**Meet the editors**

Dr Kostas Voudouris is an Assistant Professor at the Department of Geology; Laboratory of Engineering Geology & Hydrogeology, Aristotle University of Thessaloniki, Greece. He received his Bachelor of Science Degree in Geology and Mathematics from the University of Patras and a PhD degree in Hydrogeology from the Department of Geology, University of Patras.His primary re-

search interests are in Groundwater Quality, Field Hydrogeology, Aquifer Vulnerability to Pollution, Groundwater Management and Environmental Hydrogeology. He is a professional member of International Association of Hydrogeologists, Hellenic Committee of Hydrogeology, International Association of Mathematical Geosciences and European Water Resources Association. He has a number of papers published in the proceedings of international conferences and international scientific journals. He published a book on"Environmental Hydrogeology" in 2009 and has participated in a lot of European and National projects during the last years.

Dr Dimitra Voutsa is an Associate Professor at the Department of Chemistry, Laboratory of Environmental Pollution Control, Aristotle University of Thessaloniki. She received her B.S. in Chemistry and her Ph.D in Environmental Chemistry. Her research interests concern the fate of organic micropollutants and heavy metals in water cycle, drinking water quality and water treatment,

wastewater treatment processes and strategies for removal of micropollutants and chemical characterization of airborne particulate matter. She has published many articles in peer reviewed international scientific journals and chapters in books and participated in numerous national and European research projects.

Contents

**Preface IX** 

**Part 1 Statistical Analysis of Water Quality Data 1** 

Teferi Tsegaye and Mezemir Wagaw

Octavian Postolache, Pedro Silva Girão

Chapter 3 **Analysis of Water Quality Data for Scientists 65**  József Kovács, Péter Tanos, János Korponai, Ilona Kovácsné Székely, Károly Gondár,

Katalin Gondár-Sőregi and István Gábor Hatvani

**Trends of Water Quality Parameters 95** 

**Quality Monitoring in a Hydrological Basin –** 

Chapter 6 **Statistical Tools for Analyzing Water Quality Data 143** 

**and Multivariate Statistical Investigations of Groundwater Chemical Quality of Umm Rijam Aquifer (B4), North Yarmouk River Basin, Jordan 169** 

Mutewekil M. Obeidat, Muheeb Awawdeh, Fahmi Abu Al-Rub and Ahmad Al-Ajlouni

Chapter 5 **Combining Statistical Methodologies in Water** 

**Space and Time Approaches 121**  Marco Costa and A. Manuela Gonçalves

Liya Fu and You-GanWang

Chapter 7 **An Innovative Nitrate Pollution Index** 

Chapter 2 **Water Quality Monitoring and Associated** 

and José Miguel Dias Pereira

Chapter 4 **Detecting and Estimating** 

Janina Mozejko

Chapter 1 **Spatial Decision Support System (SDSS) for Stormwater Management and Water Quality Assessment 3**  Nally Kaunda-Bukenya, Wubishet Tadesse, Yujian Fu,

**Distributed Measurement Systems: An Overview 25** 

### Contents

#### **Preface XIII**


Contents VII

Chapter 19 **Determination and Speciation** 

Chapter 20 **Evaluation of Drinking Water Quality** 

Fabiola S. Sosa-Rodriguez

Chapter 23 **Water Quality Degradation Trends**

Shadrack Mulei Kithiia

Chapter 24 **Water Pollution of Oued Medjerda** 

**in Three Municipalities of Romania: The Influence of Municipal and Customer's**

Daniela Gheorghe and Aurel Ciupe

Amra Odobasic

Chapter 21 **Water Quality Monitoring** 

O.A.A. Eletta

**of Trace Heavy Metals in Natural Water by DPASV 429**

**Distribution Systems Concerning Trace Metals 457**  Gabriela Vasile, Liliana Cruceru, Cristina Dinu, Epsica Chiru,

**and Assessment in a Developing Country 481**

Chapter 22 **Assessing Water Quality in the Developing World: An Index for Mexico City 495**

**in Kenya over the Last Decade 509**

**in Algerian Souk Ahras Region 527**  A. Nait Merzoug and H. Merazig

Chapter 25 **Water Quality Issues in Developing Countries –** 

Adegbenro P. Daso and Oladele Osibanjo

Chapter 26 **Groundwater Quality Development in Area Suffering** 

Z. Hrkal, J. Burda, D. Fottová, M. Hrkalová,

Chapter 27 **Don't Know Responses in Water Quality Surveys 585** 

H. Nováková and E. Novotná

Zhihua Hu and Lois Wright Morton

**A Case Study of Ibadan Metropolis, Nigeria 541** 

**from Long Term Impact of Acid Atmospheric Deposition – The Role of Forest Cover in Czech Republic Case Study 561** 

	- **Part 2 Water Quality Monitoring Studies 267**

VI Contents

Chapter 8 **Monitoring and Modelling of Water Quality 189** 

Chapter 9 **Exploring Potentially Hazardous Areas** 

Chapter 10 **Assessment of Groundwater Quality** 

Nives Štambuk–Giljanović

Chapter 12 **Groundwater Quality Degradation** 

Chapter 14 **Temporal Water Quality Assessment** 

Chapter 15 **Mining and Water Pollution 347**  Hlanganani Tutu

Jachimko Jachimko Barbara

Chapter 17 **Relationship Between Water Quality** 

Aare Selberg and Malle Viik

**the Shallow Groundwater Quality** 

György Szabó, Tímea Vince and Éva Bessenyei

Chapter 18 **Study of the Factors Influencing**

Papiya Mandal and Sunil Kumar

**Part 2 Water Quality Monitoring Studies 267**

Chapter 11 **Sodium Levels in the Spring Water, Surface** 

Chapter 13 **Surface Water Quality Monitoring in Nigeria:** 

**of Langat River from 1995-2006 321**

Mahendran Shitan, Hafizan Juahir and Faridatul Azna Ahmad Shahabuddin

Chapter 16 **The Influence of Lignite Mining on Water Quality 373** 

**and Oil-Shale Mines in Northern Estonia 391**

**in Two Settlements with Different Characteristics 407**

**in Obrenovac Municipality, Serbia 283** 

Katarzyna Samborska, Rafal Ulanczyk and Katarzyna Korszun

**for Water Quality Using Dynamic Factor Analysis 227**  József Kovács, László Márkus, József Szalai, Márton Barcza, György Bernáth, Ilona Kovácsné Székely and Gábor Halupka

**in Industrial Areas of Delhi, India by Indexing Method 257** 

**and Groundwater in Dalmatia (Southern Croatia) 269** 

Nenad Zivkovic, Slavoljub Dragicevic, Ilija Brceski, Ratko Ristic, Ivan Novkovic, Slavoljub Jovanovic, Mrdjan Djokic and Sava Simic

**Situational Analysis and Future Management Strategy 301** A.M. Taiwo, O.O. Olujimi, O. Bamgbose and T.A. Arowolo

Zalina Mohd Ali, Noor Akma Ibrahim, Kerrie Mengersen,


Preface

scarcity and poor water quality.

around the world and water levels are dropping rapidly.

Water is a valuable and finite resource on Earth. Both water quantity and quality are becoming dominant issues in many countries. European Environment Agency notes that except in some northern countries that possess abundant water resources, water scarcity occurs in many countries, particularly in the Mediterranean, Middle East, Africa etc, confronted with a crucial combination of a severe lack of and high demand for water. The growth of world population, up to 9 billion by 2050, leading to increase demands of water, growing urbanization and high living standards, intensive agricultural activities and industrial demands as well as climate change with droughts and floods episodes are significant pressures for the available water resources. Consequently, many countries have significant problems concerning both severe water

Surface water and groundwater that are the main sources of fresh water for drinking purposes, irrigation and various other uses, represent as small fraction of water burden on earth. It is pointed out that only 30% of the freshwater (3% of the total volume of water) on Earth is groundwater. In many areas, water needs are mainly covered by groundwater abstracted from the aquifers via numerous wells and boreholes. As a result, a negative water balance is established in the aquifer systems

Point and non-point sources such as sewage effluents, wastewater discharges, agricultural runoff, industrial and mining activities, atmospheric deposition may seriously affect these water resources. As a consequence various pollutants such as pathogen microorganisms, nutrients, heavy metals, toxic elements, pesticides, pharmaceuticals and various other organic micropollutants may occur in water resulting in degradation of water quality. Another, severe problem, especially in coastal areas is the increase salinity of groundwater, due to seawater intrusion in

The access to good quality freshwater is a decisive factor for socio-economic development of the countries. Recently, the European Community through Water Directive 2000/60/EC, established the framework for actions in the field of water policy for the protection of inland surface waters, transitional waters, coastal waters and groundwater. This Directive aims at the protection and enhancement of the aquatic

coastal aquifers as a cause of high water demands and overexploitation.

### Preface

Water is a valuable and finite resource on Earth. Both water quantity and quality are becoming dominant issues in many countries. European Environment Agency notes that except in some northern countries that possess abundant water resources, water scarcity occurs in many countries, particularly in the Mediterranean, Middle East, Africa etc, confronted with a crucial combination of a severe lack of and high demand for water. The growth of world population, up to 9 billion by 2050, leading to increase demands of water, growing urbanization and high living standards, intensive agricultural activities and industrial demands as well as climate change with droughts and floods episodes are significant pressures for the available water resources. Consequently, many countries have significant problems concerning both severe water scarcity and poor water quality.

Surface water and groundwater that are the main sources of fresh water for drinking purposes, irrigation and various other uses, represent as small fraction of water burden on earth. It is pointed out that only 30% of the freshwater (3% of the total volume of water) on Earth is groundwater. In many areas, water needs are mainly covered by groundwater abstracted from the aquifers via numerous wells and boreholes. As a result, a negative water balance is established in the aquifer systems around the world and water levels are dropping rapidly.

Point and non-point sources such as sewage effluents, wastewater discharges, agricultural runoff, industrial and mining activities, atmospheric deposition may seriously affect these water resources. As a consequence various pollutants such as pathogen microorganisms, nutrients, heavy metals, toxic elements, pesticides, pharmaceuticals and various other organic micropollutants may occur in water resulting in degradation of water quality. Another, severe problem, especially in coastal areas is the increase salinity of groundwater, due to seawater intrusion in coastal aquifers as a cause of high water demands and overexploitation.

The access to good quality freshwater is a decisive factor for socio-economic development of the countries. Recently, the European Community through Water Directive 2000/60/EC, established the framework for actions in the field of water policy for the protection of inland surface waters, transitional waters, coastal waters and groundwater. This Directive aims at the protection and enhancement of the aquatic

#### XIV Preface

ecosystems, promotion of sustainable water use based on a long-term protection of available water resources, progressive reduction or cessation of discharges of hazardous substances into aquatic environment and mitigation the effects of floods and droughts. These actions contribute to the provision of sufficient supply of good quality surface water and groundwater as needed for sustainable as well as to balanced and equitable water use.

Preface XI

**List of reviewers** 

Thessaloniki

Thessaloniki

Ioannina

Thessaloniki

Thessaloniki

Thessaloniki

Thessaloniki, Greece

Thessaloniki, Greece

**Bobori Dimitra**, Lab. of Ichthyology, School of Biology, Aristotle University of

**Genitsaris Savvas**, Dep. of Botany, School of Biology, Aristotle University of

**Karayanni Hera**, Dep. of Biological Applications and Technology, University of

**Katsiapi Maria,** Dep. of Botany, School of Biology, Aristotle University of Thessaloniki **Kormas Kostas**, Dep. of Ichthyology and Aquatic Environment, University of Thessaly

**Michaloudi Evagelia**, Lab. of Ichthyology, School of Biology, Aristotle University of

**Moustaka-Gouni Maria**, Dept. of Botany, School of Biology, Aristotle University of

**Theodosiou Nikolaos,** Assistant Professor, Civil Engineering, Aristotle University of

**Voudouris Kostas**, Ass. Professor of Hydrogeology, Aristotle University of

**Voutsa Dimitra**, Assistant Professor, Dep. of Chemistry, Aristotle University of

**Polemio Maurizio,** Istituto di Ricerca per la Protezione Idrogeologica, Bari, Italy

**Georgiou Pantazis,** School of Agriculture, Aristotle University of Thessaloniki

**Kaklis Akis,** Dr. of Hydrogeology, Aristotle University of Thessaloniki

**Lazaridou Maria**, Professor of Biology, Aristotle University of Thessaloniki **Mattas Christos,** Dr. of Hydrogeology, Aristotle University of Thessaloniki **Melfos Basil**, Lecturer of Geology, Aristotle University of Thessaloniki

This book entitled "**Water Quality Monitoring and Assessment**" attempts to covers the main fields of water quality issues presenting case studies in various countries concerning the physicochemical characteristics of surface and groundwaters and possible pollution sources as well as methods and tools for the evaluation of water quality status. Particularly, this book is divided **into** two sections:

#### **1) Statistical analysis of water quality data**

The first ten chapters focus on the evaluation of water quality data by employing conventional hydrochemical techniques and statistical analysis (e.g. cluster, factor and trend analysis, risk analysis and decision support systems).

#### **2) Water quality monitoring studies**

This section includes seventeen chapters related to the water quality and the assessment of water pollution. These chapters represent case studies from different countries of the world regarding the quality of surface and groundwater.

We would like to express our thanks to the authors who contributed to this volume, to the reviewers for their valuable assistance, as well as to the organizers and the staff of the INTECH Open Access Publisher, especially **Marija Radja**, for their efforts to publish this book.

> **Dr. Kostas Voudouris**  Laboratory of Engineering Geology & Hydrogeology, Department of Geology, Aristotle University of Thessaloniki, Greece

> > **Dr. Dimitra Voutsa**  Department of Chemistry, Laboratory of Environmental Pollution Control, Aristotle University of Thessaloniki, Greece

#### **List of reviewers**

X Preface

balanced and equitable water use.

ecosystems, promotion of sustainable water use based on a long-term protection of available water resources, progressive reduction or cessation of discharges of hazardous substances into aquatic environment and mitigation the effects of floods and droughts. These actions contribute to the provision of sufficient supply of good quality surface water and groundwater as needed for sustainable as well as to

This book entitled "**Water Quality Monitoring and Assessment**" attempts to covers the main fields of water quality issues presenting case studies in various countries concerning the physicochemical characteristics of surface and groundwaters and possible pollution sources as well as methods and tools for the evaluation of water

The first ten chapters focus on the evaluation of water quality data by employing conventional hydrochemical techniques and statistical analysis (e.g. cluster, factor and

This section includes seventeen chapters related to the water quality and the assessment of water pollution. These chapters represent case studies from different

We would like to express our thanks to the authors who contributed to this volume, to the reviewers for their valuable assistance, as well as to the organizers and the staff of the INTECH Open Access Publisher, especially **Marija Radja**, for their efforts to

**Dr. Kostas Voudouris**

Department of Geology,

**Dr. Dimitra Voutsa**  Department of Chemistry,

Greece

Greece

Aristotle University of Thessaloniki,

Aristotle University of Thessaloniki,

Laboratory of Engineering Geology & Hydrogeology,

Laboratory of Environmental Pollution Control,

countries of the world regarding the quality of surface and groundwater.

quality status. Particularly, this book is divided **into** two sections:

**1) Statistical analysis of water quality data**

**2) Water quality monitoring studies** 

publish this book.

trend analysis, risk analysis and decision support systems).

**Bobori Dimitra**, Lab. of Ichthyology, School of Biology, Aristotle University of Thessaloniki

**Genitsaris Savvas**, Dep. of Botany, School of Biology, Aristotle University of Thessaloniki

**Georgiou Pantazis,** School of Agriculture, Aristotle University of Thessaloniki

**Kaklis Akis,** Dr. of Hydrogeology, Aristotle University of Thessaloniki

**Karayanni Hera**, Dep. of Biological Applications and Technology, University of Ioannina

**Katsiapi Maria,** Dep. of Botany, School of Biology, Aristotle University of Thessaloniki **Kormas Kostas**, Dep. of Ichthyology and Aquatic Environment, University of Thessaly **Lazaridou Maria**, Professor of Biology, Aristotle University of Thessaloniki

**Mattas Christos,** Dr. of Hydrogeology, Aristotle University of Thessaloniki

**Melfos Basil**, Lecturer of Geology, Aristotle University of Thessaloniki

**Michaloudi Evagelia**, Lab. of Ichthyology, School of Biology, Aristotle University of Thessaloniki

**Moustaka-Gouni Maria**, Dept. of Botany, School of Biology, Aristotle University of Thessaloniki

**Polemio Maurizio,** Istituto di Ricerca per la Protezione Idrogeologica, Bari, Italy

**Theodosiou Nikolaos,** Assistant Professor, Civil Engineering, Aristotle University of Thessaloniki

**Voudouris Kostas**, Ass. Professor of Hydrogeology, Aristotle University of Thessaloniki, Greece

**Voutsa Dimitra**, Assistant Professor, Dep. of Chemistry, Aristotle University of Thessaloniki, Greece

**Part 1** 

**Statistical Analysis of Water Quality Data** 

## **Part 1**

## **Statistical Analysis of Water Quality Data**

**1** 

*USA* 

**Spatial Decision Support System (SDSS)** 

**for Stormwater Management and** 

Nally Kaunda-Bukenya, Wubishet Tadesse, Yujian Fu,

*Alabama A&M University & City of Huntsville, AL, Planning Division* 

Land use policy in the United States is a predominantly local issue (Giannotti & Arnold, 2002). The challenge is that land use policies and decisions are made by elected and appointed municipal officials (Stocker et al., 1999) whose training may not necessarily be in environmental management. Because of the critical importance of their work, and because they deal with land-use planning and regulation on a daily basis, local officials need decision tools that can allow them to place case-by-case land-use decisions within the broader context of the watershed. These land use managers need tools for assisting them to evaluate environmental impacts of their land-use decisions, visualize alternative scenarios, and educate their constituency (Arnold, 2000). Historically, decision makers have indicated that inaccessibility of required geographic data and difficulties in synthesizing various recommendations are primary obstacles to spatial problem solving (Ascough et al., 2002). Indeed studies have shown that the ability to produce meaningful solutions can be improved if these obstacles are lessened or removed through an integrated systems approach, such as a Spatial Decision Support System (SDSS). As Ascough et al. (2002) have observed, a SDSS makes a positive contribution to decision-makers' task if it enables them to reach: (i) a more accurate solution, (ii) a faster solution to a given problem, or (iii) both of

The driving force for developing this SDSS is the limited use of Geographic Information Systems (GIS) for environmental planning in municipalities. This limitation is due to the fact that, even though GIS software is available in most municipal land management data centers, it is too complex for policy makers and environmental officials to use "out of the box" without acquiring expertise in GIS. Thus, there is a need to develop custom tools that are less intimidating to non-GIS audience or users, but robust enough to perform complex geoprocessing tasks and hydrological models in the background. The goal of this Chapter is to develop an adaptive and customer-driven environmental SDSS to assist municipal officials fulfill environmental legislations and minimize the impact of pollution resulting from urban development. There are two specific objectives that we will address. The first objective is to develop a front-end graphical user interface (GUI) that is robust

**1. Introduction** 

these.

**Water Quality Assessment** 

Teferi Tsegaye and Mezemir Wagaw

### **Spatial Decision Support System (SDSS) for Stormwater Management and Water Quality Assessment**

Nally Kaunda-Bukenya, Wubishet Tadesse, Yujian Fu, Teferi Tsegaye and Mezemir Wagaw *Alabama A&M University & City of Huntsville, AL, Planning Division USA* 

#### **1. Introduction**

Land use policy in the United States is a predominantly local issue (Giannotti & Arnold, 2002). The challenge is that land use policies and decisions are made by elected and appointed municipal officials (Stocker et al., 1999) whose training may not necessarily be in environmental management. Because of the critical importance of their work, and because they deal with land-use planning and regulation on a daily basis, local officials need decision tools that can allow them to place case-by-case land-use decisions within the broader context of the watershed. These land use managers need tools for assisting them to evaluate environmental impacts of their land-use decisions, visualize alternative scenarios, and educate their constituency (Arnold, 2000). Historically, decision makers have indicated that inaccessibility of required geographic data and difficulties in synthesizing various recommendations are primary obstacles to spatial problem solving (Ascough et al., 2002). Indeed studies have shown that the ability to produce meaningful solutions can be improved if these obstacles are lessened or removed through an integrated systems approach, such as a Spatial Decision Support System (SDSS). As Ascough et al. (2002) have observed, a SDSS makes a positive contribution to decision-makers' task if it enables them to reach: (i) a more accurate solution, (ii) a faster solution to a given problem, or (iii) both of these.

The driving force for developing this SDSS is the limited use of Geographic Information Systems (GIS) for environmental planning in municipalities. This limitation is due to the fact that, even though GIS software is available in most municipal land management data centers, it is too complex for policy makers and environmental officials to use "out of the box" without acquiring expertise in GIS. Thus, there is a need to develop custom tools that are less intimidating to non-GIS audience or users, but robust enough to perform complex geoprocessing tasks and hydrological models in the background. The goal of this Chapter is to develop an adaptive and customer-driven environmental SDSS to assist municipal officials fulfill environmental legislations and minimize the impact of pollution resulting from urban development. There are two specific objectives that we will address. The first objective is to develop a front-end graphical user interface (GUI) that is robust

Spatial Decision Support System (SDSS)

addresses this deficiency.

2006).

simultaneously fulfill environmental legislation.

for Stormwater Management and Water Quality Assessment 5

and Nonpoint Sources (BASINS) interface called WinHSPF. A simplified windows interface called Latis was developed by the authors as a simpler replacement of BASINS. (Wilkerson et al., 2010) The Latis model involves using specific rain event to model and compare scenarios under pre-development, and "as-built" with BMP options, and even worst case scenario (100 percent impervious). Later on, Latis, was further improved into Latis-LIDIA to estimate runoff based on pre- and post-developed site conditions using the widely-used Soil Conservation Service (SCS) runoff curve number (CN) method. The first step in the model requires user input of project information, site dimensions, and precipitation data. Precipitation data are automatically generated by selecting state and county, or manually entered by user-defined values. The precipitation database is tailored for sites within Alabama, Louisiana, and Mississippi (Wilkerson et al., 2010). The user then characterizes land use and land cover for each respective hydrologic soil group (HSG), cover type, and size. The model then generates runoff coefficients. Although this is significant contribution to the existing body of knowledge, the authors admit that the model needs further development to accommodate pollutant loading computation. This current research

In another instance, the Decision Evaluation in Complex Risk Network Systems (DECERNS) is a similar SDSS tool which focuses on land use planning and management (DECERNS - Team, 2006). DECERNS provides spatial data visualization for vector and raster models. It is used in the development of alternatives and criteria specification; implementation of the basic and advanced multi-criteria decision; and generation of various reports, including text descriptions, tables, diagrams, and maps (DECERNS -Team, 2006). It is a powerful commercial SDSS that caters for a larger community of state and regional officials, educators, and researchers. However, local governments who make critical land-use decisions are not direct beneficiaries because of the costs associated with this commercial software and also because this system lacks specific direct benefits to municipalities such as environmental legislation compliance tools. Thus, justifying the need to develop a municipal SDSS targeted for local communities to visualize the impacts of their decisions and

In another example, Rodman and Jackson (2006) used Python programming and the ArcGIS geoprocessor to develop a standalone spatial application, for the US Army Corps of Engineers (USACE), which performs data mining in geographic datasets. Python was selected as the language of choice because it is a powerful open source cross-platform programming language, that can run on Windows, Mac, or Unix and has a wealth of available code and tools that connect to databases for developing graphical user interfaces (GUI). The authors' goal was to create an application that relies on ArcGIS for handling spatial data formats, geographic coordinate system transformation, mapping, and geoprocessing. The resulting data mining application known as *Aspect*, is used to discover association rules that describe spatial relationships between geographic features. This allows decision makers to have tools that combine knowledge of terrain, travel routes, structure, land use/cover to improve situational awareness (Rodman & Jackson,

In this research, several implementation options were examined to determine the best approach to develop the graphical user interface. The option selected would need to take

enough to quantify and geolocate pollution hot-spots in an urban area, but simple enough for use by land use decision-makers whose expertise is neither GIS nor water quality assessment. The second objective is to demonstrate the use of the SDSS for generating environmental compliance reports and for assessing pre-development and postdevelopment conditions of a land use change. The rest of the chapter is organized into four sections. First, a review of previous studies is presented in section 2, the methods for developing the SDSS components in section 3. and the last two sections are the results and summary.

#### **2. Literature review**

A spatial decision support system (SDSS) is an interactive, computer-based system designed to support a user or group of users in achieving a higher effectiveness of decision making while solving a spatial problem (Sprague, 1982). According to Sprague (1982), a SDSS has three primary components: a geographic database management system for handling geographic data; a number of potential models that can be used to forecast the possible outcomes of decisions; and a user interface to provide interaction of the user to model scenarios. Similarly, Armstrong and Densham (1990) suggest that five key modules are needed in a SDSS: (i) a database management system (DBMS), (ii) analysis procedures in a model base management system (MBMS), (iii) a display generator, (iv) a report generator, and (v) a user interface.

Purdue Research Foundation (2010) developed a model called Long Term Hydrologic Impact Analysis (L-THIA). This model takes land use, soil, and long-term precipitation data as input and computes changes in recharge, runoff, and nonpoint source pollution resulting from past or proposed development as an output. The L-THIA modeling program is available in three forms as an online spreadsheet, as Avenue scripts that run as an extension of ArcView 3.x, or as an interactive mapping application developed using Java programming (Purdue Research Foundation, 2010). Although L-THIA was originally developed for municipal planners, its main focus seems to be non-point source (NPS) pollution. A similar tool is needed for addressing point source pollution in addition to NPS pollution, and for assisting municipalities with tools for environmental legislation compliance. The current research fills this gap by providing the ability to quantify point source pollution and use the newly developed user interface to generate reports for environmental compliance. The idea is that, having an environmental compliance tool that produces qualitative pollution hotspot maps or charts in addition to quantitative outputs (such as tabular data), enables decision-makers to track the environmental status of their watersheds and monitor the long-term effect of land use on the environment.

Wilkerson et al. (2010) developed a SDSS that allows users to balance watershed protection with smart growth/low-impact site development strategies. The SDSS was developed to calculate: time-varying runoff and water quality as a function of rainfall, site characteristics, and BMPs for development sites within the Southeastern U.S; and BMP cost, and compare various scenarios for effectiveness and cost. The authors used an existing Hydrological Simulation Program-FORTRAN (HSPF) for computing movement of water through a complete hydrologic cycle—rainfall, interception, evapo-transpiration, runoff, infiltration, and flow through the ground. HSPF runs on Better Assessment Science Integrating point

enough to quantify and geolocate pollution hot-spots in an urban area, but simple enough for use by land use decision-makers whose expertise is neither GIS nor water quality assessment. The second objective is to demonstrate the use of the SDSS for generating environmental compliance reports and for assessing pre-development and postdevelopment conditions of a land use change. The rest of the chapter is organized into four sections. First, a review of previous studies is presented in section 2, the methods for developing the SDSS components in section 3. and the last two sections are the results and

A spatial decision support system (SDSS) is an interactive, computer-based system designed to support a user or group of users in achieving a higher effectiveness of decision making while solving a spatial problem (Sprague, 1982). According to Sprague (1982), a SDSS has three primary components: a geographic database management system for handling geographic data; a number of potential models that can be used to forecast the possible outcomes of decisions; and a user interface to provide interaction of the user to model scenarios. Similarly, Armstrong and Densham (1990) suggest that five key modules are needed in a SDSS: (i) a database management system (DBMS), (ii) analysis procedures in a model base management system (MBMS), (iii) a display generator, (iv) a report generator,

Purdue Research Foundation (2010) developed a model called Long Term Hydrologic Impact Analysis (L-THIA). This model takes land use, soil, and long-term precipitation data as input and computes changes in recharge, runoff, and nonpoint source pollution resulting from past or proposed development as an output. The L-THIA modeling program is available in three forms as an online spreadsheet, as Avenue scripts that run as an extension of ArcView 3.x, or as an interactive mapping application developed using Java programming (Purdue Research Foundation, 2010). Although L-THIA was originally developed for municipal planners, its main focus seems to be non-point source (NPS) pollution. A similar tool is needed for addressing point source pollution in addition to NPS pollution, and for assisting municipalities with tools for environmental legislation compliance. The current research fills this gap by providing the ability to quantify point source pollution and use the newly developed user interface to generate reports for environmental compliance. The idea is that, having an environmental compliance tool that produces qualitative pollution hotspot maps or charts in addition to quantitative outputs (such as tabular data), enables decision-makers to track the environmental status of their

watersheds and monitor the long-term effect of land use on the environment.

Wilkerson et al. (2010) developed a SDSS that allows users to balance watershed protection with smart growth/low-impact site development strategies. The SDSS was developed to calculate: time-varying runoff and water quality as a function of rainfall, site characteristics, and BMPs for development sites within the Southeastern U.S; and BMP cost, and compare various scenarios for effectiveness and cost. The authors used an existing Hydrological Simulation Program-FORTRAN (HSPF) for computing movement of water through a complete hydrologic cycle—rainfall, interception, evapo-transpiration, runoff, infiltration, and flow through the ground. HSPF runs on Better Assessment Science Integrating point

summary.

**2. Literature review** 

and (v) a user interface.

and Nonpoint Sources (BASINS) interface called WinHSPF. A simplified windows interface called Latis was developed by the authors as a simpler replacement of BASINS. (Wilkerson et al., 2010) The Latis model involves using specific rain event to model and compare scenarios under pre-development, and "as-built" with BMP options, and even worst case scenario (100 percent impervious). Later on, Latis, was further improved into Latis-LIDIA to estimate runoff based on pre- and post-developed site conditions using the widely-used Soil Conservation Service (SCS) runoff curve number (CN) method. The first step in the model requires user input of project information, site dimensions, and precipitation data. Precipitation data are automatically generated by selecting state and county, or manually entered by user-defined values. The precipitation database is tailored for sites within Alabama, Louisiana, and Mississippi (Wilkerson et al., 2010). The user then characterizes land use and land cover for each respective hydrologic soil group (HSG), cover type, and size. The model then generates runoff coefficients. Although this is significant contribution to the existing body of knowledge, the authors admit that the model needs further development to accommodate pollutant loading computation. This current research addresses this deficiency.

In another instance, the Decision Evaluation in Complex Risk Network Systems (DECERNS) is a similar SDSS tool which focuses on land use planning and management (DECERNS - Team, 2006). DECERNS provides spatial data visualization for vector and raster models. It is used in the development of alternatives and criteria specification; implementation of the basic and advanced multi-criteria decision; and generation of various reports, including text descriptions, tables, diagrams, and maps (DECERNS -Team, 2006). It is a powerful commercial SDSS that caters for a larger community of state and regional officials, educators, and researchers. However, local governments who make critical land-use decisions are not direct beneficiaries because of the costs associated with this commercial software and also because this system lacks specific direct benefits to municipalities such as environmental legislation compliance tools. Thus, justifying the need to develop a municipal SDSS targeted for local communities to visualize the impacts of their decisions and simultaneously fulfill environmental legislation.

In another example, Rodman and Jackson (2006) used Python programming and the ArcGIS geoprocessor to develop a standalone spatial application, for the US Army Corps of Engineers (USACE), which performs data mining in geographic datasets. Python was selected as the language of choice because it is a powerful open source cross-platform programming language, that can run on Windows, Mac, or Unix and has a wealth of available code and tools that connect to databases for developing graphical user interfaces (GUI). The authors' goal was to create an application that relies on ArcGIS for handling spatial data formats, geographic coordinate system transformation, mapping, and geoprocessing. The resulting data mining application known as *Aspect*, is used to discover association rules that describe spatial relationships between geographic features. This allows decision makers to have tools that combine knowledge of terrain, travel routes, structure, land use/cover to improve situational awareness (Rodman & Jackson, 2006).

In this research, several implementation options were examined to determine the best approach to develop the graphical user interface. The option selected would need to take

Spatial Decision Support System (SDSS)

Tk/Tcl website, 2010).

**2.3 Web mapping application** 

methodology.

**2.2 Tkinter/python-based graphical user interface** 

for Stormwater Management and Water Quality Assessment 7

Tkinter Tool Command Language (Tk/Tcl) is an integral part of Python that provides a platform-independent windowing toolkit that is available to Python programmers using the Tkinter module (Official Tk/Tcl website, 2010). The Tkinter module (renamed to tkinter in Python version 3.x) is the standard Python interface to the Tk GUI toolkit (Official Python Website, 2011). Tkinter is basically a set of wrappers that implement the Tk widgets as Python classes. Tcl (Tool Command Language) is a dynamic programming language that is suitable for a very wide range of uses, including Web and desktop applications, network programming, embedded development, testing, general purpose programming, system administration, database work, and many more (Official Tk/Tcl website, 2010). (Official

Tk, a graphical user interface toolkit can be used in Tcl or in Perl language to create a number of GUI components such as buttons, labels and canvas. These components are known as widgets (Official Tk/Tcl website, 2010). Once these widgets have been created, three geometry managers are used to display them in relation to each other: pack, grid, and place. The "pack" geometry manager allows one to place your widgets in a row or column. "Grid" geometry manager allows the placement of widgets in a matrix. The third geometry manager, "place" provides the ability to place widgets by pixel or by the proportion of the way across the window that the programmer wants them to appear. More complex windows can be built using frames or nested frames (Well House Consultants, 2006). The python-based Tkinter appears to be a good option since the underlying geoprocessing scripts developed for this research are also Python-based. The limitation with this application is that the spatial components of the GUI such as coordinate system handling and spatial analysis would need to be programmed, which could be time-consuming. The

Esri's software, ArcGIS Server enables the user to create, manage, and distribute GIS services over the Web (Esri, 2011). Different Application Programming Interfaces (API) are available for web application development on various platforms. The Esri APIs include JavaScript, Flex, Silverlight, and Java Web Application Development Framework (ADF), and the NET Web ADF (Esri, 2011). ArcGIS Explorer Online is another option to present web map services, add other content to it, navigate, present and share the map. Explorer Online makes it possible to disseminate work on the Web and integrate map services from various sources (Esri, 2011). ArcGIS Explorer Online allows one to save maps to ArcGIS.com and choose to save them privately, share with a group or share over the Internet (ArcGIS Explorer Online Team, 2011). While the web-mapping approach is not emphasized in this research, a prototype web-mapping application was developed for spatial data editing and for interactive mapping. The next section further describes the SDSS development

The overall methodology for this research encompasses a three-tier approach that leads to the development of a spatial decision support system. The three levels are the data level, the

third option examined for the GUI was a web-mapping application.

**3. Spatial Decision Support System methodology** 

advantage of the already established ArcGIS geoprocessor for handling spatial data processing, coordinate system, and data editing. For this reason, some of the options examined for use include Esri's ArcGIS Add-ins, Tkinter/python-based GUI, a webmapping application, and a Java-based custom desktop application. Each of these applications is briefly described below while highlighting possible shortcomings as they relate to the current research.

#### **2.1 ArcGIS Add-ins**

In the 2011 version of ArcGIS 10 software, the concept of Add-ins was introduced to expand ArcGIS desktop's functionality and extend the interface (Burke and Elkins, 2010). Using Esri's ArcObjects, one can create new custom functionality using Add-ins to create a button or tool that a user interacts with to do something with the map or with GIS data. There are several types of Add-ins ranging from menus, buttons, toolbars, dockable windows, tool palettes, or applications and extensions. Add-in extensions are invisible to the user, but are event listeners that react to events by running code attached to them. The advantage here is that the complexity of the code is hidden from the end user, making complex processes more user-friendly. Add-ins can be built with C++ programming and often require a lot less code since the programmer simply adds more functionality instead of creating a new software package. They are also portable as they can be easily shared by email or file transfer from one user to another using a few steps to install (Burke & Elkins, 2010).

There are several ways to create Add-ins. Burke and Elkins (2010) use Microsoft Visual Studio 2008 and .NET to create a button Add-in using a wizard, assign an image icon, a name, and add reference to it. Microsoft Visual Studio 2008 Express is a free version that can be used to get started with Add-in development. Add-Ins can also be built with Java, for example using Eclipse development environment. To do any Add-in development, ArcObjects Software Development Kit (for .NET if using Visual Studio) needs to be installed. An Add-in is just one file with *EsriAddin* file extension, a folder-list container that contains everything needed for it to work without much setup. Add-ins can be placed in a specific directory where ArcGIS checks every time the software is launched. In summary, the Add-in creation process in Visual Studio involves making a Visual Studio project, creating an add-in, adding type to it, writing the business logic (what it does, and code behind it), and then testing it (Burke & Elkins, 2010).

Although Add-Ins seems straightforward, a Java or C++ programming skill is required to get the full effects (Burke & Elkins, 2010). Also, they are relatively new in ArcGIS and do not have a large user community yet, thus they were not selected for implementation in this research. Alternatively, Esri's ArcGIS Engine software can be used with ArcObjects to add dynamic mapping and GIS capabilities to existing applications, build custom mapping applications, or add geoprocessing scripts using application programming interfaces (APIs) for COM, .NET, Java, and C++ (Burke & Elkins, 2010). This option was not utilized in the current research because ArcGIS Engine software package was not available. The second option that was considered using a python-based module called Tkinter.

advantage of the already established ArcGIS geoprocessor for handling spatial data processing, coordinate system, and data editing. For this reason, some of the options examined for use include Esri's ArcGIS Add-ins, Tkinter/python-based GUI, a webmapping application, and a Java-based custom desktop application. Each of these applications is briefly described below while highlighting possible shortcomings as they

In the 2011 version of ArcGIS 10 software, the concept of Add-ins was introduced to expand ArcGIS desktop's functionality and extend the interface (Burke and Elkins, 2010). Using Esri's ArcObjects, one can create new custom functionality using Add-ins to create a button or tool that a user interacts with to do something with the map or with GIS data. There are several types of Add-ins ranging from menus, buttons, toolbars, dockable windows, tool palettes, or applications and extensions. Add-in extensions are invisible to the user, but are event listeners that react to events by running code attached to them. The advantage here is that the complexity of the code is hidden from the end user, making complex processes more user-friendly. Add-ins can be built with C++ programming and often require a lot less code since the programmer simply adds more functionality instead of creating a new software package. They are also portable as they can be easily shared by email or file transfer from one user to another using a few steps to install (Burke & Elkins,

There are several ways to create Add-ins. Burke and Elkins (2010) use Microsoft Visual Studio 2008 and .NET to create a button Add-in using a wizard, assign an image icon, a name, and add reference to it. Microsoft Visual Studio 2008 Express is a free version that can be used to get started with Add-in development. Add-Ins can also be built with Java, for example using Eclipse development environment. To do any Add-in development, ArcObjects Software Development Kit (for .NET if using Visual Studio) needs to be installed. An Add-in is just one file with *EsriAddin* file extension, a folder-list container that contains everything needed for it to work without much setup. Add-ins can be placed in a specific directory where ArcGIS checks every time the software is launched. In summary, the Add-in creation process in Visual Studio involves making a Visual Studio project, creating an add-in, adding type to it, writing the business logic (what it does, and code

Although Add-Ins seems straightforward, a Java or C++ programming skill is required to get the full effects (Burke & Elkins, 2010). Also, they are relatively new in ArcGIS and do not have a large user community yet, thus they were not selected for implementation in this research. Alternatively, Esri's ArcGIS Engine software can be used with ArcObjects to add dynamic mapping and GIS capabilities to existing applications, build custom mapping applications, or add geoprocessing scripts using application programming interfaces (APIs) for COM, .NET, Java, and C++ (Burke & Elkins, 2010). This option was not utilized in the current research because ArcGIS Engine software package was not available. The second option that was considered using a python-based module called

behind it), and then testing it (Burke & Elkins, 2010).

relate to the current research.

**2.1 ArcGIS Add-ins** 

2010).

Tkinter.

#### **2.2 Tkinter/python-based graphical user interface**

Tkinter Tool Command Language (Tk/Tcl) is an integral part of Python that provides a platform-independent windowing toolkit that is available to Python programmers using the Tkinter module (Official Tk/Tcl website, 2010). The Tkinter module (renamed to tkinter in Python version 3.x) is the standard Python interface to the Tk GUI toolkit (Official Python Website, 2011). Tkinter is basically a set of wrappers that implement the Tk widgets as Python classes. Tcl (Tool Command Language) is a dynamic programming language that is suitable for a very wide range of uses, including Web and desktop applications, network programming, embedded development, testing, general purpose programming, system administration, database work, and many more (Official Tk/Tcl website, 2010). (Official Tk/Tcl website, 2010).

Tk, a graphical user interface toolkit can be used in Tcl or in Perl language to create a number of GUI components such as buttons, labels and canvas. These components are known as widgets (Official Tk/Tcl website, 2010). Once these widgets have been created, three geometry managers are used to display them in relation to each other: pack, grid, and place. The "pack" geometry manager allows one to place your widgets in a row or column. "Grid" geometry manager allows the placement of widgets in a matrix. The third geometry manager, "place" provides the ability to place widgets by pixel or by the proportion of the way across the window that the programmer wants them to appear. More complex windows can be built using frames or nested frames (Well House Consultants, 2006). The python-based Tkinter appears to be a good option since the underlying geoprocessing scripts developed for this research are also Python-based. The limitation with this application is that the spatial components of the GUI such as coordinate system handling and spatial analysis would need to be programmed, which could be time-consuming. The third option examined for the GUI was a web-mapping application.

#### **2.3 Web mapping application**

Esri's software, ArcGIS Server enables the user to create, manage, and distribute GIS services over the Web (Esri, 2011). Different Application Programming Interfaces (API) are available for web application development on various platforms. The Esri APIs include JavaScript, Flex, Silverlight, and Java Web Application Development Framework (ADF), and the NET Web ADF (Esri, 2011). ArcGIS Explorer Online is another option to present web map services, add other content to it, navigate, present and share the map. Explorer Online makes it possible to disseminate work on the Web and integrate map services from various sources (Esri, 2011). ArcGIS Explorer Online allows one to save maps to ArcGIS.com and choose to save them privately, share with a group or share over the Internet (ArcGIS Explorer Online Team, 2011). While the web-mapping approach is not emphasized in this research, a prototype web-mapping application was developed for spatial data editing and for interactive mapping. The next section further describes the SDSS development methodology.

#### **3. Spatial Decision Support System methodology**

The overall methodology for this research encompasses a three-tier approach that leads to the development of a spatial decision support system. The three levels are the data level, the

Spatial Decision Support System (SDSS)

diffuse basins (non- point source of pollution).

for Stormwater Management and Water Quality Assessment 9

• Land use characterization was achieved by performing a spatial union of an existing land use layer with the mini-basin polygons. This resulted in one of the parameters in

the hydrological model; area values of each land use type in each mini-basin. • A geospatial approach for hydrological modeling was developed using python programming to determine the water quality and runoff effect of land use change. The uniqueness of the model component of the SDSS is in the integration of Java and Python programming languages. The user interface was written using Java, but an existing hydrological model was programmed into a geospatially-enabled hydrological model using Python programming. The hydrological model adopted is a series of three equations based on "The Simple Method" by Schueler (1987), often called the Curve Number Method. The equations were spatially enabled by encoding them into ArcGIS scripting environment called arcpy (python for ArcGIS). First, an area-weighted runoff coefficient (weighted Cvalue) script was written using Python's mathematical operations and program looping to

calculate weighted runoff coefficient (Rvi) for each outfall basin (**equation 1**).

Where *Rv*= Runoff coefficient for each land use within the outfall drainage area.

Where *Ai*=Area of a specific land use within the outfall drainage basin - *AT*=Total land area within the drainage area of the major outfall - *Rv*=Runoff coefficient (C-value) for a particular land use - *Rvi*=weighted C-value for the drainage area of the major outfall


Where *Li*= Seasonal pollutant load in pounds per outfall/season



following equation:



Rvi=∑(Ai\*Rv)/ ∑Ai (1)

*Ai*=Land area of each land use within the drainage area of the major outfall. Next, Event Mean Concentration (EMC) of each pollutant was calculated using the result from equation 1 above.

EMC=∑(Ai/AT)\*(Rv/Rvi)\*Ci (2)

Finally, the pollutant loadings (weight of pollutant per season) were calculated using the

These three equations have been sequentially executed and the resulting table would have the output that is used to create thematic maps that show pollutant hotspots in the study

Li=[(P\*CF\*Rvi)/12](EMC\*Ai) (3)

result was a delineation of basins, subbasins, and mini-basins. It is important to note that mini-basins are categorized into outfall basins (point source of pollution) and

model development level, and the development of a graphical user interface. The data level was implemented primarily for creating a comprehensive database that stores all the data needed for water quality evaluation. Modeling the effects of land use change on water quantity and quality requires a multidisciplinary approach which incorporates many different data types. Therefore, several data sets such as sewer infrastructure data, rainfall data, soils/curve number, and pollutant sampling data were compiled into a central geodatabase (geographic geodatabase).

#### **3.1 Database component**

Geodatabase development for stormwater management in particular can be challenging due to the multitude of different types of stormwater features involved, and the complex topological relationships that exist between them. In this research a geodatabase schema was developed that can be adopted by other municipalities as a template for stormwater mapping. Specifically, ArcGIS software suite by Esri was used to develop an enterprise level geodatabase for managing the watershed and subsurface infrastructure data. Geodatabases are needed for the successful storage, access, retrieval, manipulation, and management of massive data sets typical of municipalities. The most critical data set emphasized in this research was the sewer infrastructure data. Specifically, two separate datasets were developed for stormwater networks and sanitary sewer (wastewater) networks. A dataset is a set of georeferenced data layers that are topologically related and are in the same spatial extent (Esri, 2010). The stormwater dataset includes locations of stormwater inlets, pipes, headwalls, and culverts captured using mapping-grade Global Positioning Systems (GPS), as well as creeks, rivers, and delineated drainage basins from topographic mapping. Similarly, the sanitary sewer database consists of GPS locations and dimensions of manholes, sewer lines, pump stations, wastewater treatment plants, and sewer drainage basins. Network topology rules were developed to enforce the connectivity of the sewer features such that all features that participate in a network are topologically connected. For example, if an inlet is not connected to any pipe or waterway, it would be marked as an error. Similarly, attribute validation rules are also established for the sewer features to minimize errors when editing the sewer data. After the database design the next step was to populate the "spatial container" with data from multiple sources including GPS, aerial photos, engineering design drawings, and digital elevation models. Once all the data had been collected, the next step in the SDSS development was model development.

#### **3.2 Model development component**

The model development stage is critical to the full-functioning of any SDSS because data has to be processed to make it meaningful for decision-making. The general methodology of the model development stage is as follows:


model development level, and the development of a graphical user interface. The data level was implemented primarily for creating a comprehensive database that stores all the data needed for water quality evaluation. Modeling the effects of land use change on water quantity and quality requires a multidisciplinary approach which incorporates many different data types. Therefore, several data sets such as sewer infrastructure data, rainfall data, soils/curve number, and pollutant sampling data were compiled into a central

Geodatabase development for stormwater management in particular can be challenging due to the multitude of different types of stormwater features involved, and the complex topological relationships that exist between them. In this research a geodatabase schema was developed that can be adopted by other municipalities as a template for stormwater mapping. Specifically, ArcGIS software suite by Esri was used to develop an enterprise level geodatabase for managing the watershed and subsurface infrastructure data. Geodatabases are needed for the successful storage, access, retrieval, manipulation, and management of massive data sets typical of municipalities. The most critical data set emphasized in this research was the sewer infrastructure data. Specifically, two separate datasets were developed for stormwater networks and sanitary sewer (wastewater) networks. A dataset is a set of georeferenced data layers that are topologically related and are in the same spatial extent (Esri, 2010). The stormwater dataset includes locations of stormwater inlets, pipes, headwalls, and culverts captured using mapping-grade Global Positioning Systems (GPS), as well as creeks, rivers, and delineated drainage basins from topographic mapping. Similarly, the sanitary sewer database consists of GPS locations and dimensions of manholes, sewer lines, pump stations, wastewater treatment plants, and sewer drainage basins. Network topology rules were developed to enforce the connectivity of the sewer features such that all features that participate in a network are topologically connected. For example, if an inlet is not connected to any pipe or waterway, it would be marked as an error. Similarly, attribute validation rules are also established for the sewer features to minimize errors when editing the sewer data. After the database design the next step was to populate the "spatial container" with data from multiple sources including GPS, aerial photos, engineering design drawings, and digital elevation models. Once all the data had

been collected, the next step in the SDSS development was model development.

by the United States Environmental Protection Agency (EPA, 1992).

The model development stage is critical to the full-functioning of any SDSS because data has to be processed to make it meaningful for decision-making. The general methodology of the

• Stormwater outfalls were extracted by querying for stormwater pipes (diameter of at least 12 inches in industrial areas and at least 36 inches in all other land uses) that empty into major rivers and creeks. This query is based on the definition of an outfall

• Drainage boundaries were delineated using topographic elevation and the spatial and topologic locations of underground and above-ground stormwater infrastructure. The

geodatabase (geographic geodatabase).

**3.2 Model development component** 

model development stage is as follows:

**3.1 Database component** 

result was a delineation of basins, subbasins, and mini-basins. It is important to note that mini-basins are categorized into outfall basins (point source of pollution) and diffuse basins (non- point source of pollution).


The uniqueness of the model component of the SDSS is in the integration of Java and Python programming languages. The user interface was written using Java, but an existing hydrological model was programmed into a geospatially-enabled hydrological model using Python programming. The hydrological model adopted is a series of three equations based on "The Simple Method" by Schueler (1987), often called the Curve Number Method. The equations were spatially enabled by encoding them into ArcGIS scripting environment called arcpy (python for ArcGIS). First, an area-weighted runoff coefficient (weighted Cvalue) script was written using Python's mathematical operations and program looping to calculate weighted runoff coefficient (Rvi) for each outfall basin (**equation 1**).

$$\mathbf{Rv\_{i}=\sum(A\_{i}\mathbf{^{\*}Rv})/\sum A\_{i}}\tag{1}$$

Where *Rv*= Runoff coefficient for each land use within the outfall drainage area.

*Ai*=Land area of each land use within the drainage area of the major outfall. Next, Event Mean Concentration (EMC) of each pollutant was calculated using the result from equation 1 above.

$$\text{EMC} \equiv \sum (\text{A}\_{i}/\text{A}\_{\text{T}})^{\*} (\text{Rv}/\text{Rv}\_{i})^{\*} \text{C}\_{i} \tag{2}$$

Where *Ai*=Area of a specific land use within the outfall drainage basin


Finally, the pollutant loadings (weight of pollutant per season) were calculated using the following equation:

$$\text{Li} = [(\text{P\*CF\*Rvi})/12](\text{EMC\*Ai})\tag{3}$$

Where *Li*= Seasonal pollutant load in pounds per outfall/season


These three equations have been sequentially executed and the resulting table would have the output that is used to create thematic maps that show pollutant hotspots in the study

Spatial Decision Support System (SDSS)

Fig. 1. The hierarchy of Java Swing Components

for Stormwater Management and Water Quality Assessment 11

area. The equations would be encapsulated in the user interface and simplified as a onebutton click. The next section describes the user interface in more detail.

#### **3.3 SDSS GUI implementation**

The Java-based custom desktop application was selected for the development of the SDSS user interface. GUI programming is a complicated task involving multiple steps, because many interface widgets need to be generated and specified. Furthermore, the alignment and appearance of these interface widgets also required a lot of programming. The GUI was developed using version 1.0.23 of Java Software Development Kit (JDK) and version 6.9.1 of NetBeans Integrated Development Environment (IDE). NetBeans and JDK are free programs obtained from NetBeans' website (NetBeans official website, 2011). NetBeans is an IDE that simplifies the programmer's task and ease the burden of GUI programming. To make the GUI components produce results that react to user input, many kinds of events must be handled properly and programmed to make the widgets functional. Almost every GUI widget is a generator that is capable of initiating multiple kinds of events. Java uses "event listener" to capture events that are generated from user interactions. A "Listener" is a an interface that a programmer can put all information that is needed to make the buttons functional. An event object is a holder of generated events. The "Listener" can only capture the registered event handler. By this observer pattern, a Java GUI program can properly handle and process user interactions, providing a flexible and extendable programming strategy.

In our GUI development, the typical events generated were action events and mouse events. An action event is generated when a button is pressed, or by pressing *Enter* in a text field, or when a menu item is selected. (Schildt, 2001). The programmer has to implement an action listener to define what methods are invoked when a user performs certain operations. With registered necessary event handlers, an "actionPerformed" message is implemented to handle all generated events on the relevant component. For example, the *Save/Add* button on the user interface listens for the user to click the button, then performs the save operation to save the user input from the text field into the corresponding table in the basins ArGIS personal geodatabase (Microsoft Access database).

Swing components were developed by SUN Microsystem's, to provide a more sophisticated and user-friendly GUI programming paradigm, including frames, buttons, panels, text fields, and labels (NetBeans official website, 2011). **Figure 1** shows the component hierarchy of the Java Swing as illustrated by Reddy (2007). Other modules used include input/output (java.io), abstract windowing toolkit event (java.awt.event), and java.lang.reflect methods. Due to the feature of not invoking OS resource, Swing is considered as lightweight component that is extended on top of many widgets of an AWT packet.

**Figure 2** illustrates a use-case scenario that shows how a user would interact with the system and how the system responds to the user's interaction. In general, the SDSS allows the users to:


area. The equations would be encapsulated in the user interface and simplified as a one-

The Java-based custom desktop application was selected for the development of the SDSS user interface. GUI programming is a complicated task involving multiple steps, because many interface widgets need to be generated and specified. Furthermore, the alignment and appearance of these interface widgets also required a lot of programming. The GUI was developed using version 1.0.23 of Java Software Development Kit (JDK) and version 6.9.1 of NetBeans Integrated Development Environment (IDE). NetBeans and JDK are free programs obtained from NetBeans' website (NetBeans official website, 2011). NetBeans is an IDE that simplifies the programmer's task and ease the burden of GUI programming. To make the GUI components produce results that react to user input, many kinds of events must be handled properly and programmed to make the widgets functional. Almost every GUI widget is a generator that is capable of initiating multiple kinds of events. Java uses "event listener" to capture events that are generated from user interactions. A "Listener" is a an interface that a programmer can put all information that is needed to make the buttons functional. An event object is a holder of generated events. The "Listener" can only capture the registered event handler. By this observer pattern, a Java GUI program can properly handle and process user interactions, providing a flexible and extendable programming

In our GUI development, the typical events generated were action events and mouse events. An action event is generated when a button is pressed, or by pressing *Enter* in a text field, or when a menu item is selected. (Schildt, 2001). The programmer has to implement an action listener to define what methods are invoked when a user performs certain operations. With registered necessary event handlers, an "actionPerformed" message is implemented to handle all generated events on the relevant component. For example, the *Save/Add* button on the user interface listens for the user to click the button, then performs the save operation to save the user input from the text field into the corresponding table in the basins ArGIS

Swing components were developed by SUN Microsystem's, to provide a more sophisticated and user-friendly GUI programming paradigm, including frames, buttons, panels, text fields, and labels (NetBeans official website, 2011). **Figure 1** shows the component hierarchy of the Java Swing as illustrated by Reddy (2007). Other modules used include input/output (java.io), abstract windowing toolkit event (java.awt.event), and java.lang.reflect methods. Due to the feature of not invoking OS resource, Swing is considered as lightweight

**Figure 2** illustrates a use-case scenario that shows how a user would interact with the system and how the system responds to the user's interaction. In general, the SDSS allows

1. input data into simplified forms on the interface and save edits at the click of a button,

3. use one-button click to generate maps and output reports that are needed for informed

2. use one-button click to run complex geoprocessing in the background, and

decision-making and for environmental legislation compliance.

component that is extended on top of many widgets of an AWT packet.

button click. The next section describes the user interface in more detail.

**3.3 SDSS GUI implementation** 

personal geodatabase (Microsoft Access database).

strategy.

the users to:

Fig. 1. The hierarchy of Java Swing Components

Spatial Decision Support System (SDSS)

Huntsville uses.

this purpose.

Fig. 3a. ArcGIS 10 interface

for Stormwater Management and Water Quality Assessment 13

web browser showing a FlexViewer web map developed by the authors for viewing and editing geographic data associated with the SDSS. Finally, the *Open GTViewer* button opens an internal GIS desktop application (by Graphics Technologies, Inc.) that the City of

With this application, non-GIS-expert users can collect pollutant data such as pollutant name, concentrations, etc., and save into the master database tables via the *Pollutant Editor* form. The data entry is developed in such a way that the pollutant data can be edited by adding new records or editing existing records, **Figure 4** shows the pollutant editor form for

Since different municipalities use different land use classification schemes, they may find it easier to manually enter literature-based runoff coefficients from a land use lookup table. For this reason, a runoff coefficient editor was created for manually populating C-

The core of the model component is encapsulated in the "*Calculate Loadings"* Button on the GUI. When this button is pressed, the hydrological modeling equations (1-3) described in the model development stage would be executed. **Figure 6** shows the two buttons for processing pollutant loading calculations and for viewing output maps. To demonstrate the SDSS application, two applications were illustrated: a report generation function and a

Values/runoff coefficients; **Figure 5** shows the *Curve Number Editor* form.

comparison of pre-development and post-development pollution contributions.

Fig. 2. Use Case Diagram showing how a user interacts with the interface

#### **4. Results**

A graphical user interface for the SDSS was developed for non-GIS professionals as a frontend for data collection as well as for executing hydrological modeling scripts. The developed geospatially-enabled hydrological model can optionally be executed in ArcGIS as a Python script tool, but the complexity of the ArcGIS interface **(Figure 3a)** has been simplified for non-GIS users into a simple, straightforward interface as shown in **Figure 3b.** The resulting GUI allows the complex pollutant loadings scripts to be executed in a less intimidating environment.

The *Calculate Loadings* in Figure 3b button runs a Java command that executes arcpy scripts (ArcGIS 10 Python scripts), which then progress by invoking classes and methods on the ArcGIS geoprocessor object. Specifically, the *Generate Loading* scripts fetch pollutant data such as pollutant concentration and curve number information entered by the user through the Java interface and use that as input for calculating the Estimated Mean Concentrations and pollutants loadings. Pollutant hotspot maps indicating the spatial distribution of outfall basin pollutant loadings are also generated upon pressing the *Calculate Loadings* button. The *Loadings Maps* button fetches the generated maps and displays them using a PDF reader such as Adobe Reader, ready for inclusion in reports, and for immediate decision-making. The user also has the option to press the *Generate Report* function on the file menu executes Java commands to generate water quality reports based on EMC's and Pollutant loadings fields from the ArcGIS geodatabase. When clicked, the *Open Online Maps* button opens a

Fig. 2. Use Case Diagram showing how a user interacts with the interface

A graphical user interface for the SDSS was developed for non-GIS professionals as a frontend for data collection as well as for executing hydrological modeling scripts. The developed geospatially-enabled hydrological model can optionally be executed in ArcGIS as a Python script tool, but the complexity of the ArcGIS interface **(Figure 3a)** has been simplified for non-GIS users into a simple, straightforward interface as shown in **Figure 3b.** The resulting GUI allows the complex pollutant loadings scripts to be executed in a less

The *Calculate Loadings* in Figure 3b button runs a Java command that executes arcpy scripts (ArcGIS 10 Python scripts), which then progress by invoking classes and methods on the ArcGIS geoprocessor object. Specifically, the *Generate Loading* scripts fetch pollutant data such as pollutant concentration and curve number information entered by the user through the Java interface and use that as input for calculating the Estimated Mean Concentrations and pollutants loadings. Pollutant hotspot maps indicating the spatial distribution of outfall basin pollutant loadings are also generated upon pressing the *Calculate Loadings* button. The *Loadings Maps* button fetches the generated maps and displays them using a PDF reader such as Adobe Reader, ready for inclusion in reports, and for immediate decision-making. The user also has the option to press the *Generate Report* function on the file menu executes Java commands to generate water quality reports based on EMC's and Pollutant loadings fields from the ArcGIS geodatabase. When clicked, the *Open Online Maps* button opens a

**4. Results** 

intimidating environment.

web browser showing a FlexViewer web map developed by the authors for viewing and editing geographic data associated with the SDSS. Finally, the *Open GTViewer* button opens an internal GIS desktop application (by Graphics Technologies, Inc.) that the City of Huntsville uses.

With this application, non-GIS-expert users can collect pollutant data such as pollutant name, concentrations, etc., and save into the master database tables via the *Pollutant Editor* form. The data entry is developed in such a way that the pollutant data can be edited by adding new records or editing existing records, **Figure 4** shows the pollutant editor form for this purpose.

Since different municipalities use different land use classification schemes, they may find it easier to manually enter literature-based runoff coefficients from a land use lookup table. For this reason, a runoff coefficient editor was created for manually populating C-Values/runoff coefficients; **Figure 5** shows the *Curve Number Editor* form.

The core of the model component is encapsulated in the "*Calculate Loadings"* Button on the GUI. When this button is pressed, the hydrological modeling equations (1-3) described in the model development stage would be executed. **Figure 6** shows the two buttons for processing pollutant loading calculations and for viewing output maps. To demonstrate the SDSS application, two applications were illustrated: a report generation function and a comparison of pre-development and post-development pollution contributions.

Fig. 3a. ArcGIS 10 interface

Spatial Decision Support System (SDSS)

Fig. 5. The Curve Number Editor

Fig. 6. Pollutant loadings and output map button

for Stormwater Management and Water Quality Assessment 15


Fig. 3b. The newly developed, less intimidating graphical user interface


Fig. 4. Pollutant Editor Form

Fig. 3b. The newly developed, less intimidating graphical user interface

Fig. 4. Pollutant Editor Form


Fig. 5. The Curve Number Editor


Fig. 6. Pollutant loadings and output map button

Spatial Decision Support System (SDSS)

for Stormwater Management and Water Quality Assessment 17

Table 1. Land use area of each mini-basin in acres. ( Rvi = weighted runoff coefficient)

#### **4.1 SDSS Demonstration 1: Report generation and pollution contributions of land use change (Bridge Street Town Center)**

First, a report generation function for municipal environmental compliance and pollution contributions of a land use change are illustrated. To test the report function, a report was generated for the City of Huntsville stormwater outfalls using 2004 data. The *Generate Report* function on the file menu generates a report ready for submission to the United States Environmental Protection Agency (US EPA) for stormwater regulation compliance. Stormwater pollution is regulated under the Clean Water Act (CWA) of 1972. Under the CWA, the US EPA has implemented pollution control programs and set standards that make it unlawful for industries, municipalities, and other facilities to discharge any pollutant into navigable waters, without a permit (US Congress, 1972). US EPA's National Pollutant Discharge Elimination System (NPDES) permit program controls these point source discharges and has put specific regulation in place as a guide for NPDES permit *applicants* (US Congress, 1972). **Figure 7** illustrates the report generation workflow. The user interface simplifies this process for land use decision-makers by using just a few clicks to quantify pollution and generate maps for decision support or reports for environmental legislation programs such as NPDES permit compliance. Instant report generation saves time for municipal officials so that they focus more on decision-making instead of technical setbacks. The added value of the geospatially-enabled hydrological model is the ability to produce pollutant hotspot maps that unveil spatial trends, allowing land use policy makers to visualize the environmental impact of their decisions.

Fig. 7. Complete workflow showing land use characterization, model execution, and report generation

The first set of processes in **Figure 7** calculate the land use acreage in each mini-basin, the second calculates EMC's and pollutant loadings, and the last step selects outfall polygons and generates summarized Pivot tables with a report output. The report includes the stormwater outfall ID, weighted runoff coefficients, event mean concentration of pollutants, and pollutant loading in each outfall basin. **Tables 1-3** are respectively, land use summaries, Event Mean Concentrations, pollutant loadings generated as part of the report. **Figure 8** is an example of pollutant hotspot maps generated using the tools from the user interface.


Table 1. Land use area of each mini-basin in acres. ( Rvi = weighted runoff coefficient)

16 Water Quality Monitoring and Assessment

**4.1 SDSS Demonstration 1: Report generation and pollution contributions of land use** 

First, a report generation function for municipal environmental compliance and pollution contributions of a land use change are illustrated. To test the report function, a report was generated for the City of Huntsville stormwater outfalls using 2004 data. The *Generate Report* function on the file menu generates a report ready for submission to the United States Environmental Protection Agency (US EPA) for stormwater regulation compliance. Stormwater pollution is regulated under the Clean Water Act (CWA) of 1972. Under the CWA, the US EPA has implemented pollution control programs and set standards that make it unlawful for industries, municipalities, and other facilities to discharge any pollutant into navigable waters, without a permit (US Congress, 1972). US EPA's National Pollutant Discharge Elimination System (NPDES) permit program controls these point source discharges and has put specific regulation in place as a guide for NPDES permit *applicants* (US Congress, 1972). **Figure 7** illustrates the report generation workflow. The user interface simplifies this process for land use decision-makers by using just a few clicks to quantify pollution and generate maps for decision support or reports for environmental legislation programs such as NPDES permit compliance. Instant report generation saves time for municipal officials so that they focus more on decision-making instead of technical setbacks. The added value of the geospatially-enabled hydrological model is the ability to produce pollutant hotspot maps that unveil spatial trends, allowing land use policy makers

Fig. 7. Complete workflow showing land use characterization, model execution, and report

The first set of processes in **Figure 7** calculate the land use acreage in each mini-basin, the second calculates EMC's and pollutant loadings, and the last step selects outfall polygons and generates summarized Pivot tables with a report output. The report includes the stormwater outfall ID, weighted runoff coefficients, event mean concentration of pollutants, and pollutant loading in each outfall basin. **Tables 1-3** are respectively, land use summaries, Event Mean Concentrations, pollutant loadings generated as part of the report. **Figure 8** is an example of pollutant hotspot maps generated using the tools from the user interface.

**change (Bridge Street Town Center)** 

to visualize the environmental impact of their decisions.

generation


TSS= Total Suspended Solids, TDS = Total Dissolved Solids, PO4 = Phosphates) Table 2. Annual Event Mean Concentration of each pollutant in each mini-basin

(Rvi = weighted runoff coefficient, EMC= Event Mean Concentration, BOD=Biological Oxygen Demand, COD= Chemical Oxygen Demand,

Spatial Decision Support System (SDSS)

for Stormwater Management and Water Quality Assessment 19

Table 3. Annual pollutant loadings for each pollutant in each mini-basin in pounds/year. (oPO4\_Load = orthophosphates, other pollutant acronyms as described in Table 2)

The second application of the SDSS is the assessment of pre-development (2004) and postdevelopment (2010) conditions of an existing commercial establishment to evaluate the pollution contribution of Bridge Street Town Center in Huntsville, Alabama. Bridge Street is located in Cummings Research Park and is mostly a commercial development that also has condominiums, hotels and recreational facilities. TThis section demonstrates the use of the SDSS by comparing Pre and Post Development conditions of mini-basin IND06018 which encompasses Bridge Street Town Center. In 2004 Bridge Street did not exist, but in 2010 the

The land use changes shown in **Figure 10** indicate that cropland (-38%) and campus/institutional (-30) land use areas decreased while all other land uses increased. The highest land use change after the development was commercial, at 41% increase. **Figure 11** shows that runoff coefficients increased by 12%, and the event mean concentration for all pollutants increased, with Oil and grease showing the highest increase (50%). The increase

**4.2 SDSS Demonstration 2: Pollution contribution of land use change** 

land was highly developed as shown in **Figure 9.**

(Rvi = weighted runoff coefficient, EMC= Event Mean Concentration, BOD=Biological Oxygen Demand, COD= Chemical Oxygen Demand,

TSS= Total Suspended Solids, TDS = Total Dissolved Solids, PO4 = Phosphates)

Table 2. Annual Event Mean Concentration of each pollutant in each mini-basin


Table 3. Annual pollutant loadings for each pollutant in each mini-basin in pounds/year. (oPO4\_Load = orthophosphates, other pollutant acronyms as described in Table 2)

#### **4.2 SDSS Demonstration 2: Pollution contribution of land use change**

The second application of the SDSS is the assessment of pre-development (2004) and postdevelopment (2010) conditions of an existing commercial establishment to evaluate the pollution contribution of Bridge Street Town Center in Huntsville, Alabama. Bridge Street is located in Cummings Research Park and is mostly a commercial development that also has condominiums, hotels and recreational facilities. TThis section demonstrates the use of the SDSS by comparing Pre and Post Development conditions of mini-basin IND06018 which encompasses Bridge Street Town Center. In 2004 Bridge Street did not exist, but in 2010 the land was highly developed as shown in **Figure 9.**

The land use changes shown in **Figure 10** indicate that cropland (-38%) and campus/institutional (-30) land use areas decreased while all other land uses increased. The highest land use change after the development was commercial, at 41% increase. **Figure 11** shows that runoff coefficients increased by 12%, and the event mean concentration for all pollutants increased, with Oil and grease showing the highest increase (50%). The increase

Spatial Decision Support System (SDSS)

development




41.3


Fig. 10. Percent Change in Land use Acreage after Bridge Street Development


0

% Change

20

40

60

for Stormwater Management and Water Quality Assessment 21

Fig. 9. 2004 and 2010 Orthophotos at the Bridge Street Town Center location: pre and post

0.1 0.0

CAMP COM CRP HDR LDR OPN TRS WTR

Land Use Types

9.4

13.6

3.4

in runoff coefficients can be attributed to an increase in impervious surface as the area was mostly cropland in 2004 and in 2011 it is mostly urban. The higher oil and grease increase can be attributed to the increase in parking areas where oil leaks are possible from parked vehicles, and grease from the commercial establishments in the town center are inevitable. The percent change in pollutant loadings for oil and grease were also analyzed and shown in **Figure 12.** Consequently, the loadings for oil and grease also increased for the dry, wet, transitional and annual seasons.

Fig. 8. Example of hotspot map for Total Suspended Solids (TSS) loads in Huntsville, Alabama (units in pounds/year) The map is presented in State Plane Coordinate System, North American Datum of 1983, Alabama East FIPS 0101.

in runoff coefficients can be attributed to an increase in impervious surface as the area was mostly cropland in 2004 and in 2011 it is mostly urban. The higher oil and grease increase can be attributed to the increase in parking areas where oil leaks are possible from parked vehicles, and grease from the commercial establishments in the town center are inevitable. The percent change in pollutant loadings for oil and grease were also analyzed and shown in **Figure 12.** Consequently, the loadings for oil and grease also increased for the dry, wet,

Fig. 8. Example of hotspot map for Total Suspended Solids (TSS) loads in Huntsville, Alabama (units in pounds/year) The map is presented in State Plane Coordinate System,

North American Datum of 1983, Alabama East FIPS 0101.

transitional and annual seasons.

Fig. 9. 2004 and 2010 Orthophotos at the Bridge Street Town Center location: pre and post development

Fig. 10. Percent Change in Land use Acreage after Bridge Street Development

Spatial Decision Support System (SDSS)

**5. Summary and conclusions** 

where they are needed the most.

that may result in pollution.

*Water*. Washington, DC.

\_ins-for-ArcGIS-Desktop-10.aspx

http://www.Esri.com/

use\_10year.pdf

**6. References** 

for Stormwater Management and Water Quality Assessment 23

The first objective of this study was to develop a front-end graphical user interface (GUI) that is robust enough to facilitate data collection, quantify and geolocate urban pollution, but simple enough for land use decision-makers, whose expertise is neither GIS nor water quality assessment. The second objective was to demonstrate the use of the SDSS for generating reports and assessing pre-development and post-development conditions of a land use change*.* A desktop application has been designed and implemented using Java programming in NetBeans IDE. The GUI is a user-friendly interface that conceals program details, saving the user valuable time from focusing on technical complications, while still getting a powerful tool for the intended needs. The GUI provides custom tools for quick data input, spatial analysis, report generation, and environmental regulation compliance. The output is a graphical user interface for municipal officials or other land-use decisionmakers and watershed managers for visualizing and quantifying the effects of land use on the environment. A robust, but user-friendly custom interface for local land-use officials is necessary for decision-makers to be more environmentally aware and to channel resources

One shortcoming of the desktop application is that it only allows for pollutant and curve number editing, but does not support editing GIS data. As a result, a web-mapping application is under development to address this limitation. A link to the website was established in the user interface using the *Open Online Maps* button. The web mapping application was developed using Esri's FlexViewer API and hosted on the City of Huntsville's ArcGIS server as a prototype that is not yet available for public view. The web map can be loaded by field inspectors onto a mobile device or compatible smart phones to edit and modify GIS data, sending data back to the master database via a cell phone network or the Internet. Similarly, it can be used by local citizens to pinpoint incidents such as locations of illegal pollutant discharge, sanitary sewer overflows, or flooding complaints

ArcGIS Explorer Online Team. (2011). Online Electronic Documentation. Accessed February

DECERNS–Team. (2006). Decision Evaluation in Complex Risk Network Systems. Accessed

EPA. (1992). Guidance Manual for the Preparation of Part 2 of the NPDES Permit

http://blogs.esri.com/Dev/blogs/arcgisdesktop/archive/2010/05/05/Add\_2D00

Giannotti, L A. & Arnold, C L. (2011). Changing Land use Decision Making One Town at a Time. *North Carolina State University.* Accessed March 1, 2011. Available from

Esri. (2010). ArcGIS Server Technology. Accessed February 10, 2010. Available from

Esri. (2011). Add-ins for ArcGIS Desktop. Accessed March 26, 2011. Available from

http://nemo.uconn.edu/publications/about\_nemo/changing\_land

Applications for Discharge from Municipal Separate Storm Systems. *EPA Office of* 

4, 2011. Available from http://explorer.arcgis.com/

2/4/11. Available from http://195.112.127.216/?q=rd

Fig. 11. Percent change in runoff coefficients and Event Mean Concentrations

Fig. 12. Change in Seasonal Pollutant Loadings for Bridge Street area

#### **5. Summary and conclusions**

22 Water Quality Monitoring and Assessment

12.0

RV TSS TKN TDS PO4 oPO4 OG NT COD BOD Pollutants

21.0

50.0

5.0

12.0

3.0

0

10

12.0

23.0

4.0 4.0

Fig. 11. Percent change in runoff coefficients and Event Mean Concentrations

Fig. 12. Change in Seasonal Pollutant Loadings for Bridge Street area

20

30

% Change

40

50

60

The first objective of this study was to develop a front-end graphical user interface (GUI) that is robust enough to facilitate data collection, quantify and geolocate urban pollution, but simple enough for land use decision-makers, whose expertise is neither GIS nor water quality assessment. The second objective was to demonstrate the use of the SDSS for generating reports and assessing pre-development and post-development conditions of a land use change*.* A desktop application has been designed and implemented using Java programming in NetBeans IDE. The GUI is a user-friendly interface that conceals program details, saving the user valuable time from focusing on technical complications, while still getting a powerful tool for the intended needs. The GUI provides custom tools for quick data input, spatial analysis, report generation, and environmental regulation compliance. The output is a graphical user interface for municipal officials or other land-use decisionmakers and watershed managers for visualizing and quantifying the effects of land use on the environment. A robust, but user-friendly custom interface for local land-use officials is necessary for decision-makers to be more environmentally aware and to channel resources where they are needed the most.

One shortcoming of the desktop application is that it only allows for pollutant and curve number editing, but does not support editing GIS data. As a result, a web-mapping application is under development to address this limitation. A link to the website was established in the user interface using the *Open Online Maps* button. The web mapping application was developed using Esri's FlexViewer API and hosted on the City of Huntsville's ArcGIS server as a prototype that is not yet available for public view. The web map can be loaded by field inspectors onto a mobile device or compatible smart phones to edit and modify GIS data, sending data back to the master database via a cell phone network or the Internet. Similarly, it can be used by local citizens to pinpoint incidents such as locations of illegal pollutant discharge, sanitary sewer overflows, or flooding complaints that may result in pollution.

#### **6. References**


**2** 

**An Overview** 

*Portugal* 

*1ESTSetúbal-LabIM/IPS, Setúbal,* 

**Water Quality Monitoring and Associated** 

Octavian Postolache1,2, Pedro Silva Girão2 and José Miguel Dias Pereira1,2

Water is essential to life, as we know it. However, statistics reveal that, in 2000, one billion people lacked access to safe drinking water and 2.4 billion to adequate sanitation. To achieve United Nations target of reducing by half the proportion of people without sustainable access to safe drinking water by 2015, an additional 1.5 billion people would require access to some form of improved water supply by 2015, that is an additional 100 million people

Because water sources are limited, it is of paramount importance to keep its quality at the highest level possible. Threats to water are manifold, from industry to natural phenomena, and water quality assurance is a basic environmental issue involving from political to technical aspects and options, but it is obvious that no assessment of water quality is possible without a quantitative identification of some characteristics, a process commonly

This chapter is an overview on water quality and on its monitoring. The text reflects he experience of the authors on the subject, presents some research and development results they obtained in the last decade and includes data gathered from different sources, namely from USEPA reports and North Caroline State University Water Quality Group documents. The text includes remarks about measuring techniques for different water quality parameters that result from the experience acquired by the authors in the implementation of several water quality measuring units. The last part of the chapter proposes architectures and intelligent signal processing techniques for distributed water quality monitoring

Water quality is commonly defined by its physical, chemical, biological and aesthetic (appearance and smell) characteristics. Water may be used for drinking, irrigating crops and watering stock, industrial processes, production of fish, shellfish and crustaceans, wildlife

**1. Introduction** 

each year (or 274,000/day) until 2015.

called water quality monitoring.

networks.

**2. Water quality** 

**Distributed Measurement Systems:** 

*2Instituto de Telecomunicações/Instituto Superior Técnico, Lisboa,* 


https://engineering.purdue.edu/~lthia/


http://www.hindawi.com/journals/ace/2010/810402/

### **Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview**

Octavian Postolache1,2, Pedro Silva Girão2 and José Miguel Dias Pereira1,2 *1ESTSetúbal-LabIM/IPS, Setúbal, 2Instituto de Telecomunicações/Instituto Superior Técnico, Lisboa, Portugal* 

#### **1. Introduction**

24 Water Quality Monitoring and Assessment

NetBeans Official Website. (2011). NetBeans Integrated Development Environment.

Official Python Website. (2011). Python Programming Language. Accessed January 25, 2011.

Official Tk/Tcl website. (2011). The Tcl Developer Exchange. Accessed January 25, 2011.

Purdue Research Foundation. (2010). Long Term Hydrologic Impact Analysis. Accessed

Reddy, D. (2007). UI's and Java Swing -Java Swing Components Hierarchy. Accessed Feb. 4,

http://www.google.com/imgres?imgurl=http://jroller.com/DhilshukReddy/reso

Rodman L C. & Jackson, J. (2006). Creating Standalone Spatially Enabled Python

http://proceedings.esri.com/library/userconf/proc06/papers/papers/pap\_1091.

Schildt, H. (2001). *The Complete Java Reference: Java 2 Fourth Edition*. New York:

Singh,V P. & Frevert, D.K. (2010). Hydrologic Modeling Inventory. Texas A&M University

Sprague, R. H. & Carlson, E.D. (1982). Building effective Decision Support Systems.

Stocker, J., C., Prisloe, A.S. & Civco, D. (1999). Putting Geospatial Information into the

U.S. Congress. (1972). Federal Water Pollution Control Act (33 U.S.C. §1251 et seq. Accessed September 18, 2010. Available from http://epw.senate.gov/water.pdf Well House Consultants. (2006). Tk - Laying Out Your GUI with Frames, Pack And Grid.

http://www.wellho.net/mouth/787\_Tk-laying-out-your-GUI-with-frames-pack-

Wilkerson, G W., McAnally, W.H., Martin, J.L. Ballweber, J.A., Pevey, K.C., Diaz-Ramirez, J.

& Moore, A. (2010). Latis: a Spatial Decision Support System to Assess Low-Impact Site Development Strategies." *Advances in Civil Engineering* Volume 2010. Accessed

Applications using the Arcgis Geoprocessor. Accessed Dec. 8, 2010. Available from

and the Bureau of Reclamation. Accessed December 8, 2010. Available from

Hands of the Real Natural Resource Managers. *Proceedings of the 1999 ASPRS* 

Accessed January 25, 2011. Available from http://NetBeans.org/

Available from http:// www.python.org

urce/JavaSwing/JavaSwingCompoentsList.PNG

Available from http://www.tcl.tk/

December 8, 2011. Available from https://engineering.purdue.edu/~lthia/

2011. Available from

Osborne/McGraw-Hill.

http://hydrologicmodels.tamu.edu/

Accessed Feb. 1, 2011. Available from

March 18, 2011. Available from

and-grid.html

Englewood Cliffs, N.J.:Prentice-Hall, Inc.

*Annual Convention*, Portland, Oregon. 1070-1076.

http://www.hindawi.com/journals/ace/2010/810402/

pdf

Water is essential to life, as we know it. However, statistics reveal that, in 2000, one billion people lacked access to safe drinking water and 2.4 billion to adequate sanitation. To achieve United Nations target of reducing by half the proportion of people without sustainable access to safe drinking water by 2015, an additional 1.5 billion people would require access to some form of improved water supply by 2015, that is an additional 100 million people each year (or 274,000/day) until 2015.

Because water sources are limited, it is of paramount importance to keep its quality at the highest level possible. Threats to water are manifold, from industry to natural phenomena, and water quality assurance is a basic environmental issue involving from political to technical aspects and options, but it is obvious that no assessment of water quality is possible without a quantitative identification of some characteristics, a process commonly called water quality monitoring.

This chapter is an overview on water quality and on its monitoring. The text reflects he experience of the authors on the subject, presents some research and development results they obtained in the last decade and includes data gathered from different sources, namely from USEPA reports and North Caroline State University Water Quality Group documents. The text includes remarks about measuring techniques for different water quality parameters that result from the experience acquired by the authors in the implementation of several water quality measuring units. The last part of the chapter proposes architectures and intelligent signal processing techniques for distributed water quality monitoring networks.

#### **2. Water quality**

Water quality is commonly defined by its physical, chemical, biological and aesthetic (appearance and smell) characteristics. Water may be used for drinking, irrigating crops and watering stock, industrial processes, production of fish, shellfish and crustaceans, wildlife

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 27

**Physical:** temperature, turbidity and clarity, color, salinity, suspended solids, dissolved

**Chemical:** pH, dissolved oxygen, biological oxygen demand, nutrients (including nitrogen

Measurements of these indicators can be used to determine and monitor changes in water quality and to determine whether the quality of the water is suitable for the health of the

The design of water quality monitoring systems is a complex and specialized field. The range of indicators that can be measured is wide and other indicators may be adopted in the future. The cost of a monitoring system to assess them all would be prohibitive, so resources are usually directed towards assessing contaminants that are important for the local

The paragraphs that follow detail several aspects of these quantities, algae, bacteria and radiations excluded. The paper includes a short reference to systems for on-line, in-situ

Temperature is an important water parameter because it is an influence quantity for the generality of other water parameters and also because it determines many physical characteristics of a water body. In the winter, water's temperature-dependent density allows aquatic life to survive. Ice is formed at 0 ºC and thus remains at the top of the water body. Sun shining through the ice will serve to warm the water below slightly, keeping the temperature just above freezing. Water at 4 ºC is the densest, and will sink to the bottom and be replaced by lighter 1 - 3.9 ºC water. The continual process of heating and sinking keeps

In addition, temperate lakes stratify during the summer because of water's temperaturedependent density. Stratification prevents the mixing of oxygen and nutrients in the water body, and often encourages dissolved oxygen depletion. During the spring, stratification will break down allowing mixing of oxygen and nutrients. During the fall, the water body loses heat until its temperature is uniform at 4 ºC. Wind creates circulation, which distributes oxygen and nutrients throughout the water body (fall overturn). Eventually, the surface water layer falls below 4 ºC, becomes less dense, and remains at the surface. Ice will form if temperatures are low enough; otherwise, this upper layer will remain just above

Higher temperatures often exacerbate low dissolved oxygen level problems in lakes and reservoirs. High temperatures encourage the microbial breakdown of organic matter, a process that requires dissolved oxygen. Unfortunately, warm water naturally holds less

and phosphorus), organic and inorganic compounds (including toxicants)

**Aesthetic:** odors, taints, color, floating matter

environment or for a specific use of the water.

the water body from freezing entirely [1].

**Radioactive:** alpha, beta and gamma radiation emitters.

natural environment and the uses for which the water is required.

water quality monitoring and ends with a list of references.

**3. Water quality parameters and measuring techniques** 

0 ºC. Deeper water will remain roughly at 4 ºC until spring [1].

solids, sediment

**3.1 Temperature** 

habitats, protection of aquatic ecosystems, navigation and shipping, recreation (swimming, boating), and scientific study and education.

#### **2.1 Factors influencing water quality**

Water quality is closely linked to the surrounding environment and land use. Liquid water is never pure and is affected by agriculture, urban, industrial and recreation uses. The modification of natural stream flows and the weather can also have a major impact on water quality.

Groundwater is a major source of water and, when close to urban or industrial development, is vulnerable to contamination.

Generally, water quality of rivers is best in the headwaters, where rainfall is often abundant, declining as rivers flow through regions where land use and water use are intense and pollution from intensive agriculture, large towns, industry and recreation areas increases. There are of course exceptions to the rule and water quality may improve downstream, behind dams and weirs, at points where tributaries or better quality groundwater enter the mainstream, and in wetlands.

Rivers frequently act as conduits for pollutants by collecting and carrying wastewater from catchments and, ultimately, discharging it into the ocean. Storm water, which can also be rich in nutrients, organic matter and pollutants, finds its way into rivers and oceans mostly via the storm water drain network.

#### **2.2 Water quality and ecosystems**

An ecosystem is a community of organisms - plants, animals, fungi and bacteria - interacting with one another and with the environment in which they live. Protecting aquatic ecosystems is in many ways as important as maintaining water quality, for the following reasons:


#### **2.3 Water quality assessment**

The presence of contaminants and the characteristics of water are used to indicate the quality of water. These water quality indicators can be categorized as:

**Biological:** algae, bacteria

**Physical:** temperature, turbidity and clarity, color, salinity, suspended solids, dissolved solids, sediment

**Chemical:** pH, dissolved oxygen, biological oxygen demand, nutrients (including nitrogen and phosphorus), organic and inorganic compounds (including toxicants)

**Aesthetic:** odors, taints, color, floating matter

**Radioactive:** alpha, beta and gamma radiation emitters.

Measurements of these indicators can be used to determine and monitor changes in water quality and to determine whether the quality of the water is suitable for the health of the natural environment and the uses for which the water is required.

The design of water quality monitoring systems is a complex and specialized field. The range of indicators that can be measured is wide and other indicators may be adopted in the future. The cost of a monitoring system to assess them all would be prohibitive, so resources are usually directed towards assessing contaminants that are important for the local environment or for a specific use of the water.

The paragraphs that follow detail several aspects of these quantities, algae, bacteria and radiations excluded. The paper includes a short reference to systems for on-line, in-situ water quality monitoring and ends with a list of references.

#### **3. Water quality parameters and measuring techniques**

#### **3.1 Temperature**

26 Water Quality Monitoring and Assessment

habitats, protection of aquatic ecosystems, navigation and shipping, recreation (swimming,

Water quality is closely linked to the surrounding environment and land use. Liquid water is never pure and is affected by agriculture, urban, industrial and recreation uses. The modification of natural stream flows and the weather can also have a major impact on water

Groundwater is a major source of water and, when close to urban or industrial

Generally, water quality of rivers is best in the headwaters, where rainfall is often abundant, declining as rivers flow through regions where land use and water use are intense and pollution from intensive agriculture, large towns, industry and recreation areas increases. There are of course exceptions to the rule and water quality may improve downstream, behind dams and weirs, at points where tributaries or better quality groundwater enter the

Rivers frequently act as conduits for pollutants by collecting and carrying wastewater from catchments and, ultimately, discharging it into the ocean. Storm water, which can also be rich in nutrients, organic matter and pollutants, finds its way into rivers and oceans mostly

An ecosystem is a community of organisms - plants, animals, fungi and bacteria - interacting with one another and with the environment in which they live. Protecting aquatic ecosystems is in many ways as important as maintaining water quality, for the following

• Aquatic ecosystems are an integral part of our environment. They need to be maintained if the environment is to continue to support people. World conservation strategies stress the importance of maintaining healthy ecosystems and genetic

• Aquatic ecosystems play an important role in maintaining water quality and are a valuable indicator of water quality and the suitability of the water for other uses. • Aquatic ecosystems are valuable resources. Aquatic life is a major source of protein for humans. In most countries, like Portugal, commercial and sport fishing is economically

The presence of contaminants and the characteristics of water are used to indicate the

quality of water. These water quality indicators can be categorized as:

boating), and scientific study and education.

development, is vulnerable to contamination.

mainstream, and in wetlands.

via the storm water drain network.

**2.2 Water quality and ecosystems** 

**2.1 Factors influencing water quality** 

quality.

reasons:

diversity.

important.

**2.3 Water quality assessment** 

**Biological:** algae, bacteria

Temperature is an important water parameter because it is an influence quantity for the generality of other water parameters and also because it determines many physical characteristics of a water body. In the winter, water's temperature-dependent density allows aquatic life to survive. Ice is formed at 0 ºC and thus remains at the top of the water body. Sun shining through the ice will serve to warm the water below slightly, keeping the temperature just above freezing. Water at 4 ºC is the densest, and will sink to the bottom and be replaced by lighter 1 - 3.9 ºC water. The continual process of heating and sinking keeps the water body from freezing entirely [1].

In addition, temperate lakes stratify during the summer because of water's temperaturedependent density. Stratification prevents the mixing of oxygen and nutrients in the water body, and often encourages dissolved oxygen depletion. During the spring, stratification will break down allowing mixing of oxygen and nutrients. During the fall, the water body loses heat until its temperature is uniform at 4 ºC. Wind creates circulation, which distributes oxygen and nutrients throughout the water body (fall overturn). Eventually, the surface water layer falls below 4 ºC, becomes less dense, and remains at the surface. Ice will form if temperatures are low enough; otherwise, this upper layer will remain just above 0 ºC. Deeper water will remain roughly at 4 ºC until spring [1].

Higher temperatures often exacerbate low dissolved oxygen level problems in lakes and reservoirs. High temperatures encourage the microbial breakdown of organic matter, a process that requires dissolved oxygen. Unfortunately, warm water naturally holds less

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 29

Medium 0.0117 (not visible to the human eye)

**Sediment class Size (mm)** 

V. Coarse 1.5 Medium 0.375 V. Fine 0.094

V. Coarse 0.047

V. Fine 0.0049 Clay < 0.00195

Table 1. Size classification of sediments (adapted from [3]).

Aquatic Life < 50 NTU instantaneously or

< 10 NTU for trout waters or

demonstrate that this level does not interfere with:

2. maintenance of a disinfecting agent 3. microbiological determination [7]

**3.2.2 Measuring techniques [8]** 

of 0 to 100 NTU.

scattered by a reference solution [9].

< 25 NTU for a 10 day average [5]

disease. Inorganic constituents have no notable health effects.

Fish kills often result from extensive oxygen depletion.

< 25 NTU for streams (non-trout waters) or

< 50 NTU for lakes and reservoirs (non-trout waters) [6]

Human Consumption 1 to 5 NTU (up to 5 NTU is allowed if the water supplier can

Turbidity may be due to organic and/or inorganic constituents. Organic particulates may harbor micro organisms. Thus, turbid conditions may increase the possibility for waterborne

If turbidity is largely due to organic particles, dissolved oxygen depletion may occur in the water body. The excess nutrients available will encourage microbial breakdown, a process that requires dissolved oxygen. In addition, excess nutrients may result in algal growth. Although photosynthetic by day, algae respire at night, using valuable dissolved oxygen.

**Nephelometric Method:** Comparison of the light scattered by the sample and the light

• *Detection limits:* Should be able to detect turbidity differences of 0.02 NTU with a range

**Designated Use Acceptable Ranges**

Recreation 5 NTU [4]

Sand

Silt

**3.2.1 Numerical categories** 

1. disinfection

dissolved oxygen. Thus, persistent warm conditions may lead to a depletion of dissolved oxygen in the water body.

#### **3.1.1 Measuring techniques**

#### **3.1.1.1 Temperature probes**

Temperature range is usually from 0 to 30 ºC. Thus thermistor, platinum or even electronic based probes are adequate. Some manufacturers, like Quanta, commercializes probes that can measure several water parameters, including temperature (multi-parameter probes).

#### **3.2 Turbidity**

Turbidity is a quantity quantifying the degree to which light traveling through a water column is scattered by the suspended organic (including algae) and inorganic particles. Light scattering increases with the quantity of solids suspended in water. According to the research work developed by Campbell Scientific, usually the values of turbidity are correlated with the suspended solids concentration –SCC (Fig. 1); however, cases are also reported where no correlation between these two quantities is registered. Turbidity is commonly measured in Nephelometric Turbidity Units (NTU).

Fig. 1. The graph on the left provides measurements of runoff from a freeway, which indicates a bad correlation between SSC and turbidity. The graph on the right provides measurements from San Francisco Bay that indicates a good correlation between SSC and turbidity (Campbell Scientific document)

The velocity of the water resource largely determines the composition of the suspended load. Suspended loads are carried in both the gentle currents of lentic (lake) waters and the fast currents of lotic (flowing) waters. Even in flowing waters, the suspended load usually consists of grains less than 0.5 mm in diameter (Table 1). Suspended loads in lentic waters usually consist of the smallest sediment fractions, such as silt and clay [2].


Table 1. Size classification of sediments (adapted from [3]).

#### **3.2.1 Numerical categories**

28 Water Quality Monitoring and Assessment

dissolved oxygen. Thus, persistent warm conditions may lead to a depletion of dissolved

Temperature range is usually from 0 to 30 ºC. Thus thermistor, platinum or even electronic based probes are adequate. Some manufacturers, like Quanta, commercializes probes that can measure several water parameters, including temperature (multi-parameter probes).

Turbidity is a quantity quantifying the degree to which light traveling through a water column is scattered by the suspended organic (including algae) and inorganic particles. Light scattering increases with the quantity of solids suspended in water. According to the research work developed by Campbell Scientific, usually the values of turbidity are correlated with the suspended solids concentration –SCC (Fig. 1); however, cases are also reported where no correlation between these two quantities is registered. Turbidity is

Fig. 1. The graph on the left provides measurements of runoff from a freeway, which indicates a bad correlation between SSC and turbidity. The graph on the right provides measurements from San Francisco Bay that indicates a good correlation between SSC and

usually consist of the smallest sediment fractions, such as silt and clay [2].

The velocity of the water resource largely determines the composition of the suspended load. Suspended loads are carried in both the gentle currents of lentic (lake) waters and the fast currents of lotic (flowing) waters. Even in flowing waters, the suspended load usually consists of grains less than 0.5 mm in diameter (Table 1). Suspended loads in lentic waters

turbidity (Campbell Scientific document)

commonly measured in Nephelometric Turbidity Units (NTU).

oxygen in the water body.

**3.2 Turbidity** 

**3.1.1 Measuring techniques 3.1.1.1 Temperature probes** 


Human Consumption 1 to 5 NTU (up to 5 NTU is allowed if the water supplier can demonstrate that this level does not interfere with:


Turbidity may be due to organic and/or inorganic constituents. Organic particulates may harbor micro organisms. Thus, turbid conditions may increase the possibility for waterborne disease. Inorganic constituents have no notable health effects.

If turbidity is largely due to organic particles, dissolved oxygen depletion may occur in the water body. The excess nutrients available will encourage microbial breakdown, a process that requires dissolved oxygen. In addition, excess nutrients may result in algal growth. Although photosynthetic by day, algae respire at night, using valuable dissolved oxygen. Fish kills often result from extensive oxygen depletion.

#### **3.2.2 Measuring techniques [8]**

**Nephelometric Method:** Comparison of the light scattered by the sample and the light scattered by a reference solution [9].

• *Detection limits:* Should be able to detect turbidity differences of 0.02 NTU with a range of 0 to 100 NTU.

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 31

The presence in water of different anions and cations in different proportions leads to different values of water electric conductivity. However, even if it exists almost a linear relation between salinity and conductivity, that relation depends on the type of the dissolved salt. Moreover, conductivity is a non-selective measurement because instead of salinity, it gives the contribution of all charge carriers and not of a specific one. Notwithstanding, commonly, salinity is indirectly measured using conductivity meters.

As an example, Fig. 2 represents the experiment results of the relation that is obtained between TDS and conductivity for a variable amount of NaCl dissolved in water. In this case, the correlation coefficient between both variables is almost equal to 1, meaning that the

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

**experimental value**

**Conductivity (S/m)**

Fig. 2. Experiment results of the relation that is obtained between TDS and conductivity for different values of NaCl dissolved in water (square symbol: experimental data; circle

It is important to underline that the equation of the straight line that is represented in the graph, namely its slope, depends mainly on the type of salt and on the ionic activity of the

1. **Electric Conductivity (EC):** Uses a conductivity bridge calibrated with standard seawater solution. The solution's ability to transmit electricity is facilitated by increasing

elements that are dissolved in the water and on the temperature of the solution.

**y = 5.6695x - 0.0456** 

**cc=0.9994**

**theoretical value**

**3.3.2 Total dissolved solids and conductivity** 

relation between both variables is practically linear.

0

symbol: theoretical data).

**3.3.3 Measuring techniques [14][15]** 

2

4

6

8

10

12

**TDS (ppm)**

• *Interferences:* Rapidly settling coarse debris, dirty glassware, presence of air bubbles, and surface vibrations.

It is important to underline that turbidity is a measurement of the light scattering intensity relatively to the one that is obtained with the turbidity calibration standards. Visual clarity, measured as Secchi [10][11] or black disc visibility, is a direct measurement of the amount of suspended solids in water but its measurement requires more expensive equipments. Nevertheless, clarity measurements are more precise than turbidity measurements.

#### **3.3 Salinity**

The total dissolved solids (TDS) in water consist of inorganic salts and dissolved materials. In natural waters, salts are chemical compounds comprised of anions such as carbonates, chlorides, sulphates, and nitrates (primarily in ground water), and cations such as potassium (K), magnesium (Mg), calcium (Ca), and sodium (Na) [12]. In ambient conditions, these compounds are present in proportions that create a balanced solution. If there are additional inputs of dissolved solids to the system, the balance is altered and detrimental effects may be seen. Inputs include both natural and anthropogenic sources.

#### **3.3.1 Numerical categories**


\*Industry can de-ionize water to meet requirements; economics is the limiting factor [12].

Table 2. Salinity: designated use limits [13].

#### **3.3.2 Total dissolved solids and conductivity**

The presence in water of different anions and cations in different proportions leads to different values of water electric conductivity. However, even if it exists almost a linear relation between salinity and conductivity, that relation depends on the type of the dissolved salt. Moreover, conductivity is a non-selective measurement because instead of salinity, it gives the contribution of all charge carriers and not of a specific one. Notwithstanding, commonly, salinity is indirectly measured using conductivity meters.

As an example, Fig. 2 represents the experiment results of the relation that is obtained between TDS and conductivity for a variable amount of NaCl dissolved in water. In this case, the correlation coefficient between both variables is almost equal to 1, meaning that the relation between both variables is practically linear.

**TDS (ppm)**

30 Water Quality Monitoring and Assessment

• *Interferences:* Rapidly settling coarse debris, dirty glassware, presence of air bubbles, and

It is important to underline that turbidity is a measurement of the light scattering intensity relatively to the one that is obtained with the turbidity calibration standards. Visual clarity, measured as Secchi [10][11] or black disc visibility, is a direct measurement of the amount of suspended solids in water but its measurement requires more expensive equipments.

The total dissolved solids (TDS) in water consist of inorganic salts and dissolved materials. In natural waters, salts are chemical compounds comprised of anions such as carbonates, chlorides, sulphates, and nitrates (primarily in ground water), and cations such as potassium (K), magnesium (Mg), calcium (Ca), and sodium (Na) [12]. In ambient conditions, these compounds are present in proportions that create a balanced solution. If there are additional inputs of dissolved solids to the system, the balance is altered and detrimental

Nevertheless, clarity measurements are more precise than turbidity measurements.

effects may be seen. Inputs include both natural and anthropogenic sources.

**Designated Use (mg/l)** 

Irrigation 500-1 000 TDS (dependent upon crop sensitivity)

250 mg/l chloride

 Human Consumption 500 TDS 250 chloride 250 sulphate

 Light beer 500 TDS Dark beer 1000 TDS

 Fine paper 200 TDS Ground wood paper 500 TDS

Canning/Freezing 850 TDS

Table 2. Salinity: designated use limits [13].

Boiler feed water 50 to 3000 TDS depending on pressure

Aquatic Life Varies, depending on natural conditions

\*Industry can de-ionize water to meet requirements; economics is the limiting factor [12].

surface vibrations.

**3.3.1 Numerical categories** 

**3.3 Salinity** 

Industry \*

Brewing

Pulp and paper

Fig. 2. Experiment results of the relation that is obtained between TDS and conductivity for different values of NaCl dissolved in water (square symbol: experimental data; circle symbol: theoretical data).

It is important to underline that the equation of the straight line that is represented in the graph, namely its slope, depends mainly on the type of salt and on the ionic activity of the elements that are dissolved in the water and on the temperature of the solution.

#### **3.3.3 Measuring techniques [14][15]**

1. **Electric Conductivity (EC):** Uses a conductivity bridge calibrated with standard seawater solution. The solution's ability to transmit electricity is facilitated by increasing

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 33

**I/V conductance**

**reference voltage** 

Ferromagnetic Winding Core

**Rw**

(b)

Based on the equivalent circuit of the sensor, presented in Fig.4 (b), the relationship

<sup>1</sup> <sup>2</sup>

ω

1 11 1 1 *w*

*I j C U nR j L*

 = ++

ω

Fig. 4. (a) Inductive sensor with a single ferromagnetic core; (b) Equivalent electrical circuit

between the applied voltage, u1, and the sensor output current, i, is:

⇔

(a)

Current lines

**u1**

**i**

**n1 2n R1 w**

(2)

**<sup>C</sup> L1**

**I**

**V**

(a) (b)

Fig. 3. Four-electrode conductivity cell: (a) measuring circuit and (b) cell structure.

**current source**

**n1 n2=1**

**i**

**C**

**L1**

**u1**

(Rw- water resistance).

**voltage measurement**

**current measurement**

**R**

**conductivity cell** 

salt content. The EC is normally measured in mhos/cm, mmho/cm, or umho/cm (non-SI units) or siemens per meter (S/m) (SI units), depending on dissolved salt concentration.

The conversion of EC values into the total quantity of dissolved salts depends on the dissolved salts and manufacturers of this type of meter use conversion factors such that:

$$\text{TDS (mg/l)} = 50 \text{ to } 70^{\circ} \text{EC (S/m)} = 500 \text{ to } 700^{\circ} \text{EC (mmhto/cm)}.$$

Note that, in this context, 1 mg/l is approximately equal to 1 ppm and that these conversion factors are some ten times higher than the value obtained from Fig. 2.

A simple sensor for water (or any liquid) electrical conductivity measurement is the two electrode cell [16][17]. The current that flows between two electrodes immersed in the water, when a voltage is applied between them, is a function of all dissolved ionized solids in the solution. Theoretical results of conductivity, σ, are determined from the sensor measured conductance (1/R), using a geometric coefficient or "cell constant" (KC) that reflects the ratio between the length (d) and the cross-section area (A) of the sampled water volume in which the electrical current actually flows:

$$
\sigma = \frac{d}{A} \frac{1}{R} = K\_c \frac{1}{R} \tag{1}
$$

The cell and associated conditioning circuits are projected to reduce electrode-water interface contributions to the measurement. For instance, almost invariably, alternating current is used in order to minimize the polarization effects [18]. Notwithstanding, these effects cause a nonlinear variation of the cell constant KC, that can only be compensated by calibration. Thus, it is not convenient to consider KC as the geometric value of the ratio d/A for the cell, but the value obtained experimentally using solutions of accurate known conductivity.

For high conductance measurements, when coating and electrodes fouling are a concern, the four-electrode based conductivity measuring circuit, as the one presented in Fig. 3, is a more suitable solution [19]. Current is imposed across two drive electrodes and the other two sense electrodes are used for a null-current voltage measurement. Polarization at the drive electrodes has no effect on the measurement, provided the drive voltage is able to maintain the control current through the cell.

Another way to eliminate polarization effects and to obtain contact less measurements is to use an inductive sensing structure [20]. The basic principle consists in the presence of eddy currents induced in the interior of the body under test. These currents are due to the time variation of the magnetic flux originated by the primary coil current. The intensity of the induced currents is related to the electric conductivity of the salty water and there is a correlation between the electrical parameters that can be assessed and the medium conductivity to be determined. In Fig. 4 the simplest form of inductive conductivity sensor is depicted. The sensor is shunted with a capacitance C. It is a transformer whose secondary coil is the surrounding liquid. The primary current phasor, *I* , is the sensor signal output.

concentration.

salt content. The EC is normally measured in mhos/cm, mmho/cm, or umho/cm (non-SI units) or siemens per meter (S/m) (SI units), depending on dissolved salt

The conversion of EC values into the total quantity of dissolved salts depends on the dissolved salts and manufacturers of this type of meter use conversion factors such that:

Note that, in this context, 1 mg/l is approximately equal to 1 ppm and that these conversion factors are some ten times higher than the value obtained from Fig. 2.

A simple sensor for water (or any liquid) electrical conductivity measurement is the two electrode cell [16][17]. The current that flows between two electrodes immersed in the water, when a voltage is applied between them, is a function of all dissolved ionized solids in the solution. Theoretical results of conductivity, σ, are determined from the sensor measured conductance (1/R), using a geometric coefficient or "cell constant" (KC) that reflects the ratio between the length (d) and the cross-section area (A) of the

> 1 1 *C <sup>d</sup> <sup>K</sup> AR R*

The cell and associated conditioning circuits are projected to reduce electrode-water interface contributions to the measurement. For instance, almost invariably, alternating current is used in order to minimize the polarization effects [18]. Notwithstanding, these effects cause a nonlinear variation of the cell constant KC, that can only be compensated by calibration. Thus, it is not convenient to consider KC as the geometric value of the ratio d/A for the cell, but the value obtained experimentally using

For high conductance measurements, when coating and electrodes fouling are a concern, the four-electrode based conductivity measuring circuit, as the one presented in Fig. 3, is a more suitable solution [19]. Current is imposed across two drive electrodes and the other two sense electrodes are used for a null-current voltage measurement. Polarization at the drive electrodes has no effect on the measurement, provided the

Another way to eliminate polarization effects and to obtain contact less measurements is to use an inductive sensing structure [20]. The basic principle consists in the presence of eddy currents induced in the interior of the body under test. These currents are due to the time variation of the magnetic flux originated by the primary coil current. The intensity of the induced currents is related to the electric conductivity of the salty water and there is a correlation between the electrical parameters that can be assessed and the medium conductivity to be determined. In Fig. 4 the simplest form of inductive conductivity sensor is depicted. The sensor is shunted with a capacitance C. It is a transformer whose secondary coil is the surrounding liquid. The primary current

= = (1)

sampled water volume in which the electrical current actually flows:

σ

drive voltage is able to maintain the control current through the cell.

solutions of accurate known conductivity.

phasor, *I* , is the sensor signal output.

TDS (mg/l) = 50 to 70\*EC (S/m) = 500 to 700\*EC (mmho/cm).

Fig. 3. Four-electrode conductivity cell: (a) measuring circuit and (b) cell structure.

Fig. 4. (a) Inductive sensor with a single ferromagnetic core; (b) Equivalent electrical circuit (Rw- water resistance).

Based on the equivalent circuit of the sensor, presented in Fig.4 (b), the relationship between the applied voltage, u1, and the sensor output current, i, is:

$$
\overline{I} = \left(\frac{1}{n\_1^2 R\_w} + \frac{1}{j\alpha \mathcal{L}\_{11}} + j\alpha \mathcal{C}\right) \overline{\mathcal{U}}\_1 \tag{2}
$$

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 35

1. **Extrabasinal (terrigenous) particles:** Terrigenous particles have been eroded from the land outside the body of water experiencing the deposition. The particles either retain

3. **Pyroclastic particles:** These particles are derived during the explosive action of a volcano. Particles include rock fragment, single crystals, or bits of volcanic glass. 4. **Intrabasinal particles:** These particles grow biochemically or chemically in the waters experiencing deposition. These particles include carbonate biocrystals, silica biocrystals, particles composed of evaporated minerals, and minerals that grow at the

Sediment introduced into surface water is either deposited on the bed of the stream or lake or suspended in the water column (suspended load). Bed load is large sediment particles that move by bouncing along the bottom. Generally, the suspended loads in lotic (flowing) water consist of grains less than 0.5 mm in diameter [2]. Lentic (lake) suspended loads usually consist of the smallest sediment fractions, such as silt and clays [2]. A water body's suspended load is a component of the total turbidity. Any sediment transported by water is

The amount of sediment deposited on a rocky substrate can be quantitatively defined by an estimation of the percent embeddedness. The percent embeddedness is the degree to which fine sediments such as sand, silt, and clay fill the interstitial spaces between rocks on a

Aquatic life [22]

Industry (total solids) [4]

their chemical make-up, or become chemically altered to clays and iron oxides. 2. **Carbonaceous particles:** These particles are organic in nature and are derived from either solid carbonaceous material (coal, amber, wax, and kerogen), reworked from

other geologic formations, or from modern plant detritus.

water/sediment interface.

substrate.

**3.4.1 Numerical categories** 

subject to deposition as flow velocity decreases [21].

Acceptable Ranges to Maintain Designated Use

**Optimal Ranges Designated Use** 

 Boiler Feedwater 500 – 3000 mg/l 0 - 150 psi 500 – 2500 mg/l 150 - 250 psi 100 – 1500 mg/l 250 - 400 psi 50 mg/l > 400 psi

100 mg/l Pulp Production

Table 4. Sediment: designated use and optimal ranges.

200 mg/l Photographic Processing 200 mg/l Clear Plastic Production

< 25% embeddedness Excellent Conditions 25 - 50% embeddedness Good Conditions 50 - 75% embeddedness Fair Conditions > 75% embeddedness Poor Conditions

where n1 is the number of windings, L11 the inductance of the coil, C is the capacitance of the shunt capacitor, ω=2πf and f the frequency. By choosing C so that ωC=1/ωL, the quadrature component of the current is null and the primary current phasor becomes an exact measure of RW ( ) <sup>2</sup> *I U nR* = 1 1 / *<sup>W</sup>* .

Finally, it is important to underline that conductivity depends on temperature, exhibiting a dependence coefficient approximately equal to 2 %/ºC for salt waters. Hence, it is necessary to normalize the conductivity measurements to a reference temperature, usually equal to 25 ºC.


Interferences: Presence of aluminium, calcium, iron, manganese, silica, strontium, and suspended matter might interfere with test. Solution should not contain more than 3.5 g NH4Cl.

#### **3.4 Sediment**

Sediment is composed of organic and inorganic particles of various sizes. The major classes of sediment, from largest to smallest, are boulders, cobbles, pebbles, sand, silt, and clay [3],


(Adapted from [3]).

Table 3. Sediment classes.

Sediments are classified into four broad categories, according to their origin in relation to the basin of water in which they are deposited: extrabasinal, carbonaceous, pyroclastic, and intrabasinal.


Sediment introduced into surface water is either deposited on the bed of the stream or lake or suspended in the water column (suspended load). Bed load is large sediment particles that move by bouncing along the bottom. Generally, the suspended loads in lotic (flowing) water consist of grains less than 0.5 mm in diameter [2]. Lentic (lake) suspended loads usually consist of the smallest sediment fractions, such as silt and clays [2]. A water body's suspended load is a component of the total turbidity. Any sediment transported by water is subject to deposition as flow velocity decreases [21].

The amount of sediment deposited on a rocky substrate can be quantitatively defined by an estimation of the percent embeddedness. The percent embeddedness is the degree to which fine sediments such as sand, silt, and clay fill the interstitial spaces between rocks on a substrate.

#### **3.4.1 Numerical categories**

34 Water Quality Monitoring and Assessment

an exact measure of RW ( ) <sup>2</sup> *I U nR* = 1 1 / *<sup>W</sup>* .

temperature, usually equal to 25 ºC.

dried and weighed.

V. Coarse 1.5 Medium 0.375 V.Fine 0.094

V. Coarse 0.047

NH4Cl.

**3.4 Sediment** 

Sand

Silt

V.Fine Clay

(Adapted from [3]).

intrabasinal.

Table 3. Sediment classes.

2. **Density Method:** Uses a precise vibrating flow densimeter.

where n1 is the number of windings, L11 the inductance of the coil, C is the capacitance of the shunt capacitor, ω=2πf and f the frequency. By choosing C so that ωC=1/ωL, the quadrature component of the current is null and the primary current phasor becomes

Finally, it is important to underline that conductivity depends on temperature, exhibiting a dependence coefficient approximately equal to 2 %/ºC for salt waters. Hence, it is necessary to normalize the conductivity measurements to a reference

3. **Gravimetric Method** [8]: As an example, magnesium is measured using the gravimetric method. Diammonium hydrogen phosphate precipitates magnesium in ammonical solution as magnesium ammonium phosphate. The test can be performed two ways. First, the ammonium salts and oxalate can be destroyed, followed by precipitation of magnesium ammonium phosphate. Second, the diammonium hydrogen phosphate can undergo double precipitation without pre-treatment (preferable option). Sample is then

Interferences: Presence of aluminium, calcium, iron, manganese, silica, strontium, and suspended matter might interfere with test. Solution should not contain more than 3.5 g

Sediment is composed of organic and inorganic particles of various sizes. The major classes of sediment, from largest to smallest, are boulders, cobbles, pebbles, sand, silt, and clay [3],

Class Size (mm) Approx size Boulders > 256 > Volleyball Cobbles > 64 > Tennis ball Pebbles > 2 > Match head

Medium 0.0117 (no longer visible to the human eye)

Sediments are classified into four broad categories, according to their origin in relation to the basin of water in which they are deposited: extrabasinal, carbonaceous, pyroclastic, and

0.0049 < 0.00195 Acceptable Ranges to Maintain Designated Use


Table 4. Sediment: designated use and optimal ranges.

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 37

The rate of sediment-induced storage loss in lakes and reservoirs can be measured with a series of sedimentation or bathymetric surveys. Transects and sampling points are often established perpendicular to the main axis of tributary inflow or the main axis of the lake. Some base strata (original lake or reservoir bottom) should be established to track the rate (cm/yr) of deposition. A long pole may be used in shallow areas to measure sediment depth. For deeper areas, sonar or a SCUBA diver can record sediment depth measurements. Although measurements every year may not be needed, the same transects and stations should be monitored periodically. Major changes in land use, best management practices, or stream bank erosion that could increase sedimentation should also be monitored [25].

Sediment accumulation (deposition rates) can also be determined by measuring radionuclides in the sediment [26]. Cesium-137, a fallout product of nuclear testing, binds tightly to soil particles and can be used to estimate the time of sediment deposition. Measurable levels of Cesium-137 were introduced into the atmosphere during the beginning

Lead-210 may also be used to measure sedimentation rates [26]. Lead-210 is a naturallyoccurring uranium isotope that is a decay product of radon. When atmospheric radon decays, the lead-210 is deposited on the earth's surface. Lead-210 will bind to soil particles

As long as the hydrogen ion concentration is not too high (less than 10-3 M), this activity is

[aH+]=γ⋅[H+] (4)

This coefficient is always lower than one and is almost equal to one for low hydrogen ion

The pH is a log-base 10 scale that measures acidity of a solution on a scale of 0 to 14. The pH of neutral solutions, such as pure water, is equal to 7. Alkaline solutions will have high pHs

Since the pH is a log-base-10 scale, the pH changes 1 unit for every power of ten changes in [H+]. For example, water with a pH of 3 has 100 times the amount of [H+] that is found in a pH 5 water. Because pH = - log10 [H+], the pH will decrease as the [H+] increases [27].

Although near-neutral pH values are preferred, industry as a whole can tolerate a wide pH range, depending on the intended water use. The Environmental Protection Agency reports

pH = - log10 [aH+] (3)

of the nuclear age and can only be used to estimate deposition after 1957.

and can be used to measure sedimentation rates for the past 100 years [26].

approximately proportional to the hydrogen ion concentration, more precisely:

pH is defined as the negative log-base 10 of the hydrogen ion activity:

where γ represents the activity coefficient.

(8-14) and acidic solutions will have low pHs (1-6).

**3.5 pH** 

concentration.

**3.5.1 Numerical categories** 

**3.4.2.3 Assessment of capacity loss due to sedimentation** 

#### **3.4.2 Measuring techniques**

#### **3.4.2.1 Total Suspended Solids (TSS) sampling technique**

A number of different methods are available for sampling suspended sediment in streams. Two types of suspended sediment samplers are available for perennial streams: depthintegrating and point-integrating. Both samplers are usually made from cast aluminium or bronze and have a tail fin to orient the sampler's intake nozzle upstream. The depthintegrating sampler is designed to sample continuously as it is lowered at a constant speed from the water surface to the stream bed and back. The point-integrating sampler is equipped with a mechanism at the end of the sampler that can open and collect a sample at a specified depth in the stream [23].

Ephemeral streams are sampled using rising-stage samplers. These samplers have a number of bottles arranged on top of one another in a frame. Each bottle is equipped with kinked tubing pointed into the flow and will collect a sample as the water rises. This sampler is better suited for sampling silts and clays because the intake flow velocity is slower than the stream velocity and larger particles might settle before entering the sampler [23].

Many automated samplers are also available. They differ in cost, ease of maintenance, run time, and ability to extract a representative sample. The samplers extract samples by pumping water from the stream and retaining some in a sample bottle. The samplers can be set to extract samples at set intervals or can be attached to a float that will automatically trigger the sampling process when a pre-selected stream stage is reached [23].

Finally, suspended sediment concentration is often monitored with a turbidity meter. Although turbidity can be influenced by other factors such as size distribution, shape, and absorptivity of the sediment, and the color of the water, turbidity meters give satisfactory estimates [23].

#### **3.4.2.2 Percent embeddedness technique**

The amount of sediment deposited on a rocky substrate can be quantitatively defined by an estimation of the percent embeddedness. This method is only applicable on substrates of coarse pebbles, cobbles, or rubble. Percent embeddedness is measured at transect points of an area of known size (e.g. a square 0.5m X 0.5m). The percentages normally assigned to various substrate conditions are [24]:


#### **3.4.2.3 Assessment of capacity loss due to sedimentation**

The rate of sediment-induced storage loss in lakes and reservoirs can be measured with a series of sedimentation or bathymetric surveys. Transects and sampling points are often established perpendicular to the main axis of tributary inflow or the main axis of the lake. Some base strata (original lake or reservoir bottom) should be established to track the rate (cm/yr) of deposition. A long pole may be used in shallow areas to measure sediment depth. For deeper areas, sonar or a SCUBA diver can record sediment depth measurements. Although measurements every year may not be needed, the same transects and stations should be monitored periodically. Major changes in land use, best management practices, or stream bank erosion that could increase sedimentation should also be monitored [25].

Sediment accumulation (deposition rates) can also be determined by measuring radionuclides in the sediment [26]. Cesium-137, a fallout product of nuclear testing, binds tightly to soil particles and can be used to estimate the time of sediment deposition. Measurable levels of Cesium-137 were introduced into the atmosphere during the beginning of the nuclear age and can only be used to estimate deposition after 1957.

Lead-210 may also be used to measure sedimentation rates [26]. Lead-210 is a naturallyoccurring uranium isotope that is a decay product of radon. When atmospheric radon decays, the lead-210 is deposited on the earth's surface. Lead-210 will bind to soil particles and can be used to measure sedimentation rates for the past 100 years [26].

#### **3.5 pH**

36 Water Quality Monitoring and Assessment

A number of different methods are available for sampling suspended sediment in streams. Two types of suspended sediment samplers are available for perennial streams: depthintegrating and point-integrating. Both samplers are usually made from cast aluminium or bronze and have a tail fin to orient the sampler's intake nozzle upstream. The depthintegrating sampler is designed to sample continuously as it is lowered at a constant speed from the water surface to the stream bed and back. The point-integrating sampler is equipped with a mechanism at the end of the sampler that can open and collect a sample at

Ephemeral streams are sampled using rising-stage samplers. These samplers have a number of bottles arranged on top of one another in a frame. Each bottle is equipped with kinked tubing pointed into the flow and will collect a sample as the water rises. This sampler is better suited for sampling silts and clays because the intake flow velocity is slower than the stream velocity and larger particles might settle before entering the

Many automated samplers are also available. They differ in cost, ease of maintenance, run time, and ability to extract a representative sample. The samplers extract samples by pumping water from the stream and retaining some in a sample bottle. The samplers can be set to extract samples at set intervals or can be attached to a float that will automatically

Finally, suspended sediment concentration is often monitored with a turbidity meter. Although turbidity can be influenced by other factors such as size distribution, shape, and absorptivity of the sediment, and the color of the water, turbidity meters give satisfactory

The amount of sediment deposited on a rocky substrate can be quantitatively defined by an estimation of the percent embeddedness. This method is only applicable on substrates of coarse pebbles, cobbles, or rubble. Percent embeddedness is measured at transect points of an area of known size (e.g. a square 0.5m X 0.5m). The percentages normally assigned to

• *100% embeddedness*= Rocks are completely surrounded by sediment and completely

• *75% embeddedness*= Rocks are completely surrounded by sediment and half covered by

• *50% embeddedness*= Rocks are completely surrounded by sediment but are not covered

• *25% embeddedness*= Rocks are half surrounded by sediment and are not covered by

trigger the sampling process when a pre-selected stream stage is reached [23].

**3.4.2 Measuring techniques** 

a specified depth in the stream [23].

**3.4.2.2 Percent embeddedness technique**

various substrate conditions are [24]:

• *0% embeddedness*= No fine sediments on substrate.

covered by sediment.

sediment.

sediment.

by sediment.

sampler [23].

estimates [23].

**3.4.2.1 Total Suspended Solids (TSS) sampling technique** 

pH is defined as the negative log-base 10 of the hydrogen ion activity:

$$\mathbf{h}\mathbf{pH} = -\log\_{10}\left[\mathbf{a}\_{\mathrm{H}\*}\right] \tag{3}$$

As long as the hydrogen ion concentration is not too high (less than 10-3 M), this activity is approximately proportional to the hydrogen ion concentration, more precisely:

$$[\mathbf{a}\_{\rm It^{\ast}}] \overline{\boldsymbol{\simeq}} \boldsymbol{\mu} [\![\![\mathbf{H}^{+}]\!]\!] \tag{4}$$

where γ represents the activity coefficient.

This coefficient is always lower than one and is almost equal to one for low hydrogen ion concentration.

The pH is a log-base 10 scale that measures acidity of a solution on a scale of 0 to 14. The pH of neutral solutions, such as pure water, is equal to 7. Alkaline solutions will have high pHs (8-14) and acidic solutions will have low pHs (1-6).

Since the pH is a log-base-10 scale, the pH changes 1 unit for every power of ten changes in [H+]. For example, water with a pH of 3 has 100 times the amount of [H+] that is found in a pH 5 water. Because pH = - log10 [H+], the pH will decrease as the [H+] increases [27].

#### **3.5.1 Numerical categories**

Although near-neutral pH values are preferred, industry as a whole can tolerate a wide pH range, depending on the intended water use. The Environmental Protection Agency reports

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 39

1. **Electronic pH Meter:** A probe containing an acidic aqueous solution encased in a special glass membrane allows migration of hydrogen ions (H+). If the water has a pH

( )( ) 2,3R T 273,15 <sup>E</sup> 7 pH nF

Usually, electrochemical sensors are used to measure pH. According to the Nernst equation, the voltage difference (E) that is obtained from these sensors is given by:

where E represents the voltage difference between measuring and reference electrodes, T represents temperature in ºC, F represents the Faraday constant and n represents the number of electrodes that participate in the REDOX reaction. This value is equal to one

Fig. 5 represents a pH measuring diagram based on a glass membrane pH sensor and its equivalent electrical circuit. The electrical circuit contains multiple resistors and voltage sources that are associated with all resistive and polarization effects existing

> **Measuring electrode**

<sup>Δ</sup><sup>E</sup> **r.e. m.e.**

<sup>+</sup> = − <sup>−</sup> (5)

2 3 2H O OH H O ↔ +− + (6)

<sup>+</sup> E5

<sup>+</sup> E4

R5

R4

**r.e. m.e.**

R1

<sup>+</sup> E1

R2

<sup>+</sup> E2

R3

<sup>+</sup> E3

different from that of the solution within the probe, an electric potential results.

**3.5.2 Measuring techniques [27]** 

for the auto-ionization reaction of water:

Eléctrodos metálicos

**Glass membrane**

Fig. 5. Structure of a glass membrane pH sensor and its equivalent electrical circuit.

product of the water [32][33] with temperature that is approximately given by:

Regarding temperature effects, pH measurements must be compensated for the explicit effect of temperature defined in Nernst equation and also for the variation of the ionic

() ()<sup>2</sup> 14.00 0.0331 25 T 0.00017 25 T 2 2 K 10 <sup>W</sup> mol .l − − −+ ⋅ − <sup>−</sup> <sup>=</sup> (7)

**Metalic electrodes (Ag/AgCl)** 

**reference solution**

ΔE

inside the measuring cell.

**reference electrode**

**liquid junction**


that the widest pH range is between 3.0 and 11.7 for process waters, and between 5.0 and 8.9 for cooling waters. Specific industries will require more limited ranges. An industry can usually prepare water of the proper pH to meet its needs [12].

Table 5. pH ranges for different designated use.

A reduction in pH (more acidic) may allow the release of toxic metals that would otherwise be absorbed to sediment and essentially removed from the water system. Once mobilized, these metals are available for uptake by organisms. For many metals, the rate of uptake is directly proportional to the levels of metal availability in the environment. Thus, a decrease in pH increases metal availability, lending itself to greater metal uptake by organisms. Metal uptake can cause extreme physiological damage to aquatic life [29].

An increase in pH may cause heightened ammonia concentrations [12]. At low pH, ammonia combines with water (H2O) to produce an ammonium ion (NH4+) and a hydroxide ion (OH-). The ammonium ion is non-toxic and not of concern to organisms. Above a pH of 9, ammonia (un-ionized) is the predominant species [30]. The un-ionized ammonia (NH3) is very toxic to organisms. Thus, organisms experience ammonia toxicity more readily at higher pH [31].

Experiments have shown that a pH decrease of 1.4 units of pH can disturb the aquatic community.

#### **3.5.2 Measuring techniques [27]**

38 Water Quality Monitoring and Assessment

that the widest pH range is between 3.0 and 11.7 for process waters, and between 5.0 and 8.9 for cooling waters. Specific industries will require more limited ranges. An industry can

usually prepare water of the proper pH to meet its needs [12].

**Optimal pH Ranges Designated Use** 

6.0 -8.5 General Agriculture [28]

6.8 - 8.5 Dairy Sanitation 4.5 - 9.0 Irrigation water [12] 5.0 - 9.0 Human Consumption [28] 6.5 - 9.0 Freshwater aquatic life 6.5 - 8.5 Marine aquatic life [12] Industry [28] > 8.0 Boiler Feedwater

6.5 - 7.0 Brewery 6.5 - 7.5 Cooling Water > 7.5 Cannery 6.0 - 6.8 Laundering

> 7.0 Oil Well Flooding 7.8 - 8.3 Rayon Manufacturing 6.8 - 7.0 Steel Manufacturing

A reduction in pH (more acidic) may allow the release of toxic metals that would otherwise be absorbed to sediment and essentially removed from the water system. Once mobilized, these metals are available for uptake by organisms. For many metals, the rate of uptake is directly proportional to the levels of metal availability in the environment. Thus, a decrease in pH increases metal availability, lending itself to greater metal uptake by organisms. Metal

An increase in pH may cause heightened ammonia concentrations [12]. At low pH, ammonia combines with water (H2O) to produce an ammonium ion (NH4+) and a hydroxide ion (OH-). The ammonium ion is non-toxic and not of concern to organisms. Above a pH of 9, ammonia (un-ionized) is the predominant species [30]. The un-ionized ammonia (NH3) is very toxic to organisms. Thus, organisms experience ammonia toxicity

Experiments have shown that a pH decrease of 1.4 units of pH can disturb the aquatic

6.8 - 8.0 Tanning

uptake can cause extreme physiological damage to aquatic life [29].

Table 5. pH ranges for different designated use.

more readily at higher pH [31].

community.

1. **Electronic pH Meter:** A probe containing an acidic aqueous solution encased in a special glass membrane allows migration of hydrogen ions (H+). If the water has a pH different from that of the solution within the probe, an electric potential results.

Usually, electrochemical sensors are used to measure pH. According to the Nernst equation, the voltage difference (E) that is obtained from these sensors is given by:

$$\mathbf{E} = -\frac{2\beta \mathbf{R} \left(\mathbf{T} + 273.15\right)}{\mathbf{nF}} \text{(7-pH)}\tag{5}$$

where E represents the voltage difference between measuring and reference electrodes, T represents temperature in ºC, F represents the Faraday constant and n represents the number of electrodes that participate in the REDOX reaction. This value is equal to one for the auto-ionization reaction of water:

$$\text{CH}\_2\text{O} \leftrightarrow \text{OH}^- + \text{H}\_3\text{O}^\* \tag{6}$$

Fig. 5 represents a pH measuring diagram based on a glass membrane pH sensor and its equivalent electrical circuit. The electrical circuit contains multiple resistors and voltage sources that are associated with all resistive and polarization effects existing inside the measuring cell.

Fig. 5. Structure of a glass membrane pH sensor and its equivalent electrical circuit.

Regarding temperature effects, pH measurements must be compensated for the explicit effect of temperature defined in Nernst equation and also for the variation of the ionic product of the water [32][33] with temperature that is approximately given by:

$$\mathbf{K}\_{\mathrm{W}} = 10^{-14.00 - 0.0531(25 - \mathrm{I}) + 0.00017 \left(25 - \mathrm{I}\right)^{2}} \mathrm{mol}^{2} \,\mathrm{I}^{-2} \tag{7}$$

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 41

*Criteria to maintain designated use*

Warm water fish 5.0 Cold water fish 6.0 Spawning season 7.0 Estuarine biota 5.0

Primary Contact 3.0 Secondary Contact 3.0

High Pressure 0 Low Pressure 0.1 - 1.4

Table 6. Dissolved oxygen designated use and lowest acceptable levels.

\* Summary of state standards.

*Preferred ranges for designated use*

**Designated Use Ranges (mg/l DO)** 

1. **Iodometric Method:** Most reliable method. Requires the addition of divalent manganese solution, followed by a strong alkali, in a stoppered glass bottle. The dissolved oxygen oxidizes to manganous hydroxide precipitate. With the addition of iodide, the oxidized manganese reverts back to divalent state, releasing iodide equivalent to original dissolved oxygen content. Iodide is titrated with thiosulphate. • *Interferences:* Oxidizing agents may release iodines from iodides (false positive). Reducing agents may reduce iodine to iodide (false negative). Air entrapped in the

2. **Azide Method:** This method is suitable for samples containing more than 50 ug/l nitrite

3. **Permanganate Modification:** This method is suitable for samples containing ferrous

• *Interferences:* Oxidizing and reducing agents may provide false positives or

• *Interferences:* High concentrations of ferric iron will interfere. Can be overcome by

**Designated Use Lowest acceptable DO levels (mg/l)\*** 

**3.6.1 Numerical categories** 

 Aquatic life

Recreation

 Industry Boiler Feed Water

sample bottle will also interfere.

and not more than 1 mg/l of ferrous iron.

the addition of 1 ml KF.

**3.6.2 Measuring techniques** 

negatives.

iron.

where T represents the temperature in º C.

Finally, it is important to underline that because the internal resistance (RS) is very high, sometimes higher than 500 MΩ, an accurate measurement of the low voltage delivered by the pH transducer requires an amplifier with a very high input resistance. Usually the maximum bias currents of the amplifier must be lower than a few pA. As an example, Fig. 6 represents an equivalent electrical circuit of a signal conditioner that can be used for such a pH sensor.

Fig. 6. Signal conditioning for a glass membrane pH sensor.

2. **Electronic devices:** a semiconductor device, usually a JFET or MOSFET, whose electrical characteristics change with the hydrogen ion concentration.

#### **3.6 Dissolved oxygen**

Dissolved oxygen (DO) refers to the volume of oxygen that is contained in water. Oxygen enters the water by photosynthesis of aquatic biota and by the transfer of oxygen across the air-water interface. The amount of oxygen that can be held by the water depends on the water temperature, salinity, and pressure. Gas solubility increases with decreasing temperature (colder water holds more oxygen). Gas solubility increases with decreasing salinity (freshwater holds more oxygen than does saltwater). Both the partial pressure and the degree of saturation of oxygen will change with altitude. Finally, gas solubility decreases as pressure decreases. Thus, the amount of oxygen absorbed in water decreases as altitude increases because the atmospheric pressure decreases [1].

Microbes play a key role in the loss of oxygen from surface waters because they use oxygen as energy to break down long-chained organic molecules into simpler, more stable endproducts such as carbon dioxide, water, phosphate and nitrate [2]. If high levels of organic matter are present in water, microbes may use all available oxygen.

#### **3.6.1 Numerical categories**

40 Water Quality Monitoring and Assessment

Finally, it is important to underline that because the internal resistance (RS) is very high, sometimes higher than 500 MΩ, an accurate measurement of the low voltage delivered by the pH transducer requires an amplifier with a very high input resistance. Usually the maximum bias currents of the amplifier must be lower than a few pA. As an example, Fig. 6 represents an equivalent electrical circuit of a signal conditioner that can be used for

Coaxial cable

where T represents the temperature in º C.

10M

Fig. 6. Signal conditioning for a glass membrane pH sensor.

increases because the atmospheric pressure decreases [1].

matter are present in water, microbes may use all available oxygen.

10 MΩ

10M

10 MΩ

**pH cell** 

+VCC

2. **Electronic devices:** a semiconductor device, usually a JFET or MOSFET, whose

Dissolved oxygen (DO) refers to the volume of oxygen that is contained in water. Oxygen enters the water by photosynthesis of aquatic biota and by the transfer of oxygen across the air-water interface. The amount of oxygen that can be held by the water depends on the water temperature, salinity, and pressure. Gas solubility increases with decreasing temperature (colder water holds more oxygen). Gas solubility increases with decreasing salinity (freshwater holds more oxygen than does saltwater). Both the partial pressure and the degree of saturation of oxygen will change with altitude. Finally, gas solubility decreases as pressure decreases. Thus, the amount of oxygen absorbed in water decreases as altitude

Microbes play a key role in the loss of oxygen from surface waters because they use oxygen as energy to break down long-chained organic molecules into simpler, more stable endproducts such as carbon dioxide, water, phosphate and nitrate [2]. If high levels of organic

electrical characteristics change with the hydrogen ion concentration.


+VCC

such a pH sensor.

**3.6 Dissolved oxygen** 

E

+


Table 6. Dissolved oxygen designated use and lowest acceptable levels.

#### **3.6.2 Measuring techniques**

	- *Interferences:* Oxidizing agents may release iodines from iodides (false positive). Reducing agents may reduce iodine to iodide (false negative). Air entrapped in the sample bottle will also interfere.
	- *Interferences:* Oxidizing and reducing agents may provide false positives or negatives.
	- *Interferences:* High concentrations of ferric iron will interfere. Can be overcome by the addition of 1 ml KF.

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 43

Generally, phosphorus is the limiting nutrient in freshwater aquatic systems. If all phosphorous is used, plant growth will cease, no matter the amount of nitrogen available. In contrast to freshwater, nitrogen is the primary limiting nutrient in the seaward portions of most estuarine systems [36]. Thus, nitrogen levels control the rate of primary production. If a nitrogen limited system is supplied with high levels of nitrogen, significant increases in phytoplankton (algae) and macrophyte (larger aquatic plants) production may occur. The recommended level of nitrogen in estuaries to avoid algal blooms is 0.1 to 1 mg/l, while the

**Designated Use Limit (mg/l)(AWWA 1990 [4])** 

Human Consumption 10.0

Warm water fish 90.0

Brewing 30.0

Human Consumption 1.0

Warm water fish 5.0

 Human Consumption 10.0 Agriculture (Livestock etc.) 100.0

Table 7. Nitrogen designated use and corresponding limits.

1. **Ultraviolet Spectrophotometer Screening Method:** 

determine most suitable method.

• *Detection limits:* 0.1 mg/l nitrate

 maximum diversity 0.1 (and phosphorus 0.01) moderate diversity 1.0 (and phosphorus 0.1)

• *Interferences:* Dissolved organic matter, surfactants, NO2(-) and Cr(+).

• *Detection limits:* Used to screen non-contaminated samples (low inorganic matter) to

• *Detection limits:* NO3(-) ion activity between 0.00001 and 0.1 M (0.14 to 1400 mg/l) • *Interferences:* Chloride and bicarbonate, when their weight ratios to nitrate are >10,

phosphorus concentration is 0.01 to 0.1 mg/l. Limits suggested to maintain designated use:

 **Nitrate (NO3-N):** 

Aquatic Life

 **Nitrite (NO2-):** 

Aquatic Life

Aquatic life

Estuaries (recommended)

**3.7.2 Measuring techniques** 

2. **Ion Chromatography Method:** 

or >5, respectively.

• *Interferences:* N/A 3. **Nitrate Electrode Method:** 

**A. Nitrate-Nitrogen [8]** 

 **Nitrate+Nitrite:** 

Industry

#### 4. **Alum Flocculation Modification:**

	- *Interferences:* The presence of hydrogen sulphide gas will desensitize the electrode cells.

#### **3.7 Nitrogen**

Nitrogen makes up 78% of the atmosphere as gaseous molecular nitrogen, but most plants can use it only in the fixed forms of nitrate and ammonium. Nitrate and nitrite are inorganic ions occurring naturally as part of the nitrogen cycle [1].

The nitrogen cycle is composed of four processes. Three of the processes - fixation, ammonification, and nitrification - convert gaseous nitrogen into usable chemical forms. The fourth process, denitrification, converts fixed nitrogen back to the unusable gaseous nitrogen state [1].


In temperate zones, soil nitrate concentrations will vary seasonally with temperature and moisture levels. Fall and winter rains thoroughly remove all nitrates from the soil. During the spring and summer, the increased nitrogen-fixing activity of organisms and the addition of fertilizer cause the concentration of nitrates in the soil to steadily increase. Most of this nitrate is absorbed by plants. Thus, the removal of crops in the fall increases the chances for large flushes of nitrate from the soil to water bodies. Some leaching may occur in the spring if crops are not well- established enough to absorb the nitrogen [34].

#### **3.7.1 Numerical categories**

Water contaminated with nitrate is very difficult and costly to treat. Thus, if contamination affects a large water supply, the best alternative may be a new water source [35].

Generally, phosphorus is the limiting nutrient in freshwater aquatic systems. If all phosphorous is used, plant growth will cease, no matter the amount of nitrogen available. In contrast to freshwater, nitrogen is the primary limiting nutrient in the seaward portions of most estuarine systems [36]. Thus, nitrogen levels control the rate of primary production. If a nitrogen limited system is supplied with high levels of nitrogen, significant increases in phytoplankton (algae) and macrophyte (larger aquatic plants) production may occur. The recommended level of nitrogen in estuaries to avoid algal blooms is 0.1 to 1 mg/l, while the phosphorus concentration is 0.01 to 0.1 mg/l.


Limits suggested to maintain designated use:

42 Water Quality Monitoring and Assessment

5. **Membrane Electrode Method:** This method is ideal for field testing. Not applicable for industrial and domestic wastewater. Polarographic or galvanic oxygen-sensitive membrane electrodes are composed of two metal electrodes in contact with a supporting electrolyte that is separated from the test solution by a selective membrane. • *Interferences:* The presence of hydrogen sulphide gas will desensitize the electrode

Nitrogen makes up 78% of the atmosphere as gaseous molecular nitrogen, but most plants can use it only in the fixed forms of nitrate and ammonium. Nitrate and nitrite are inorganic

The nitrogen cycle is composed of four processes. Three of the processes - fixation, ammonification, and nitrification - convert gaseous nitrogen into usable chemical forms. The fourth process, denitrification, converts fixed nitrogen back to the unusable gaseous

• **Nitrogen fixation** is the conversion of nitrogen in its gaseous state to ammonia or nitrate. Nitrate is the product of high-energy fixation of atmospheric nitrogen and oxygen. High-energy fixation accounts for little (10%) of the nitrate entering the nitrogen cycle. In contrast, biological fixation accounts for 90% of the fixed nitrogen in the cycle. In biological fixation, molecular nitrogen (N2) is split into two free N molecules. The N molecules combine with hydrogen (H) molecules to yield ammonia

• **Ammonification** is a one-way reaction in which organisms break down amino acids

• **Nitrification** is the process in which ammonia is oxidized to nitrite and nitrate, yielding

• **Denitrification** is the process in which nitrates are reduced to gaseous nitrogen. This

In temperate zones, soil nitrate concentrations will vary seasonally with temperature and moisture levels. Fall and winter rains thoroughly remove all nitrates from the soil. During the spring and summer, the increased nitrogen-fixing activity of organisms and the addition of fertilizer cause the concentration of nitrates in the soil to steadily increase. Most of this nitrate is absorbed by plants. Thus, the removal of crops in the fall increases the chances for large flushes of nitrate from the soil to water bodies. Some leaching may occur in the spring

Water contaminated with nitrate is very difficult and costly to treat. Thus, if contamination

affects a large water supply, the best alternative may be a new water source [35].

• *Interferences:* High suspended solids may consume iodide in acid solution. This

4. **Alum Flocculation Modification:** 

cells.

**3.7 Nitrogen** 

nitrogen state [1].

(NH3).

and produce ammonia (NH3).

**3.7.1 Numerical categories** 

energy for decomposer organisms.

process is used by facultative anaerobes.

if crops are not well- established enough to absorb the nitrogen [34].

interference is removable by alum flocculation.

ions occurring naturally as part of the nitrogen cycle [1].

Table 7. Nitrogen designated use and corresponding limits.

#### **3.7.2 Measuring techniques**

#### **A. Nitrate-Nitrogen [8]**

#### 1. **Ultraviolet Spectrophotometer Screening Method:**


#### 2. **Ion Chromatography Method:**


#### 3. **Nitrate Electrode Method:**


Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 45

phosphorus, phosphorus adsorbed to particulates, and amorphous phosphorus. The dissolved phase includes inorganic phosphorus (generally in the soluble orthophosphate form), organic

The organic and inorganic particulate and soluble forms of phosphorus undergo continuous transformations. The dissolved phosphorus (usually as orthophosphate) is assimilated by phytoplankton and altered to organic phosphorus. The phytoplankton is then ingested by detritivores or zooplankton. Over half of the organic phosphorus taken up by zooplankton is excreted as inorganic phosphorus. Continuing the cycle, the inorganic P is rapidly

The EPA water quality criteria state that phosphates should not exceed 0.05 mg/l if streams discharge into lakes or reservoirs, 0.025 mg/l within a lake or reservoir, and 0.1 mg/l in streams or flowing waters not discharging into lakes or reservoirs to control algal growth [12]. Surface waters that are maintained at 0.01 to 0.03 mg/l of total phosphorus tend to

Reservoirs (CO) chlorophyll a 15 ug/l

(Minn.) Total P 0.015 mg/l

aquatic life Total P 0.025 mg/l Lakes (NC) chlorophyll a 40 ug/l

mountain lakes 0.02 mg/l (VT) Total P 0.014 mg/l

\*These figures are recommended; eutrophication is also dependent on freshwater influx, nutrient

(and nitrogen < 0.1) mg/l

Aquatic life support 0.1 ug/l elemental phosphorus

moderate diversity 0.1\* (and nitrogen < 1.0) mg/l

cycling, dilution, and flushing of a pollutant load in a particular estuary [40].

maximum diversity 0.01\* total phosphorus

streams/rivers: 0.1 mg/l streams entering lakes: 0.05 mg/l lakes/reservoirs: 0.025 mg/l

Total P 0.035 mg/l

Total P 0.05 mg/l

phosphorus excreted by organisms, and macromolecular colloidal phosphorus.

assimilated by phytoplankton [1][39].

remain uncontaminated by algal blooms.

**3.8.1 Numerical categories** 

Example. State criteria used:

Impoundments (EPA Region 4) water supply Total P .015 mg/l

Estuaries (recommended)

Table 8. Phosphorus designated use limits.

Designated Use Limit

Freshwater Aesthetics Federal criteria

	- *Detection limits:* 0.01 mg/l to 1.0 mg/l nitrate. Recommended especially for nitrate concentrations below 0.1 mg/l, when other methods lack sufficient sensitivity.
	- *Interferences:* Suspended matter in the column will restrict sample flow.

#### 5. **Automated Cadmium Reduction Method:**

	- *Detection limits:* 0.01 mg/l to 10 mg/l nitrate.
	- *Interferences:* NH3 and NO2(-), if present, are measured with NO3(-). Measure separately and subtract.
	- *Detection limits:* 0.01 mg/l to 10 mg/l nitrate.
	- *Interferences:* Color, sulphide ion concentrations of less than 10 mg/l.

#### **B. Total Kjeldahl Nitrogen [8][37]**

	- *Detection limits:* 0.05 mg/l to 2.0 mg/l.
	- *Interferences:* Iron and chromium ions tend to catalyze, while copper ions will inhibit the color reaction.

#### **3.8 Phosphorus**

Phosphorus (P) is an essential nutrient for all life forms. It plays a role in deoxyribonucleic acid (DNA), ribonucleic acid (RNA), adenosine diphosphate (ADP), and adenosine triphosphate (ATP). Phosphorus is required for these necessary components of life to occur.

Phosphorus does not exist in a gaseous state. Natural inorganic phosphorus deposits occur primarily as phosphate in the mineral apatite. Apatite is found in igneous and metamorphic rocks, and sedimentary rocks. When released into the environment, phosphates will speciate as orthophosphate according to the pH of the surrounding soil.

Phosphate is usually not readily available for uptake in soils. Phosphate is only freely soluble in acid solutions and under reducing conditions. In the soil it is rapidly immobilized as calcium or iron phosphates. Most of the phosphorus in soils is adsorbed to soil particles or incorporated into organic matter [1][38][39].

Phosphorus in freshwater and marine systems exists in either a particulate phase or a dissolved phase. Particulate matter includes living and dead plankton, precipitates of phosphorus, phosphorus adsorbed to particulates, and amorphous phosphorus. The dissolved phase includes inorganic phosphorus (generally in the soluble orthophosphate form), organic phosphorus excreted by organisms, and macromolecular colloidal phosphorus.

The organic and inorganic particulate and soluble forms of phosphorus undergo continuous transformations. The dissolved phosphorus (usually as orthophosphate) is assimilated by phytoplankton and altered to organic phosphorus. The phytoplankton is then ingested by detritivores or zooplankton. Over half of the organic phosphorus taken up by zooplankton is excreted as inorganic phosphorus. Continuing the cycle, the inorganic P is rapidly assimilated by phytoplankton [1][39].

The EPA water quality criteria state that phosphates should not exceed 0.05 mg/l if streams discharge into lakes or reservoirs, 0.025 mg/l within a lake or reservoir, and 0.1 mg/l in streams or flowing waters not discharging into lakes or reservoirs to control algal growth [12]. Surface waters that are maintained at 0.01 to 0.03 mg/l of total phosphorus tend to remain uncontaminated by algal blooms.

#### **3.8.1 Numerical categories**

Designated Use Limit

Freshwater Aesthetics Federal criteria

44 Water Quality Monitoring and Assessment

4. **Cadmium Reduction Method:** Nitrate is reduced to nitrite in the presence of cadmium. The nitrite concentration is determined by diazotizing with sulphanilamide and coupling with NED dihydrochloride to form a colored azo dye that is measured

• *Detection limits:* 0.01 mg/l to 1.0 mg/l nitrate. Recommended especially for nitrate concentrations below 0.1 mg/l, when other methods lack sufficient sensitivity.

6. **Titanous Chloride Method:** Nitrate is determined potentiometrically using an NH3 gas-sensing electrode after nitrate is reduced to NH3 by a titanous chloride reagent.

7. **Automated Hydrazine Reduction Method:** Nitrate is reduced to nitrite by hydrazine sulphate. The nitrite concentration is determined by diazotizing with sulphanilamide and coupling with NED dihydrochloride to form a colored azo dye that is measured

2. **Automated Phenate Colorimetric Method:** Reaction produces indophenol, an intensely

Phosphorus (P) is an essential nutrient for all life forms. It plays a role in deoxyribonucleic acid (DNA), ribonucleic acid (RNA), adenosine diphosphate (ADP), and adenosine triphosphate (ATP). Phosphorus is required for these necessary components of life to occur. Phosphorus does not exist in a gaseous state. Natural inorganic phosphorus deposits occur primarily as phosphate in the mineral apatite. Apatite is found in igneous and metamorphic rocks, and sedimentary rocks. When released into the environment, phosphates will speciate

Phosphate is usually not readily available for uptake in soils. Phosphate is only freely soluble in acid solutions and under reducing conditions. In the soil it is rapidly immobilized as calcium or iron phosphates. Most of the phosphorus in soils is adsorbed to soil particles

Phosphorus in freshwater and marine systems exists in either a particulate phase or a dissolved phase. Particulate matter includes living and dead plankton, precipitates of

• *Interferences:* Iron and chromium ions tend to catalyze, while copper ions will

• *Interferences:* Color, sulphide ion concentrations of less than 10 mg/l.

• *Interferences:* NH3 and NO2(-), if present, are measured with NO3(-). Measure

• *Interferences:* Suspended matter in the column will restrict sample flow.

colorimetrically.

calorimetrically.

blue compound.

**3.8 Phosphorus** 

**B. Total Kjeldahl Nitrogen [8][37]** 

1. D**igestion followed by distillation.** 

inhibit the color reaction.

or incorporated into organic matter [1][38][39].

• *Detection limits:* 0.05 mg/l to 2.0 mg/l.

as orthophosphate according to the pH of the surrounding soil.

5. **Automated Cadmium Reduction Method:** 

• *Interferences:* Turbidity, color

separately and subtract.

• *Detection limits:* 0.5 mg/l to 10 mg/l nitrate.

• *Detection limits:* 0.01 mg/l to 10 mg/l nitrate.

• *Detection limits:* 0.01 mg/l to 10 mg/l nitrate.


\*These figures are recommended; eutrophication is also dependent on freshwater influx, nutrient cycling, dilution, and flushing of a pollutant load in a particular estuary [40].

Table 8. Phosphorus designated use limits.

Water Quality Monitoring and Associated Distributed Measurement Systems: An Overview 47

tubular flow cell.

of orthophosphate present.

• *Detection limits:* 0.001 to 6 mg/l as P.

chromium, mercury, lead, arsenic, and antimony.

REDOX environment of the system [29].

**3.9.1 Numerical categories** 

for soils.

**3.9 Heavy metals** 

• *Detection limits:* 0.001 to 10.0 mg/l as P when photometric measurements are performed at 600 to 650 nm in a 15mm tubular flow cell, or 880 nm in a 50mm

• *Interferences:* >50 mg/l Fe(3+), 10 mg/l Cu, and 10 mg/l SiO2. Turbidity, color may

• *Detection limits:* 1 to 20 mg/l as P. This method is not good for water samples - best

• *Interferences:* Silica and arsenate interfere in heated samples. Blue color is formed by

• *Interferences:* Silica and arsenate interfere in heated samples. Blue color is formed by

ferrous iron, but does not interfere if iron concentration is < 100 mg/l. 4. **Stannous Chloride Method:** Molybdophosphoric acid is formed and reduced by

ferrous iron, but does not interfere if iron concentration is < 100 mg/l.

Heavy metals are elements having atomic weights between 63.546 and 200.590, and a specific gravity greater than 4.0. Living organisms require trace amounts of some heavy metals, including cobalt, copper, iron, manganese, molybdenum, vanadium, strontium, and zinc. Excessive levels of essential metals, however, can be detrimental to the organism. Nonessential heavy metals of particular concern to surface water systems are cadmium,

All heavy metals exist in surface waters in colloidal, particulate, and dissolved phases, although dissolved concentrations are generally low. The solubility of trace metals in surface waters is predominately controlled by the water pH, the type and concentration of ligands on which the metal could adsorb, and the oxidation state of the mineral components and the

The behaviour of metals in natural waters is a function of the substrate sediment composition, the suspended sediment composition, and the water chemistry. Sediment composed of fine sand and silt will generally have higher levels of adsorbed metal than will quartz, feldspar, and detrital carbonate-rich sediment. Metals also have a high affinity for

Heavy metals in surface water systems can be from natural or anthropogenic sources.

humic acids, organo-clays, and oxides coated with organic matter [29].

Numeric aquatic life guideline criteria extracted from several US codes are:

Currently, anthropogenic inputs of metals exceed natural inputs.

3. **Vanadomolybdophosphoric Acid Colorimetric Method:** Ammonium molybdate reacts under acid conditions to form a heteropolyacid. In the presence of vanadium, yellow vanadomolybdophosphoric acid is formed, the intensity of which indicates the amount

interfere. Arsenate provides a positive interference.

stannous chloride, forming intensely colored molybdenum blue.

Generally, phosphorus (as orthophosphate) is the limiting nutrient in freshwater aquatic systems; if all phosphorus is used, plant growth will cease, no matter how much nitrogen is available. The natural background levels of total phosphorus are generally less than 0.03 mg/l. The natural levels of orthophosphate usually range from 0.005 to 0.05 mg/l [2]. As mentioned before, in contrast to freshwater, nitrogen is generally the primary limiting nutrient in the seaward portions of estuarine systems. Systems may be phosphorus limited, however, or become so when nitrogen concentrations are high and N:P>16:1 [41].

#### **3.8.2 Measuring techniques**

**Total Phosphorus and Orthophosphate:** Analysis involves two procedural steps:

	- *Detection limits:* Ranges change with light path used.
	- *Interferences:* Arsenates react with the molybdate to form a similar blue color. Nitrite and hexavalent chromium interfere to yield results 3% less than actual at 1 mg/l and 10% to 15% less than actual at 10 mg/l.


Table 9. Ascorbic acid method detection limits as a function of light path length.

2. **Automated Ascorbic Acid Reduction Method:** Ammonium molybdate and potassium antimonyl tartrate react with orthophosphate in an acid medium to form an antimonyphosphomolybdate complex that forms a blue color suitable for photometric measurements when reduced by ascorbic acid.

	- *Detection limits:* 1 to 20 mg/l as P. This method is not good for water samples best for soils.
	- *Interferences:* Silica and arsenate interfere in heated samples. Blue color is formed by ferrous iron, but does not interfere if iron concentration is < 100 mg/l.
	- *Detection limits:* 0.001 to 6 mg/l as P.
	- *Interferences:* Silica and arsenate interfere in heated samples. Blue color is formed by ferrous iron, but does not interfere if iron concentration is < 100 mg/l.

#### **3.9 Heavy metals**

46 Water Quality Monitoring and Assessment

Generally, phosphorus (as orthophosphate) is the limiting nutrient in freshwater aquatic systems; if all phosphorus is used, plant growth will cease, no matter how much nitrogen is available. The natural background levels of total phosphorus are generally less than 0.03 mg/l. The natural levels of orthophosphate usually range from 0.005 to 0.05 mg/l [2]. As mentioned before, in contrast to freshwater, nitrogen is generally the primary limiting nutrient in the seaward portions of estuarine systems. Systems may be phosphorus limited, however, or become so when nitrogen concentrations are high and

**Total Phosphorus and Orthophosphate:** Analysis involves two procedural steps:

2. colorimetric evaluation of the dissolved orthophosphate concentration [8].

2. **Nitric Acid-Sulphuric Acid Method:** Recommended for most samples.

or more thorough techniques and adopted if results are identical.

• *Detection limits:* Ranges change with light path used.

mg/l and 10% to 15% less than actual at 10 mg/l.

1. conversion of the phosphorus form into dissolved orthophosphate by a digestion

1. **Perchloric Acid Digestion:** Recommended only for extremely difficult-to-analyze

3. **Persulphate Oxidation Method:** This simple method should be cross-checked with one

1. **Ascorbic Acid Method:** Ammonium molybdate and potassium antimonyl tartrate react with orthophosphate to form a heteropoly acid that is reduced to molybdenum blue by

• *Interferences:* Arsenates react with the molybdate to form a similar blue color. Nitrite and hexavalent chromium interfere to yield results 3% less than actual at 1

**Range (mg/l as P) Path (cm)** 

0.3 - 2.0 0.5

0.15 - 1.3 1.0

0.01 - 0.25 5.0

2. **Automated Ascorbic Acid Reduction Method:** Ammonium molybdate and potassium antimonyl tartrate react with orthophosphate in an acid medium to form an antimonyphosphomolybdate complex that forms a blue color suitable for photometric

Table 9. Ascorbic acid method detection limits as a function of light path length.

measurements when reduced by ascorbic acid.

N:P>16:1 [41].

**3.8.2 Measuring techniques** 

method, and

**Step 1.** Digestion methods

**Step 2.** Colorimetric methods

ascorbic acid.

samples, such as sediments.

Heavy metals are elements having atomic weights between 63.546 and 200.590, and a specific gravity greater than 4.0. Living organisms require trace amounts of some heavy metals, including cobalt, copper, iron, manganese, molybdenum, vanadium, strontium, and zinc. Excessive levels of essential metals, however, can be detrimental to the organism. Nonessential heavy metals of particular concern to surface water systems are cadmium, chromium, mercury, lead, arsenic, and antimony.

All heavy metals exist in surface waters in colloidal, particulate, and dissolved phases, although dissolved concentrations are generally low. The solubility of trace metals in surface waters is predominately controlled by the water pH, the type and concentration of ligands on which the metal could adsorb, and the oxidation state of the mineral components and the REDOX environment of the system [29].

The behaviour of metals in natural waters is a function of the substrate sediment composition, the suspended sediment composition, and the water chemistry. Sediment composed of fine sand and silt will generally have higher levels of adsorbed metal than will quartz, feldspar, and detrital carbonate-rich sediment. Metals also have a high affinity for humic acids, organo-clays, and oxides coated with organic matter [29].

Heavy metals in surface water systems can be from natural or anthropogenic sources. Currently, anthropogenic inputs of metals exceed natural inputs.

#### **3.9.1 Numerical categories**

Numeric aquatic life guideline criteria extracted from several US codes are:


\* Four-day average concentration

One-hour average concentration

+ Twenty-four hour average concentration

++ Level not to be exceeded at any time

(Adapted from [42][43][44]).
