 595 82.52 596 Nov-1996 303.84 Nov-1981 245.51 596 82.66 597 Dec-1996 236.16 Nov-1982 245.51 597 82.80 598 Jan-1997 264.48 Nov-1993 245.51 598 82.94 599 Feb-1997 242.96 Feb-1974 245.22 599 83.08 600 Mar-1997 411.44 Mar-1977 244.94 600 83.22 601 Apr-1997 777.86 Dec-1947 244.37 601 83.36 602 May-1997 791.46 Oct-1974 244.37 602 83.50 603 Jun-1997 1316.73 Feb-1997 242.96 603 83.63 604 Jul-1997 2376.92 Dec-1959 242.11 604 83.77 605 Aug-1997 3301.74 Mar-1971 242.11 605 83.91 606 Sep-1997 1565.92 Dec-1950 240.69 606 84.05 607 Oct-1997 569.17 Feb-1994 240.41 607 84.19 608 Nov-1997 413.43 Dec-1964 237.86 608 84.33 609 Dec-1997 312.33 Feb-1982 237.58 609 84.47 610 Jan-1998 267.88 Dec-1996 236.16 610 84.60 611 Feb-1998 609.94 Nov-1960 235.88 611 84.74 612 Mar-1998 917.18 Nov-1979 235.88 612 84.88 613 Apr-1998 1095.01 Nov-1948 235.03 613 85.02
213 614 May-1998 1343.07 Nov-2000 235.03 614 85.16 615 Jun-1998 1678.34 Dec-1963 234.18 615 85.30 616 Jul-1998 3016.03 Dec-2003 233.05 616 85.44 617 Aug-1998 2353.13 Nov-1965 231.35 617 85.58 618 Sep-1998 1311.92 Nov-1991 231.35 618 85.71 619 Oct-1998 618.72 Nov-1953 230.22 619 85.85 620 Nov-1998 312.33 Nov-2001 229.37 620 85.99 621 Dec-1998 247.77 Feb-1967 227.95 621 86.13 622 Jan-1999 352.26 Dec-1982 227.67 622 86.27 623 Feb-1999 317.72 Nov-1976 227.38 623 86.41 624 Mar-1999 394.17 Nov-1989 227.38 624 86.55 625 Apr-1999 707.07 Jan-1979 225.97 625 86.69 626 May-1999 1095.30 Nov-1980 225.69 626 86.82 627 Jun-1999 1273.41 Jan-1980 225.40 627 86.96 628 Jul-1999 2110.74 Nov-1994 224.55 628 87.10 629 Aug-1999 2167.65 Nov-1967 222.85 629 87.24 630 Sep-1999 1366.85 Jan-1976 222.29 630 87.38 631 Oct-1999 428.43 Jan-1949 221.72 631 87.52 632 Nov-1999 312.33 Feb-1960 220.31 632 87.66 633 Dec-1999 191.42 Nov-1951 220.02 633 87.79 634 Jan-2000 369.25 Jan-1953 220.02 634 87.93 635 Feb-2000 382.84 Jan-1960 220.02 635 88.07 636 Mar-2000 353.39 Jan-1951 219.74 636 88.21 637 Apr-2000 543.68 Feb-1953 219.46 637 88.35 638 May-2000 1136.36 Feb-1988 219.17 638 88.49 639 Jun-2000 1241.13 Nov-2004 218.89 639 88.63 640 Jul-2000 2102.53 Nov-1973 218.04 640 88.77 641 Aug-2000 1926.96 Jan-1985 218.04 641 88.90 642 Sep-2000 1063.86 Jan-2007 216.91 642 89.04 643 Oct-2000 367.84 Dec-1980 215.21 643 89.18 644 Nov-2000 235.03 Nov-1985 213.23 644 89.32 645 Dec-2000 173.02 Jan-1988 211.53 645 89.46 646 Jan-2001 167.07 Jan-1952 211.24 646 89.60 647 Feb-2001 143.28 Mar-1985 210.68 647 89.74 648 Mar-2001 192.55 Dec-1977 210.39 648 89.88 649 Apr-2001 349.15 Nov-1984 210.39 649 90.01 650 May-2001 849.79 Nov-2002 208.70 650 90.15 651 Jun-2001 1318.72 Dec-1972 208.13 651 90.29 652 Jul-2001 2131.41 Jan-1974 207.56 652 90.43 653 Aug-2001 1977.37 Feb-1956 207.00 653 90.57 654 Sep-2001 822.04 Dec-1948 205.58 654 90.71 655 Oct-2001 329.61 Jan-1982 204.73 655 90.85 656 Nov-2001 229.37 Jan-1962 203.60 656 90.98 657 Dec-2001 194.25 Dec-1992 203.03 657 91.12 658 Jan-2002 123.74 Feb-1985 201.33 658 91.26 659 Feb-2002 168.77 Dec-1987 201.33 659 91.40 660 Mar-2002 312.62 Nov-1970 201.05 660 91.54 661 Apr-2002 506.87 Dec-1989 199.35 661 91.68
214 662 May-2002 1184.78 Nov-1968 199.07 662 91.82 663 Jun-2002 1697.03 Jan-1963 198.50 663 91.96 664 Jul-2002 1846.54 Dec-1954 197.65 664 92.09 665 Aug-2002 2124.05 Nov-1972 197.37 665 92.23 666 Sep-2002 1045.74 Dec-1966 195.39 666 92.37 667 Oct-2002 366.42 Nov-1969 195.39 667 92.51 668 Nov-2002 208.70 Nov-1949 194.54 668 92.65 669 Dec-2002 171.88 Dec-1956 194.25 669 92.79 670 Jan-2003 193.12 Feb-1984 194.25 670 92.93 671 Feb-2003 798.82 Dec-2001 194.25 671 93.07 672 Mar-2003 862.81 Jan-1975 193.97 672 93.20 673 Apr-2003 834.50 Dec-1981 193.69 673 93.34 674 May-2003 1280.49 Jan-2003 193.12 674 93.48 675 Jun-2003 2093.46 Mar-2001 192.55 675 93.62 676 Jul-2003 2276.67 Feb-1963 192.27 676 93.76 677 Aug-2003 2272.14 Dec-1974 191.99 677 93.90 678 Sep-2003 1281.90 Jan-1978 191.42 678 94.04 679 Oct-2003 418.24 Dec-1999 191.42 679 94.17 680 Nov-2003 258.53 Dec-1979 190.86 680 94.31 681 Dec-2003 233.05 Dec-1983 188.31 681 94.45 682 Jan-2004 317.15 Dec-1973 186.32 682 94.59 683 Feb-2004 419.09 Dec-2005 186.32 683 94.73 684 Mar-2004 397.85 Dec-1960 185.76 684 94.87 685 Apr-2004 514.52 Dec-1976 183.78 685 95.01 686 May-2004 1007.23 Dec-1978 182.93 686 95.15 687 Jun-2004 1216.77 Dec-2004 182.36 687 95.28 688 Jul-2004 1601.03 Nov-1952 182.08 688 95.42 689 Aug-2004 1700.14 Feb-1955 178.96 689 95.56 690 Sep-2004 906.14 Dec-1991 178.40 690 95.70 691 Oct-2004 377.75 Jan-1967 174.43 691 95.84 692 Nov-2004 218.89 Dec-1993 174.15 692 95.98 693 Dec-2004 182.36 Dec-1975 173.30 693 96.12 694 Jan-2005 362.74 Dec-1984 173.02 694 96.26 695 Feb-2005 906.42 Dec-2000 173.02 695 96.39 696 Mar-2005 999.87 Jan-1969 172.73 696 96.53 697 Apr-2005 807.88 Nov-1971 172.73 697 96.67 698 May-2005 946.63 Dec-1953 171.88 698 96.81 699 Jun-2005 1695.33 Dec-2002 171.88 699 96.95 700 Jul-2005 2950.62 Dec-1951 170.75 700 97.09 701 Aug-2005 2220.61 Dec-1968 168.77 701 97.23 702 Sep-2005 1222.44 Feb-2002 168.77 702 97.36 703 Oct-2005 429.00 Jan-2001 167.07 703 97.50 704 Nov-2005 257.68 Jan-1984 166.22 704 97.64 705 Dec-2005 186.32 Dec-1949 164.80 705 97.78 706 Jan-2006 339.24 Feb-1971 163.11 706 97.92 707 Feb-2006 272.12 Dec-1965 162.26 707 98.06 708 Mar-2006 400.40 Jan-1955 160.27 708 98.20 709 Apr-2006 553.03 Dec-1952 152.06 709 98.34
215 710 May-2006 1415.84 Jan-1970 152.06 710 98.47 711 Jun-2006 1209.98 Dec-1970 149.23 711 98.61 712 Jul-2006 2314.90 Dec-1969 146.11 712 98.75 713 Aug-2006 2718.42 Feb-2001 143.28 713 98.89 714 Sep-2006 1748.00 Feb-1970 142.72 714 99.03 715 Oct-2006 535.47 Jan-1966 141.30 715 99.17 716 Nov-2006 316.30 Dec-1971 133.09 716 99.31 717 Dec-2006 408.90 Jan-1972 130.26 717 99.45 718 Jan-2007 216.91 Nov-1974 128.28 718 99.58 719 Feb-2007 306.67 Jan-2002 123.74 719 99.72 720 Mar-2007 1169.49 Jan-1971 121.76 720 99.86
216 APPENDIX IV
CALCULATION OF LOW FLOW FREQUENCY CURVE (LFFC)
Return Minimum Minimum Period flows in _ _ _ Sr. annual Rank (Wiebull) 2 3 Year ascending 2 (z −z) (z z) No. flow (x) (M) (N+1)/M z = logx (z− z) − 3 order of ((m /s)) (year) magnitude 3 (m /s) 1 1947-48 8.63 4.3 1 61.00 0.9360 -0.0985 0.0097 -9.54E-04 2 1948-49 7.26 4.37 2 30.50 0.8609 -0.0234 0.0005 -1.28E-05 3 1949-50 5.82 4.53 3 20.33 0.7649 0.0726 0.0053 3.83E-04 4 1950-51 7.76 4.6 4 15.25 0.8899 -0.0523 0.0027 -1.43E-04 5 1951-52 6.03 4.99 5 12.20 0.7803 0.0572 0.0033 1.87E-04 6 1952-53 5.37 5.04 6 10.17 0.7300 0.1076 0.0116 1.25E-03 7 1953-54 6.07 5.06 7 8.71 0.7832 0.0544 0.0030 1.61E-04 8 1954-55 5.66 5.37 8 7.63 0.7528 0.0847 0.0072 6.08E-04 9 1955-56 7.31 5.66 9 6.78 0.8639 -0.0264 0.0007 -1.83E-05 10 1956-57 6.86 5.82 10 6.10 0.8363 0.0012 0.0000 1.86E-09 11 1957-58 9.86 5.87 11 5.55 0.9939 -0.1563 0.0244 -3.82E-03 12 1958-59 12.3 5.96 12 5.08 1.0899 -0.2524 0.0637 -1.61E-02 13 1959-60 7.77 6.03 13 4.69 0.8904 -0.0529 0.0028 -1.48E-04 14 1960-61 6.56 6.07 14 4.36 0.8169 0.0206 0.0004 8.80E-06 15 1961-62 7.19 6.07 15 4.07 0.8567 -0.0192 0.0004 -7.05E-06 16 1962-63 6.79 6.11 16 3.81 0.8319 0.0057 0.0000 1.84E-07 17 1963-64 8.27 6.12 17 3.59 0.9175 -0.0800 0.0064 -5.11E-04 18 1964-65 8.4 6.15 18 3.39 0.9243 -0.0867 0.0075 -6.52E-04 19 1965-66 4.99 6.16 19 3.21 0.6981 0.1395 0.0194 2.71E-03 20 1966-67 6.16 6.3 20 3.05 0.7896 0.0480 0.0023 1.10E-04 21 1967-68 7.87 6.44 21 2.90 0.8960 -0.0584 0.0034 -1.99E-04 22 1968-69 5.96 6.46 22 2.77 0.7752 0.0623 0.0039 2.42E-04 23 1969-70 5.04 6.49 23 2.65 0.7024 0.1351 0.0183 2.47E-03 24 1970-71 4.3 6.56 24 2.54 0.6335 0.2041 0.0417 8.50E-03 25 1971-72 4.6 6.58 25 2.44 0.6628 0.1748 0.0306 5.34E-03 26 1972-73 6.97 6.58 26 2.35 0.8432 -0.0057 0.0000 -1.83E-07 27 1973-74 6.58 6.74 27 2.26 0.8182 0.0193 0.0004 7.22E-06 28 1974-75 4.53 6.76 28 2.18 0.6561 0.1815 0.0329 5.97E-03 29 1975-76 6.12 6.76 29 2.10 0.7868 0.0508 0.0026 1.31E-04 30 1976-77 6.49 6.79 30 2.03 0.8122 0.0253 0.0006 1.62E-05 31 1977-78 6.76 6.84 31 1.97 0.8299 0.0076 0.0001 4.40E-07 32 1978-79 6.46 6.86 32 1.91 0.8102 0.0273 0.0007 2.04E-05 33 1979-80 6.74 6.97 33 1.85 0.8287 0.0089 0.0001 7.03E-07 34 1980-81 7.6 7.04 34 1.79 0.8808 -0.0433 0.0019 -8.10E-05 35 1981-82 6.84 7.11 35 1.74 0.8351 0.0025 0.0000 1.56E-08 36 1982-83 8.04 7.17 36 1.69 0.9053 -0.0677 0.0046 -3.10E-04 37 1983-84 5.87 7.19 37 1.65 0.7686 0.0689 0.0047 3.27E-04 38 1984-85 6.11 7.26 38 1.61 0.7860 0.0515 0.0027 1.37E-04 39 1985-86 7.53 7.31 39 1.56 0.8768 -0.0392 0.0015 -6.04E-05 40 1986-87 10.17 7.53 40 1.53 1.0073 -0.1698 0.0288 -4.89E-03 41 1987-88 7.11 7.6 41 1.49 0.8519 -0.0143 0.0002 -2.93E-06 42 1988-89 9.19 7.66 42 1.45 0.9633 -0.1258 0.0158 -1.99E-03 43 1989-90 7.04 7.76 43 1.42 0.8476 -0.0100 0.0001 -1.01E-06 44 1990-91 8.93 7.77 44 1.39 0.9509 -0.1133 0.0128 -1.45E-03 45 1991-92 6.3 7.87 45 1.36 0.7993 0.0382 0.0015 5.58E-05 46 1992-93 7.17 7.93 46 1.33 0.8555 -0.0180 0.0003 -5.80E-06
217 47 1993-94 6.15 8.04 47 1.30 0.7889 0.0487 0.0024 1.15E-04 48 1994-95 7.93 8.23 48 1.27 0.8993 -0.0617 0.0038 -2.35E-04 49 1995-96 8.92 8.27 49 1.24 0.9504 -0.1128 0.0127 -1.44E-03 50 1996-97 8.34 8.34 50 1.22 0.9212 -0.0836 0.0070 -5.85E-04 51 1997-98 9.46 8.4 51 1.20 0.9759 -0.1383 0.0191 -2.65E-03 52 1998-99 8.75 8.63 52 1.17 0.9420 -0.1045 0.0109 -1.14E-03 53 1999-00 6.76 8.75 53 1.15 0.8299 0.0076 0.0001 4.40E-07 54 2000-01 5.06 8.92 54 1.13 0.7042 0.1334 0.0178 2.37E-03 55 2001-02 4.34 8.93 55 1.11 0.6375 0.2001 0.0400 8.01E-03 56 2002-03 6.07 9.19 56 1.09 0.7832 0.0544 0.0030 1.61E-04 57 2003-04 8.23 9.46 57 1.07 0.9154 -0.0778 0.0061 -4.72E-04 58 2004-05 6.44 9.86 58 1.05 0.8089 0.0287 0.0008 2.36E-05 59 2005-06 6.58 10.17 59 1.03 0.8182 0.0193 0.0004 7.22E-06 60 2006-07 7.66 12.3 60 1.02 0.8842 -0.0467 0.0022 -1.02E-04
Summation 50.2502 0.5077 0.002
n = 60
n z _ ∑ i=1 3 mean (z) = = 0.837553 (m /s) n
n _ 2 ∑(z− z) standard deviation (S.D) = i=1 = 0.092656 n −1
n _ coefficient of skew (G) = [ (z − z)3 ] = 0.05 3 ∑ (n −1)(n − 2)(S.D) i−1
218 Calculation of Log Normal and Log Pearson III Distribution Functions
Log Log Log _ Log Return _ Exceedence Normal Normal Pearson III Parson III Period log Q = z+ K × S.D log Q = z+ K × S.D Probability K Discharge K Discharge (years) 3 3 (G=0) ((m /s)) (G=0.1) ((m /s))
1.001 0.999 3.09023 1.123881 13.30 2.94834 1.110734391 12.90 1.002 0.998 2.87816 1.104232 12.71 2.75706 1.093011151 12.39 1.005 0.995 2.57583 1.076219 11.92 2.48187 1.067513147 11.68 1.010 0.99 2.32635 1.053103 11.30 2.25258 1.046268052 11.12 1.020 0.98 2.05375 1.027845 10.66 1.99973 1.022839983 10.54 1.042 0.96 1.75069 0.999765 9.99 1.7158 0.996532165 9.92 1.111 0.9 1.28155 0.956296 9.04 1.27037 0.955260403 9.02 1.250 0.8 0.84162 0.915534 8.23 0.84611 0.915950168 8.24 1.429 0.7 0.5244 0.886142 7.69 0.53624 0.887238853 7.71 1.667 0.6 0.25335 0.861027 7.26 0.26882 0.862460786 7.29 2.000 0.5 0 0.837553 6.88 0.01662 0.839092943 6.90 2.500 0.4 -0.25335 0.814079 6.52 -0.23763 0.815535155 6.54 3.333 0.3 -0.5244 0.788964 6.15 -0.51207 0.790106642 6.17 5.000 0.2 -0.84162 0.759572 5.75 -0.83639 0.760056448 5.76 10.000 0.1 -1.28155 0.71881 5.23 -1.29178 0.717861832 5.22 25.000 0.04 -1.75069 0.675341 4.74 -1.78462 0.672197249 4.70 50.000 0.02 -2.05375 0.647261 4.44 -2.10697 0.642329588 4.39 100.000 0.01 -2.32635 0.622003 4.19 -2.39961 0.615214736 4.12 200.000 0.005 -2.57583 0.598887 3.97 -2.66965 0.59019391 3.89 500.000 0.002 -2.87816 0.570874 3.72 -2.99978 0.559605384 3.63 1000.000 0.001 -3.09023 0.551225 3.56 -3.23322 0.537975768 3.45
219 VITA
MUHAMMAD TOUSIF BHATTI
Candidate for the Degree of
Doctor of Philosophy
Dissertation: Strategic Analysis of Spatial and Temporal Water Quality of River Chenab and Its Management
Major Field: Water Resources Engineering
Biographical Information:
Personal Data: Born in Narowal, Pakistan, November 07, 1980, Son of Muhammad Idrees Bhatti.
Education: Received secondary school certificate from Government High School Sheikhupura in 1996, received higher secondary school certificate from Government College Sheikhupura in 1998, completed Bachelor of Science in Agricultural Engineering from University of Agriculture Faisalabad in 2002, received Master of Science in Water Resources Management from Centre of Excellence in Water Resources Engineering, University of Engineering and Technology, Lahore in 2005, Completed Masters of Business Administration with major in Information Technology from COMSATS Institute of Information Technology, Lahore in 2005, completed the requirements for the Doctor of Philosophy degree in Water Resources Engineering at Centre of Excellence in Water Resources Engineering, University of Engineering and Technology, Lahore in 2009.
Professional Experience: March 2005 to May 2009, Research Associate, Centre of Excellence in Water Resources Engineering, University of Engineering and Technology, Lahore.
Affiliations: Pakistan Engineering Council, and Canadian Water Resources Association.
220