BORANG PENGESAHAN STATUS TESIS***

Transcription

1 UNIVERSITI TEKNOLOGI MALAYSIA PSZ 19:16 (Pind. 1/97) BORANG PENGESAHAN STATUS TESIS*** JUDUL: MANAGEMENT OF MELANA WATERSHED USING MULTICRITERIA DECISION MAKING APPROACHES SESI PENGAJIAN: 2006/2007 Saya AHMAD AMZARI BIN YACCOB (HURUF BESAR) mengaku membenarkan tesis (PSM/ Sarjana/ Doktor Falsafah)* ini disimpan di Perpustakaan Universiti Teknologi Malaysia dengan syarat-syarat kegunaan seperti berikut: 1. Tesis adalah hakmilik Universiti Teknologi Malaysia. 2. Perpustakaan Universiti Teknologi Malaysia dibenarkan membuat salinan untuk tujuan pengajian sahaja. 3. Perpustakaan dibenarkan membuat salinan tesis ini sebagai bahan pertukaran antara institusi pengajian tinggi. 4. **Sila tandakan ( ) SULIT TERHAD (Mengandungi maklumat yang berdarjah keselamatan atau kepentingan Malaysia seperti yang termaktub di dalam AKTA RAHSIA RASMI 1972) (Mengandungi maklumat TERHAD yang telah ditentukan oleh organisasi/ badan di mana penyelidikan dijalankan) TIDAK TERHAD Disahkan oleh: (TANDATANGAN PENULIS) (TANDATANGAN PENYELIA) Alamat Tetap: NO. 5, JALAN MANIS, TAMAN SRI KENANGAN, Dr. SUPIAH BINTI SHAMSUDDIN BATU PAHAT. Nama Penyelia Tarikh: 10 MEI 2007 Tarikh: 10 MEI 2007 CATATAN: * Potong yang tidak berkenaan. ** Jika tesis ini SULIT atau TERHAD, sila lampirkan surat daripada pihak berkuasa/ organisasi berkenaan dengan menyatakan sekali sebab dan tempoh tesis ini perlu dikelaskan sebagai SULIT atau TERHAD. *** Tesis dimaksudkan sebagai tesis bagi Ijazah Doktor Falsafah dan Sarjana secara penyelidikan, atau disertasi bagi pengajian secara kerja kursus dan penyelidikan, atau Laporan Projek Sarjana Muda (PSM).

2 I/We* declare that I/we* have read this project report and in *my/our opinion this report is sufficient in terms of scope and quality for the award of the degree of Master of Engineering (Civil-Hydraulic and Hydrology). Signature :. Name of Supervisor : DR. SUPIAH BINTI SHAMSUDDIN Date : 10 MAY 2007

3 MANAGEMENT OF MELANA WATERSHED USING MULTICRITERIA DECISION MAKING APPROACHES AHMAD AMZARI BIN YACCOB Laporan projek ini dikemukakan sebagai memenuhi sebahagian daripada syarat penganugerahan Sarjana Kejuruteraan (Awam-Hidraul dan Hidrologi) Fakulti Kejuruteraan Awam Universiti Teknologi Malaysia MEI, 2007

4 MANAGEMENT OF MELANA WATERSHED USING MULTICRITERIA DECISION MAKING APPROACHES AHMAD AMZARI BIN YACCOB A project report submitted in partial fulfillment of the requirement for the award of the degree of Master of Engineering (Civil-Hydraulic and Hydrology) Faculty of Civil Engineering Universiti Teknologi Malaysia MAY, 2007

5 ii I declare that this project entitled Management of Melana Watershed Using Multicriteria Decision Making Approaches is the result of my own research except as cited in the references. The report has not been accepted for any degree and is not concurrently submitted in candidature of any other degree. Signature :.. Name : AHMAD AMZARI BIN YACCOB Date : 10 MAY 2007

6 iii Specially for my dad, mum, sisters and brother who I love so much and grateful thanks are due to the support from all of you.

7 iv ACKNOWLEDGEMENTS Firstly I would like to express my gratitude to my supervisor Dr.Supiah Shamsudin for her continuous guidance, support and advice in carrying out this research. I would also like to thank the staff of environment lab for their assistance and cooperation during sample lab test. Not forgetting also to my friends who help me to complete this project. My gratefulness goes to my family for their support and understanding. Acknowledgements would no be complete without thanking to Aina and Low and all my friends who directly or indirectly have in many ways gave me inspiration and contributed their ideas towards my project.

8 v ABSTRACT Nowadays, watershed management is very important and need to be given attention especially by government agencies because all resources between it are related. This project proposed using Multicriteria Decision Making Approach (MCDM) to identify and rank which subwatershed in Melana need to given priority by government authorities in planning their strategies or alternatives for improvement Melana catchment. Each subwatershed was ranked using Multicriteria Decision Making Approaches specifically applying Fuzzy Composite Programming (FCP) and Analytic Hierarchy Process (AHP). There were 3 selected subwatershed which were residential, light industry and heavy industry. The objective of this project i) to identify basic indicator for management of Melana watershed, ii) to rank subwatershed in Melana watershed using MCDM approaches, iii) to compare FCP and AHP technique and analyses based on selected basic indicators. The FCP and AHP structures contained 13 first-level indicators, 5 second level indicators, 3 third level indicators and the final indicators. The highest ranking subwatershed was residential with the highest ordered sequence value. The highest ranking subwatershed was also associated with the shortest distance between the fuzzy box and ideal point. The highest ordered sequences value means that subwatershed was the best in economy, water quantity and quality. The critical subwatershed was the lowest ordered sequence value which was heavy industry. In AHP, the highest priority index value was heavy industry. This subwatershed is the critical watershed which need be given attention and conservation. This study showed that FCP has a similar result with AHP.

9 vi ABSTRAK Sejak kebelakangan ini, pengurusan tadahan menjadi sesuatu yang amat penting dan perlu diberi perhatian terutama oleh badan kerajaan kerana semua sumber di dalam tadahan adalah saling bergantungan. Project ini mencadangkan penggunaan Multicriteria Decision Making Approach (MCDM) untuk mengenal pasti dan menentukan subtadahan mana yang perlu diberi keutamaan oleh badan kerajaan untuk diambil tindakan dalam mengatur strategi atau menyediakan alternatif untuk memulihkan keadaan tadahan Melana. Setiap subtadahan akan dipangkatkan dengan menggunakan Multicriteria Decision Making Approaches specifically iaitu Fuzzy Composite Programming (FCP) dan Analytic Hierarchy Process (AHP).Terdapat tiga subtadahan yang dipilih iaitu industri berat, perumahan dan industri ringan. FCP dan AHP struktur mengandungi tiga belas petunjuk peringkat pertama, lima petunjuk peringkat kedua, tiga petunjuk peringkat ketiga dan akhirnya sistem. Pangkat teratas bagi ketiga-tiga subtadahan ialah perumahan.dengan nilai susunan jujukan tertinggi. Pemangkatan subtadahan tertinggi adalah berkaitan dengan jarak terdekat diantara kotak fuzzi dan titik unggul. Nilai susunan jujukan tertinggi menunjukkan bahawa subtadahan tersebut adalah yang terbaik dari aspek ekonomi, kuantiti dan kualiti air. Subtadahan yang kritikal adalah yang mempunyai nilai susunan jujukan yang terendah iaitu industri berat. Dalam AHP, pula, nilai indeks keutamaan tertinggi adalah subtadahan industri berat. Subtadahan ini adalah subtadahan kritikal yang mana perlu diberi perhatian. Kajian ini menunjukkan bahawa FCP mempunyai keputusan yang sama dengan AHP.

10 vii LIST OF CONTENTS CHAPTER TITLE PAGE DECLARATION DEDICATION ACKNOWLEDGEMENTS ABSTRACT ABSTRAK LIST OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF ABBREVIATIONS LIST OF APPENDIXES ii iii iv v vi vii x xi xiii xiv I INTRODUCTION Introduction Problem Statement Objective of The Study 3 II LITERATURE REVIEW Definition of Management Concept Definition of Watershed 6

11 viii 2.2 Regional Patern in Watershed Condition Defination of Water Quality Water Quality Management Assessment for Water Quality Water Pollutant and Their Source Water Quality Parameters Water Quality According to Malaysian s 20 Interim National Water Quality standards 2.9 Defination of Multicritera Decision 22 Making 2.10 The Component of The Decision Process Multictreria Decision Making Tools 27 III METHODOLOGY OF THE STUDY Site Description Heavy Industry Residential Light Industy Fuzzy Composite Programming Introduction Fuzzy Logic and Fuzziness Fuzzy Sets Fuzzy Number Membership Value Membership Function Analytic Hierarchy Process Introduction The Principale of Identity and 45 Decomposition Pairwise Comparisons The Need for a Scale of Comparison Consistency To Estimate The Cosintency 49

12 ix Ratio IV RESULTS AND ANALYSIS Data Input Data Input for FCP Data Input for AHP Hierachies Sructure Result of Fuzzy Composite Programming Numerical Output Graphical Output Result of Analytic Hierachy Process Sensivity Analysis Discussion 70 V CONCLUSION AND RECOMMENDATION Conclusion Recommendation 73 REFERENCES 75 APPENDIX Appendix A Data Input for FCP 77 Appendix B Data Input for AHP 89 Appendix C Output for FCP 93 Appendix D Output for FCP 97

13 x LIST OF TABLE NO. OF TABLE TITLE PAGE 2.1 Major pollutant categories and priciples sources 15 of pollutants 2.2 Interim National Water Quality Standards 20 For Malaysia 3.1 Intensity of relative importance Data input for subwatershed heavy industy Data input for subwatershed residential Data input for subwatershed light industry Weight used in 3 trials Balancing factors used in 3 trials Average value for each parameter and subwatershed System sequence order Priority index values from AHP Sensitivity analysis trial on model 67

14 xi LIST OF FIGURES NO. OF FIGURE TITLE PAGE 2.1 llustration for alternative watershed definations Watershed diagram The hydrologic process Typical function of a) utility theory and b) prospect theory Network of Melana watershed Membership function for certain fuzzy A matrix that represnt the comparison of weights DO (mg/l) bwatershed heavy industry DO (mg/l) watershed residential DO (mg/l)bwatershed light industry Average value for DO (mg/l) Average value for COD (mg/l) Average value for BOD (mg/l) Fuzzy composite programming hierachy structure Analytic hierarchy process hierachy structure Water quality versus economy (trial 1) Water quantity versus economy (trial 1) Water quantity versus water quality (trial 1) Priority index for first analysis Priority index for second analysis Priority index for third analysis Sensivity analysis for synthesis 1 (trial 1) 68

15 xii NO. OF FIGURE TITLE PAGE 4.16 Sensivity analysis for synthesis 1 (trial 2) Sensivity analysis for synthesis 1 (trial 3) 69

16 xiii LIST OF ABBREVIATIONS BOD - Biochemical Oxygen Demand COD - Chemical Oxygen Demand DO - Dissolved Oxygen TSS - Total Suspended Solid MCDM - Multicriteria Decision Making FCP - Fuzzy Composite Programming AHP - Analytic Hierarchy Process km 3 - Kilometre cubic km 2 - Kilometre squire ET - Evapotranspiration % - Percentage F - Fahrenheit EPA - Environmental Protection Agency CaCO3 - Calcium Carbonate N - Nitrogen DA - Decision Analysis S ij - The actual value of basic index in second-level group j of basic indicators L ij - The composite index for second-level group j of basic indicators Nj - The number of basic indicators in group j α ij t - The weight expressing the relative importance of basic indicators in group j Pj - The balancing factor among indicators for group j

17 xiv LIST OF APPENDIXES APPENDIX TITLE PAGE A1 BOD (mg/l) for subwatershed heavy industry 78 A2 BOD (mg/l) for subwatershed residential 78 A3 BOD (mg/l) for subwatershed light industry 79 A4 COD (mg/l) for subwatershed heavy industry 79 A5 COD (mg/l) for subwatershed residential 80 A6 COD (mg/l) for subwatershed light industry 80 A7 TSS (mg/l) for subwatershed heavy industry 81 A8 TSS (mg/l) for subwatershed residential 81 A9 TSS (mg/l) for subwatershed light industry 82 A10 Ammonia (mg/l) for subwatershed heavy industry 82 A11 Ammonia (mg/l) for subwatershed residential 83 A12 Ammonia (mg/l) for subwatershed light industry 83 A13 Phosphorus (mg/l) for subwatershed heavy industry 84 A14 Phosphorus (mg/l) for subwatershed residential 84 A15 Phosphorus (mg/l) for subwatershed light industry 85 A16 Nitrate (mg/l) for subwatershed heavy industry 85 A17 Nitrate (mg/l) for subwatershed residential 86 A18 Nitrate (mg/l) for subwatershed light industry 86 A19 Flowrate (m 3 /s) for subwatershed heavy industry 87 A20 Flowrate (m 3 /s) for subwatershed residential 87 A21 Flowrate (m 3 /s) for subwatershed light industry 88 A22 Rainfall (mm) for all subwatershed 88 B1 Average value for TSS (mg/l) 90

18 xv APPENDIX TITLE PAGE B2 Average value for ammonia (mg/l) 90 B3 Average value for phosphorus (mg/l) 91 B4 Average value for nitrate (mg/l) 91 B5 Average value for flowrate (m 3 /s) 92 B6 Average value for rainfall (mm) 92 C1 Water quality versus economy (trial 2) 94 C2 Water quantity versus economy (trial 2) 94 C3 Water quantity versus water quality (trial 2) 95 C4 Water quality versus economy (trial 3) 95 C5 Water quantity versus economy (trial 3) 96 C6 Water quantity versus water quality (trial 3) 96 D1 Sensivity analysis for synthesis 2 (trial 1) 98 D2 Sensivity analysis for synthesis 2 (trial 2) 98 D3 Sensivity analysis for synthesis 2 (trial 3) 99 D4 Sensivity analysis for synthesis 3 (trial 1) 99 D5 Sensivity analysis for synthesis 3 (trial 2) 100 D6 Sensivity analysis for synthesis 3 (trial 3) 100

19 CHAPTER 1 INTRODUCTION 1.1 Introduction Malaysia is gifted with an enormous source of good land and fresh water supply. It supported by more than 2500 mm annual rainfall and a dense network of rivers and streams which about 150 major river basins. So that s mean, Malaysia supposedly must enjoy with these natural resources. No doubt Malaysia is called as a country of water resource-rich (Ayob Katimon and Supiah Shamsudin, 2005). In terms of hydrologic water balance, Malaysia has received 990 km 3 annual rainfalls and lost 360 km 3 due to evaporation (36 percent), which has 540 km 3 (54 percent) fresh water surpluses. The total surface runoff (the surface water generated by a combination of rainfall and watershed system) is 566 km 3 and about 64 km 3 contribute to groundwater recharge. Without considering water supply from groundwater system, it is clear that Malaysia has a surplus in fresh water supply. Water is very important for people, food and rural development, economic development and environment. But unfortunately, many states in Malaysia still have a problem with water supply especially in water supply shortages, low water quality, flash flood in urban

20 2 area and economy. So to prevent these problems, we need to manage our watershed from overall aspects. Management of watershed is suggested by using multi criteria decision making approaches. There are many techniques, adopting complex mathematical models and theories, are developed for decision making. Although all decision makers endeavor to improve final outcomes of decision making, the researcher is more concerned about the decision making process. From scientific point of view, a good final outcome depends upon a good decision making process. Principally, the purpose of MCDM is to provide help and guidance to the decision maker for further discovering his/her true preference in order to catch the most desirable solution of the problem. A large number of multicriteria decision making (MCDM) methods have been proposed in the past and applied to manage watershed or water resource planning (Keeney and Wood, 1977). In real-world decision-making processes in watershed management, decision making theory has become one of most important fields. It uses the optimisation methodology connected with a single criterion, but also satisfying concepts of multiple criteria. Fuzzy Composite Programming and Analytic Hierarchy Process are one of Multicriteria Decision Making methods. These methods were applied to manage Melana watershed and used to rank sub watershed based on their relative degree of potential to determine a potential that an individual site possesses. These methods also helped to identify sub watershed that need to pay more attention. These approaches are useful for handling watershed system complexity, use more through data and allowing for flexible analysis.

21 3 1.2 Problem Statement Nowadays, we always heard problem about water supply shortage and water quality. That happened because of we have not manage our watershed with proper methods. Management of watershed is very important to guide and coordinate the use of land and other resources in sustainable manner in order to obtain proper product such as water supply without affecting future values and productivity. This can be done through conservation of physical and environmental quality. A watershed has a strong basis for management because all resources within it are interrelated with each others. All resources include water, soil, forest, minerals, nutrient, habitats and clean air. So if we want all that resources preserve or sustainable for future generation, so we must identify, protect and improve the watershed first by using multi criteria decision making approaches. 1.3 Objectives Of The Study The objective of this study is to identify and rank manage Melana subwatershed by using multi criteria decision making approaches. Fuzzy Composite Programming and Analytic Hierarchy Process is used in this study as multi criteria decision making methods. The determination of the best strategy from a number of potential alternatives in watershed management is a complex decision making process. It may include conflicting quantitative and quality criteria and multiple decision-makers. The decision making process will be carried out using the multi-criteria decision making techniques.

22 4 The evaluation and ranking of alternatives by MCDM techniques is based on criteria values associated with each of the alternative, and the objectives and preferences by decision makers. Watershed ranking provides watershed scoring technique which would help prioritization of concerns and applicable for the purpose of preservation and mitigation potentials. Watershed preservation priorities should be defined to give special attention to watersheds in need of restoration and protection the most. There are three objectives for this study: 1) to identify basic indicator for management of Melana watershed 2) to rank Melana subwatershed using MCDM (FCP and AHP) approaches 3) to compare FCP and AHP technique in ranking Melana subwatershed.

23 CHAPTER II LITERATURE RIEW 2.0 Definition and Concept of Management "Management" (from Old French ménagement "the directing", from Latin manu agere "to lead by the hand") characterizes the process of leading and directing all or part of an organization, often a business, through the deployment and manipulation of resources (human, financial, material, intellectual or intangible). Early twentieth-century management writer Mary Parker Follett defined management as "the art of getting things done through people." One can also think of management functionally, as the action of measuring a quantity on a regular basis and of adjusting some initial plan, and as the actions taken to reach one's intended goal. This applies even in situations where planning does not take place. From this perspective, there are five management functions: planning, organizing, leading, co-ordinating and controlling. For others though, this definition, while useful, is far too narrow. The phrase "management is what managers do" is also prevalent, conveying the difficulty with which management is defined, the shifting nature of definitions, and the connection of managerial practices with the existence of a managerial cadre or class.

24 6 Management is known by some as "business administration", although this then excludes management in places outside business, eg charities and the public sector. University departments that teach management are nonetheless usually called "business schools". The term "management" may also be used as a collective word, describe the managers of an organization, for example of a corporation. 2.1 Definition of Watershed The term watershed describes an area of land that drains down slope to the lowest point. The water moves through a network of drainage pathways, both underground and on the surface. Generally, these pathways converge into streams and rivers, which become progressively larger as the water moves on downstream, eventually reaching an estuary and the ocean. Other terms used interchangeably with watershed include drainage basin or catchments basin. Land use is the key element affecting this area of land. The boundary of a watershed is defined by the highest elevations surrounding the stream. A drop of water falling outside of the boundary will drain to another watershed. A watershed, therefore, is "an area of land that drains water, sediment, and dissolved materials to a common outlet" Watersheds can be large or small. Every stream, tributary, or river has an associated watershed, and small watersheds join to become larger watersheds. It is relatively easy to delineate watersheds using a topographic map that shows stream channels. Watershed boundaries follow major ridgelines around channels and meet at the bottom, where water flows out of the watershed, a point commonly referred to as a stream or river.

25 7 The connectivity of the stream system is the primary reason for doing aquatic assessments at the watershed level. Connectivity refers to the physical connection between tributaries and the river, between surface water and groundwater, and between wetlands and water. Because water moves downstream, any activity that changes soil permeability, vegetation type or cover, water quality, quantity, or rate of flow at a location can change the characteristics of a stream or even the watershed at downstream locations. Land use practices such as clearing land for timber or agriculture, developing and maintaining roads, housing developments, and water diversions may have environmental consequences that greatly affect stream conditions even when the land use is not directly associated with a stream. Proper planning and adequate care in implementing projects can help ensure that one activity within a watershed does not detrimentally impact the downstream environment. For this reason, everyone living or working within a watershed needs to cooperate to ensure good watershed conditions. Figure 2.1: Illustration for alternative watershed definitions

26 8 Figure 2.2 : Watershed diagram Figure 2.3: The hydrologic process

27 9 Rain is the common form of precipitation, but snow, hail, dew, fog drip, and frost can also bring water into a watershed (imported water may also be introduced into the hydrologic cycle in any particular watershed). Precipitation that reaches the surface of the earth can move through three different pathways. Water can (1) be intercepted by vegetation and evaporated or transpired back to the atmosphere, (2) move down slope on the surface or through the soil to a stream system, eventually making its way back to the ocean, or (3) be stored in snow pack, groundwater, ponds, or wetlands for a variable period of time. Heat from the sun can evaporate water from any point in the cycle, from the surface of the vegetation, ground, or water bodies. The rate of evaporation in a watershed is dependent on the water surface area exposed to the air, temperature, humidity, and wind and transpiration, which is the loss of water to the atmosphere through living plants. The two processes of evaporation and transpiration are referred to as evapotranspiration (ET). ET can largely explain the difference between the amount of rainfall and the amount of runoff within a watershed. The amount of water intercepted, evaporated, and transpired in a watershed is dependent on the type and extent of vegetative cover present, and land uses that alter the vegetative cover of a watershed can affect the water available for runoff. The annual amount of water leaving a drainage basin as runoff varies, from less than 10% of yearly precipitation in hot desert climates to greater than 90% in the steep mountainous areas. When water evaporates, it enters the atmosphere as water vapor that eventually condenses to form drops of water (if the temperature is above 32ºF) or ice crystals (if below 32ºF). When the drops of water or ice crystals have grown large enough or when the wind moves them through the air, they fall to the earth. Upon reaching the earth, water will tend to move down slope, under the influence of gravity. Precipitation intensity, the rate at which water is delivered to the earth s surface, is usually less than the infiltration, the rate at which water enters the soil. Thus, under natural conditions most of the precipitation infiltrates into the soil and little, if any, surface runoff or overland flow occurs.

28 10 However, there are a few watershed conditions under which surface runoff will occur, including precipitation on very steep slopes, in watersheds with thin soils over bedrock, and in disturbed watersheds (particularly those with large areas of impervious surfaces or recently burned areas). Decreases in infiltration rates result in water moving more quickly to streams, causing peak flows to occur earlier and to be larger in magnitude, thus reducing water storage and resulting in lower flows during the dry months. Water that enters the soil infiltrates vertically toward the groundwater table. Once in the soil, water moves either toward the river channel or percolates and becomes groundwater. Water that enters the river channel becomes surface water and is called stream flow or discharge. Stream flow is conveyed by gravity through the stream channel to connect with progressively larger channels, eventually flowing into the ocean. The length of stream channels with actively flowing water depends on the volume of water moving through the system; during the dry season there are typically fewer actively flowing channels than during the wet season. Groundwater can be recharged from the nearby stream channel (a losing reach) or can provide flow to the channel (a gaining reach), depending on the relative energy (hydraulic gradient) and the ability to transmit flow. The direction of flow between connected surface- and groundwater depends on the hydraulic gradient and the ability of the soils to transmit flow. Water can be stored as groundwater storage, or in surface-water storage such as ponds, wetlands, lakes, and reservoirs. These surface water storage areas are often of special interest because they represent unique features in the watershed and may be used in special ways by plants, animals, and humans. Land uses in a watershed can have a significant impact on the amount of stored water and conditions that reduce the infiltration rate can also lead to a reduction in the amount of water held in groundwater storage.

29 Regional Patterns in Watershed Conditions The geology and climate regimes of an area influence the physical shape and condition of a watershed. Geologic uplifting from deep within the earth can influence the slope and orientation of the bedrock material, and earthquakes can cause large blocks of the earth s surface to shift, creating distinctive fault lines, which influence the direction and velocity of water flow. The position and orientation of ridges and valleys influence the local climate to determine the location, quantity and timing of precipitation that falls into the watershed. In addition, latitude, elevation, and aspect (directional position) are also important elements that influence the timing and the quantity of solar energy that a watershed receives. All of these processes and conditions influence the type and distribution of soils within the watershed, as well as the amount of erosion, character of sediments in the stream channels, and vegetation patterns. Watersheds are also subject to periodic disturbances that can have a significant influence on stream channel conditions. Watersheds will typically undergo short periods of catastrophic change, such as a large storm that produces huge floods and widespread landslides, followed by intervening periods of relative stability. Natural disturbances may include large-scale events, such as a flood or fire that impacts the entire watershed or localized events, such as a pocket of trees that are blown down into the stream channel. Both large- and small-scale natural disturbances are an integral part of the watershed s evolution, and the aquatic community has adapted to this disturbance regime. For example, floods can create new pools, and landslides can act to deliver large trees and spawning gravels to stream channels, both of which are important for forming fish and wildlife habitat. Human activities can modify the natural disturbance regimes by changing the timing and intensity of these natural processes. For example, as shown in Figure 2.5, urbanization and roads increase impervious surfaces and change the routing of water. These conditions can increase flood peaks and landslide frequency above the range

30 12 that would be expected in undeveloped conditions. Significant changes in pervious versus impervious conditions have occurred throughout the watershed. 2.3 Definition of Water Quality Water is essential to human life and to the health of the environment. As a valuable natural resource, it comprises marine, estuarine, freshwater (river and lakes) and groundwater environments, across coastal and inland areas. Water has two dimensions that are closely linked - quantity and quality. Water quality is commonly defined by its physical, chemical, biological and aesthetic (appearance and smell) characteristics. A healthy environment is one in which the water quality supports a rich and varied community of organisms and protects public health. Water quality in a body of water influences the way in which communities use the water for activities such as drinking, swimming or commercial purposes. More specifically, the water may be used by the community for: supplying drinking water recreation (swimming, boating) irrigating crops and watering stock industrial processes navigation and shipping production of edible fish, shellfish and crustaceans protection of aquatic ecosystems wildlife habitats scientific study and education

31 Water Quality Management The quality of surface and groundwater resources can significantly affect water use in many regions. In regions where water pollutants from human activities have seriously degraded water quality, the main issue in water quality management is to control pollution sources. Definition of the control levels depends on the water quality standards defined for the various water uses. The term water quality management implies that water should be managed so that no uses at any location will be detrimental to its use at another location. Water quality management deals with all aspects of water quality problems relating to the many beneficial uses of water. Water quality is a reflection or response of water composition to all inputs and processes, whether natural or cultural. Water quality management should not be equated to water pollution control which generally is the adequate treatment and disposal of wastewater. In the definition of water quality management above, water uses consist of intake; on site, and in stream flow uses. Intake uses include water for domestic, agricultural and industrial purposes, or uses that remove water from the source. On site uses primarily refer to water consumed by swamps, wetlands, evaporation from water bodies, natural vegetation, and unirrigated crops and wildlife. Flow uses include water for estuaries, wastewater dilution, navigation, hydroelectric power production, and fish, wildlife and recreation purposes. Water quality management serves to optimize water quality for all beneficial uses. Implied in this, is that water should be managed so that no use at any one location will be detrimental to its use at another location. In managing water quality, the factors and inputs that must be considered, include both man-made sources and natural causes. With respect to water quality changes, natural causes include geologic formations, vegetation, geographic factors and natural eutrophication.

32 Assessment for Water Quality Water quality is assessed by its physical, biological and chemical characteristics. Contamination can alter one or all of these characteristics and may originate from point or from ambient sources. The investigation and management of water resources systems for water quality must include consideration and evaluation of: a) the physical, chemical and biological composition of headwaters and significant groundwater discharges. b) water quantity and quality requirements for all existing and potential water uses c) the means of water withdrawal and their effect on water quality and quantity d) the existing and future water and wastewater treatment technology used to alter water quality e) the wastewater outfall configuration and effluent mixing f) the eutrophication status of the receiving waters g) the waste assimilative capacity of the receiving waters. h) the ecological changes that might be caused by wastewater discharges j) the potential effects of discharged waters. 2.6 Water Pollutants and Their Sources When the discharge of wastes disturbs the natural ecological balance of a water body, water pollution occurs. The wide range of water pollutants can be classified principal sources of each category. Point sources include domestic sewage and industrial wastes because they are collected and discharged into receiving

33 15 surface and groundwaters from a single point. The pollutant sources are non-point sources if the pollutants are discharged to water from multiple points. The major non-point sources can be classified as agricultural return flows and urban runoffs. Reduction or elimination of point sources of pollution can be implemented by proper treatment processes before discharging to receiving waters, but treatment of nonpoint effluents usually is not economically feasible. Table 2.1: Major Pollutant Categories and Principal Sources of Pollutants. (Source: Davis M. L. and Cornwell, 1961) Point Source Non-Point Sources Domestic Industrial Agricultural Urban Sewage Wastes Runoff Runoff Pollutant Category X X X X Oxygen-demanding material X X X X Nutrients X X X X Pathogents X X X X Suspended solids/sediments X X X Salts X X X Toxic metals X X Toxic organic chemicals X X Heat X 2.7 Water Quality Parameters a) Ammonia The ammonia criteria are only for the protection of aquatic life, as no criteria have been developed for the protection of human health (consumption of contaminated fish or drinking water). In United States, the data used in deriving the U.S. Environmental Protection Agency (EPA) criteria are predominantly from flowthrough tests in which ammonia concentrations were measured. Concentrations of

34 16 ammonia acutely toxic to fishes may cause loss of equilibrium, hyperexcitability, increased breathing, cardiac output and oxygen uptake, and, in extreme cases, convulsions, coma, and death. The ammonia criterion appropriate for the protection of aquatic animals will therefore in all likelihood be sufficiently protective of plant life a) Bacteria Bacteria, viruses, and other microorganisms excreted by diseased animals or persons and found in wastewater are pathogenic organisms. When pathogens containing sewage are discharged into a water body used for drinking and recreation, they pose a dangerous health hazard for the public. b) Dissolved Oxygen Demand This category of pollutants consists of materials that can be oxidized, thus consuming dissolved oxygen (DO) during the oxidation process. Human waste and food residues are sources of oxygen-demanding materials in domestic sewage. Depletion of DO can occur in a water body by discharging industrial wastes and agricultural runoffs that consist of organic matters. c) Total Dissolved Solids, Chloride and Conductivity Total dissolved solids, chlorides, and conductivity observations are typically used to indicate the magnitude of dissolved minerals in the water. The term total dissolved solids (or dissolved solids) is generally associated with fresh water and refers to the inorganic salts, small amounts of organic matter, and dissolved materials in the water. Chlorides (not chlorine) are directly related to salinity

35 17 because of the constant relationship between the major salts in seawater. Conductivity is a measure of the electrical conductivity of water and is also generally related to total dissolved solids, chlorides, or salinity. Excessive dissolved solids in drinking water can cause physiological effects, unpalatable mineral tastes, and higher costs because of corrosion or the necessity for additional treatment. The physiological effects directly related to dissolve solids include complication on cardiac disease and women with toxemia associated with pregnancy. According to Rawson and Moore (1944) lakes with dissolved solids in excess of 15,000 mg/l were unsuitable for most freshwater fishes d) Hardness Hardness commonly is reported as an equivalent concentration of calcium carbonate (CaCO3) or alkalinity. Concerns about water hardness originated because hard water requires more soap to form lather and because hard water causes scale in hot water systems. Modern use of synthetic detergents has eliminated the concern of hard water in laundries, but it is still of primary concern for many industrial water users. Many households use water softeners to reduce scale formation in hot water systems and for water taste reasons. Natural sources of hardness principally are lime stones which are dissolved by percolating rainwater. Industrial sources include the inorganic chemical industry and discharges from operating and abandoned mines. Hardness in fresh water is frequently distinguished in carbonate and non carbonate fractions

36 18 e) Heat An increase in water temperature can have a negative impact on the ecosystem. The rate of oxygen depletion increases as the temperature increases. This is important where the oxygen-demanding wastes are discharged to the water bodies. Many industries, such as thermal power plants, are the major sources of heat pollution of water bodies. f) Nutrients Phosphates and nitrates, derived from municipal wastewater, are inorganic nutrients. These nutrients are necessary for the growth of all living organisms but are classified as pollutants because excessive amounts of them can promote plant and algae growth. Fertilizers and phosphorus-based detergents and domestic wastes are the major sources of nutrients. In excessive concentrations, phosphates can stimulate plant growth. Excessive growths of aquatic plants (eutrophication) often interfere with water uses. In some aquatic systems, however, nitrogen compounds may be the most critical nutrients because of relatively large amounts of treated sewage (which is especially high in phosphates) in relation to other pollution sources, such as agricultural and urban runoff (which are high in nitrogen). In quantities normally found in food or feed, nitrates become toxic only under conditions in which they are, or may be, reduced to nitrites. High intake of nitrates constitutes a hazard primarily to warmblooded animals under conditions that are favorable to reduction to nitrite. When nitride reaches the bloodstream and reacts directly with hemoglobin to produce methemoglobin, consequently impairing oxygen transport. The reaction of nitrite with hemoglobin can be hazardous in infants under 3 months of age. Serious and occasionally fatal poisonings in infants have occurred following ingestion of untreated well waters shown to contain nitrate at concentrations greater than 10 mg/l nitrate nitrogen (N)

37 19 g) Toxic metals and organic compound Many heavy metals and chemicals are toxic and cancerous to living organisms. Pesticides in agricultural run off, lead and zinc in urban runoff, and toxic metals and toxic organic substances in industrial wastewater are some examples of toxic pollutant sources. h) Suspended solids and turbidity Suspended solids (sometimes referred to as nonfilterable residue) and turbidity are related to the solids content that is not dissolved. Turbidity refers to the blockage of light penetration beam, while suspended solids are measured by weighing the amount of dried sediment that is trapped on a 0.45m diameter filter, after filtering a known sample volume. Turbidity (and color) can be caused mostly by very small particles (less than 1m), while the suspended solids content is usually associated with more moderatesized particles (10 to 100m). Suspended solids can cause water quality problems directly, as discussed in they may also have other pollutants (such as organics and toxicants) associated with them that would cause additional problems. The control of suspended solids is required in most discharge permits because of potential sedimentation problems downstream of the discharge and the desire to control associated other pollutants. Turbid water also interferes with recreational use and aesthetic enjoyment of water.

38 Water quality according to Malaysian s Interim National Water Quality Standards In Malaysia, rivers or water catchment s water are classified base on the guideline suggested by Interim National Water Quality Standards. The suggested guidelines are summarized as below: Table 2.2 : Interim National Water Quality Standards for Malaysia Parameters Unit I IIA IIB III IV V Temperature o C - normal - normal - - ph Conductivity ìs/cm Colour Pt -CO DO mg/l < 3 < 1 BOD mg/l > 2 COD mg/l > 100 Oil & Grease mg/l natural 40; nil nil Dissolved solids mg/l Suspended solids mg/l > 300 Turbidity NTU Ammonia N mg/l > 2.7 Floatables nil nil nil Odour nil nil nil - - Salinity Taste nil nil nil - - -

39 21 E. coli MPN/100mL Total coliform MPN/100mL > Hardness mg/l natural K mg/l natural F mg/l natural > 1 NO 3 mg/l natural > 5 P mg/l natural S mg/l natural Cd mg/l natural > 0.01 Cu mg/l natural > 0.2 Fe mg/l natural > 5 Pb mg/l natural > 5 Mn mg/l natural > 0.2 Ni mg/l natural > 0.2 Class I represents water body of excellent quality. Standards are set for the conservation of natural environment in its undisturbed state. Water bodies such as those in the national park areas, fountainheads, and in high land and undisturbed areas come under this category where strictly no discharge of any kind is permitted. Water bodies in this category meet the most stringent requirements for human health and aquatic life protection. Class II represents water bodies of good quality. Most existing raw water supply sources come under this category. In practice, no body contact activity is allowed in this water for prevention of probable human pathogens. There is a need to introduce another class for water bodies not used for water supply but of similar quality, which may be referred to as Class IIB. The determination of Class IIB

40 22 standard is based on criteria for recreational use and protection of sensitive aquatic species. Class III is defined with the primary objective of protecting common and moderately tolerant aquatic species of economic value. Water under this classification may be used for water supply with extensive/advance treatment. This class of water is also defined to suit livestock drinking needs. Class IV defines water quality required for major agricultural irrigation activities, which may not cover minor applications to sensitive crops. Class V represents other waters, which do not meet any of the above uses. 2.9 Definition of Multi-Criteria Decision Making There are many techniques, adopting complex mathematical models and theories, are developed for decision making. Although all decision makers endeavor to improve final outcomes of decision making, the researcher is more concerned about the decision making process. From scientific point of view, a good final outcome depends upon a good decision making process. Principally, the purpose of MCDM is to provide help and guidance to the decision maker for further discovering his/her true preference in order to catch the most desirable solution of the problem. In essence, MCDM is trying to force the decision maker to understand his/her personal values in order to create more desirable alternatives to achieve meaningful decisions. Multi-criteria decision making comprises a broad set of methods and models that helps and guides the decision maker in discovering his or her most desired solution to the problem. The MCDM is interested in either

41 23 choosing one alternative (a choice problem) or obtaining an order of preference of the alternatives (a ranking problem). For a considerable period of time, the MCDM has been studied as part of the operation research and decision making models. Goal programming, introduced by Charnes and Cooper in 1961, was one of the early models that recognized the existence of multi-criteria for individual and organization decision making. In this model, the decision maker is forced to specify the goals of decision, and the satisfaction level of his decision making depends upon how close the goals are achieved. Multi-attribute decision theory, developed by Keeney and Wood (1977), incorporates the decision maker's preference towards risk into a mathematical function or a utility function. This approach falls into the category of decision analysis. The major difference between MCDM and decision analysis (DA) is the different treatment of preferences and alternatives. In general, MCDM considers the applicability of the preference axioms to real decision makers to be implicit and the actual decision alternatives to be explicit. In contrast with MCDM, DA explicitly determines the preference of the decision maker and considers the alternatives to be implicit or hypothetical. Under the DA, the decision maker's preference can be explicitly described by a mathematical function. Without uncertainty, the mathematical function can be referred to as a value function, and the purpose of the decision making is to maximize the value function through selecting the best alternatives. In essence, MCDM adopts the practical portion of DA but releases its strict assumptions. Under MCDM, the value function can be either stable or dynamic or do not even exist. However, the different assumptions about decision making will provide significant influence on solution methodology. The assumption of MCDM can be illustrated in three MCDM methods: The interactive solution method is a typical direct solution method that assumes the stable value function exists implicitly. The decision maker's

42 24 preference is progressively captured during a sequence of solution selection. After several interactions and elicitation of preference from the decision maker, the direct solution method stops at the solution where the maximum preference of the decision maker is found. The non-directed solution methods are based on the assumption of the implicit existence of dynamic value function. Under this method, the decision makers choose better solution at each iteration and conclude with a particular satisfied solution where the further venture is costly. The European MCDM School puts emphasis on helping the decision maker to learn about his/her preference. Figure 2.4: Typical function of a) utility theory and b) prospect theory A decision problem is based on a set of finite alternatives A: {a, b, c } and a set of judgement criteria C: {1,k, p}.given the alternatives and criteria, a matrix V:{V k (a)} is obtained reflecting the viewpoint of the DM, where V k (a) corresponds to the evaluation of alternative a (a Є A) with respect to criteria k. Value judgement for each criterion can be expressed on either a cardinal or a verbal

43 25 scale. These scales are employed for ordering alternatives with respect to criteria and to weight the criteria. By directly using verbal scales, judgmental statements are converted into numerical values read on the cardinal scale. The classical approach adopted in MCDM is to use a utility function. The shape of this function is concave-down for risk adverse persons, suggesting that after a certain value of risk the utility grows slower than expected. For risk seeking behaviour, there is a tendency to increase utility faster with increased risk, resulting in a concave-up form. Neutrality is represented as a linear function in figure 2.4(a). To get a global evaluation and a ranking of the alternatives, the value judgment matrix is normalised, say to 0-1, for each alternative and transformed using a set of weights, g k, associated with the importance of the criterion k. The final value of each alternative is obtained by summation of the transformed scores. Like the utility function, for prospect theory the shape of the value function explains risk aversion and risk seeking behaviours. While utility expresses the psychological preferences of a DM, the value function is measuring the differences in utility or preference. For prospect theory, the proposed function of Kahneman and Tversky is displayed in Figure 2.4(b). This function has an S-shape form that is concave above the reference or neutral point representing gains, and convex below the reference point representing losses. The concave part reflects risk aversion faced with gains, and the convex part is related with risk seeking behaviour when the DM is faced with losses. Because responses to losses are more extreme than to gains, the slope of the function for losses is steeper. Based on the prospect theory, an interactive multi-criteria decision making method (known as TODIM, Tomada de Decisao Interativa Multicriterio in Portuguese), has been proposed by Gomes. In this method, the matrix V, representing the value judgement of the DM, is normalized across alternatives

44 26 obtaining a matrix p criteria x n alternatives W: {w c (a)}, which is called the position matrix of the DM following Znotinas and Hipel terminology The Components of the Decision Process Traditional decision making models are focused on values, attributes, goals, and alternatives and subjective. Unlike the decision making models, MCDM is composed of the objectively defined set of alternatives and subjectively defined criteria. The criteria are independent of alternatives. In the decision making process, the most challenging work is to clarify and further construct the criteria that are close to the alternatives. In essence, the criteria and alternatives have different characteristics. Conceptually, the alternatives are determined by the criteria. However, the criteria are generally abstract and conceptual, and the alternatives are tangible in most of the cases. The attributes, determined by the decision context and the decision maker's preference, are constructed in hierarchy to fill in the gap between the criteria and the alternatives. Part of the decision making process is choosing the attributes. The chosen attributes should reflect both the measurable components of the alternatives and the decision maker's subjective criteria. The objective mapping between the attributes and the alternatives are independent of the decision maker's preference. This mapping includes data collection, statistical analysis and research models that help the decision maker evaluate possible alternatives. The subjective mapping between criteria and attributes is provided by the decision maker. This process is trying to clarify the criteria to force the decision maker to understand the reason of his/her criteria and the relationship with the attributes.

45 27 The essence of the typical decision making process includes three major steps. First of all, the criteria and the alternatives of the decision making need to be identified. Secondly, through objectively mapping attributes to alternatives and subjectively mapping criteria to attributes, the decision maker's preference is understood and clarified. Finally, through an iterative process to fully understand the preference, the decision maker has the opportunity to extend the set of alternatives. The decision making process is constrained by time, energy, and limited resources. The whole process is finished at the point that the utility level is high and further investigation is non-profitable Multi-Criteria Decision Making Tool There are several multi-criteria decision making(mcdm) tools such as linear multiobjective program (Zeleny,1982), goal programming (Lee et al,1985), compromise programming (Chankong and Haimes,1983), composite programming ( Bogardi and Bardossy, 1983), multiattribute utility( Zeleny,1982) and elimination and choice translating reality (ELECTRE) programming (Goicoechea et. al, 1982). The choice of an MCDM tool is dependent upon following factors: (1) types of attribute values(quantitative values), (2) types of alternatives (discrete or continuous options), (3) robustness of results with respect to changes in attributes values, (4) ease of computation (the time required by the analyst to master and implement the technique and to analyze the results, (5) amount of interaction required between decision maker and the system analyst, and (6) type of decision maker (whether it is an individual or group). For example, compromise programming, composite programming and multiattribute utility theory can effectively deal with either discrete or continuous

46 28 options. However, linear multiobjective programming and goal programming can only handle continuous sets of options, and ELECTRE programming can only deal with discrete sets of option. Detailed evaluation of the MCDM tools are shown in Zeleny (1982), Goicoechea et al.(1982) and Stansbury (1990). The MCDM tools describe above have been used to assist decision makers in solving problems involving multiple attributes and conflicting objectives. However, uncertainties inherent in these problems are often treated separately from the decision process itself. That is the decision analysis is often carried out using point or crisp estimates (no uncertainty) of input variables, even when uncertainties (sometimes large) exist in the values of input indicator variables. To incorporate associated uncertainties directly into the MCDM process. Michael E.Hagemeister (1996), used Fuzzy Composite Programming to rank hazard of landfills. He proposed to use a multicritera assessment system as a tool for screening and prioritizing unregulated disposal sites according to their level of environmental and health hazard. Using traditional method is time consuming and expensive. There is a need for a screening system that use more through but inexpensive data to estimate the potential hazards from a landfill site. The proposed system can be viewed as an intermediate assessment procedure, more precise than assessment procedure such as the NSWMP method, but not as comprehensive as a full-scale assessment. Therefore, it could be used to evaluate those sites believed to be potentially hazardous by a preliminary assessment procedure, thus determining which sites should receive immediate funds for further investigation. P. Anand Raj (1995), used fuzzy number to rank river basin planning and development alternatives under multi-criterion environment. The purposed is to find the most suitable planning of reservoirs with their associated purposed aimed at the development of one of the major peninsular river basin (Krishna river basin) in India. A set of 7 alternative systems with 8 main objectives, which are further subdivided into 18 criteria, are considered for ordering or ranking them employing

47 29 the opinion of three experts: an academician, a field engineer and an official from Ministry of Water Resources, using fuzzy numbers. B. Srdjevic used Analytic Hierarchy Process to evaluate management strategies in Paraguacu River. Reasonably large sets of criterion and objectives are manipulated during a three-phase process by appropriate shrinking and enlarging related decision hierarchies. In the first phase an unrestricted set of management interest are grouped into (1) long-term, and (2) middle- and short-term decision context by creating decision hierarchies and evaluating them by the Analytic Hierarchy Process (AHP). After refining the hierarchies by deleting dominated elements, management plans are introduced at their fingertips and so enlarges hierarchies are respected evaluated by AHP in the second phase. Only two bottom levels of hierarchies is necessary to consider by this break in evaluating procedure. Size of decision hierarchies is rationally preserved in both phase considering limited human abilities in handling numerous decision element while comparing them, and difficulties in attaining desired decision consistency The last phase aggregates weights of management plans derived in the second phase for two hierarchies and perform a final ranking of the management alternatives. In illustrating practical applications of FCP, Bender and Simonovic (2000), applied the technique to two water resources problems from the literature. In sorting fuzzy distance metrics (i.e., evaluating performance of alternatives), only two different ranking methods were employed, which yielded similar results. Since the literature search reveals the presence of about 30 or so ranking methods, it is reasonable to evaluate results produced by FCP by other methods as well. Therefore, the objective of this paper is to compare ranking produced by FCP and AHP.

48 Multicriteria Decision Making Under Uncertainty Because of the lack of a standard terminology, descriptions of human decision making are replete wiyth interchangeable terms. For this study, widely accepted terms ( attributes and indicators, objectives and goals, alternatives and strategies and multicriteria decision making under uncertainty are explained and carefully defined based on the literature (Zeleny, 1982). The terms attributes and indicators can be used interchangeably and they refer to descriptors of objective reality, which may be actual non subjective traits or subjectively assigned traits that are perceived as characteristics of objects. For example, one might consider several non subjective attributes (such as horsepower, gas mileage, weight, price and colour) and subjective attributes (such as comfort, styling and status image) to describe his or her choice among new automobiles. The terms objectives represents directions of desirability or improvement along individual attributes or complexes of attributes. For example, price in itself is an attribute but purchasing the cheapest automobile among the choices (minimizing price) is an objective, the direction of search related to the attribute price. However, an objective may sometimes derive from aggregates of attribute: for example, the objective of maximizing the prestige of an automobile s owner can derive from combine attributes of price, safety, comfort, horsepower and scarcity. The term goals refer quite unambiguously to specific target levels of achievement that can be defined in terms of both attributes and objectives. For example, maximizing gas mileage is an objective in the search for new automobile but achieving gas mileage of 26 miles per gallon is goal indicating a specific reference value for that objective. Alternatives are mutually exclusive activities, objects, projects or modes of behavior among which a choice is possible. For example, if one s purpose is to

49 31 travel from New York to Los Angeles, there are different means or ways of acting on that purpose: going by automobile, bus, train or airplane. He or she then chooses one of these ways to go to Las Angeles. In this study, other terms option, strategies and action are used but only as metaphoric substitutes for alternatives. Decision making is conducted by selecting or formulating different attributes, objectives, or goals. These three categories can be then be referred to as criteria. Therefore, multicriteria or multiple-criteria decision making (MCDM) indicates a concern with the general class of problems that involve multiple attributes, objectives and goals. Then the term MCDM under uncertainty can be broadly defined as a stuggle to solve MCDM problem where the values of attributes or indicators are imprecise.

50 CHAPTER III METHODOLOGY OF THE STUDY 3.1 Site Description This project under study is located at Sungai Melana catchment. This river started from Gunung Pulai until Sungai Skudai reach at Taman Perling district of Johor Bahru.Sungai Melana is one of network in Sungai Skudai basin. This watershed is located in Mutiara Rini area, Skudai in Johor Bahru district. The area of Melana watershed is about km 2 and now there are many changes in this area because of fast development. Many housing area being constructed in this watershed such as Taman Teratai, Taman Universiti, Taman Mutiara Rini, Taman Sri Pulai, and Taman Pulai Perdana to cater the increasing of population. Besides, kem tentera Diraja Malaysia also in construction in this Melana watershed. This watershed will be subdivided into three subwatershed based on the land use.

51 33 Figure 3.1: Network of Melana watershed For this study, thirty six samples of river were taken at three stations which required for research the quality of the rivers. This study will cover the area of river tributary which nearly workshops area at Taman Teratai and this area was classified as heavy industry. Taman Universiti and Taman Pulai Perdana classified as residential. Besides, Taman Peridustrian Universiti classified as light industry..

52 Heavy Industry The location of heavy industry is at Taman Teratai. That area classified as heavy industry because most of the shop at that area is vehicles workshop. The workshop repair many types of vehicles such as lorry, car, van and also school bus. Besides, all pollution waste such as oil and grease which produced by repair work of vehicles was discharge to drainage system and then entered the river. This process will reduce the water quality of that river Residential This residential area is located at Taman Pulai Perdana. Many houses at this area is double stories terraced houses. The population of people at this residential area is about 5000 peoples. From economy aspect, overall people at this residential area are having moderately income. The study focused at this area because pollutant is transported through runoff and flow to the river Light Industry This area is located at Jalan Perdagangan 1, Taman Universiti. This area coverage about 225 terrace light industry. Many small factories at this area operate as a shoe, clothes, cake, metal food made factory and many others. To complete enough the component of watershed system management, the required data such as flowrate at each station must be collect. Data of total rainfall also need to know by take the data from Department Irrigation and Drainage (DID) Kuala Lumpur. All

53 35 obtained data from lab test and site will analyse using software Fuzzy Composite Programming (FCP) and Analytic Hierarchy Process (AHP). This is important to get which watershed need to give attention or priority by government authority. 3.2 Fuzzy Composite Programming Introduction Fuzzy Composite Programming, an extension of compromise programming (Zeleny,1982) was developed by Bardossy and Duckstein(1992). This programming is a distance based technique that depends on the point of reference or ideal point. This technique attempts to minimize the distance from the ideal solution for a satisfying solution. The closest one to the ideal across all criteria is the compromise solution or compromise set. Fuzzy compromise programming is a multiobjective decision analysis tool powerful enough to accurately model subjective information while keeping interpretation of results relatively straightforward. The concept of non dominance is used in distance-based techniques to select the best solution or choice of alternative. A solution is said to be non dominated if there exists no other feasible solution that will cause an improvement in a value of the objective or criterion functions without making a value of any other objective function worse The non dominance solution concept, originating with Pareto in 1906, has been one of the cornerstones of traditional economic theory. It is usually stated as the Pareto principle:

54 Fuzzy Logic and Fuzziness The logic of fuzzy and it application was introduced by Zadeh who was expert in fuzzy logic development. Since 1965, application of fuzzy logic was grow and apply in human life especially in Japan. As example, fuzzy logic applied in camera product which need technique of automatic in focusing. Fuzziness refers to vagueness and uncertainty, in particular to the vagueness related to human language and thinking, examples as: the set of tall people all people living close to my home all areas that are very suitable for corn Fuzziness provides a way to obtain conclusions from vague, ambiguous or imprecise information. It imitates the human reasoning process of working with non precise data Fuzzy Sets and Fuzzy Number Fuzziness or uncertainty represent situations where membership in sets cannot be defined on a yes/no basis because the boundaries of the sets are vague. The central concept of fuzzy-set theory is the membership function, which represent numerically the degree to which an element belongs to a set. In a classical set, a sharp or unambiguous distinction exists between the members and nonmembers of the set. In other words, the value of the membership

55 37 function of each elements in the classical set is either 1 for members (those that certainly belong to the set) or 0 for nonmembers (those that certainly do not). However, it is sometimes difficult to make a sharp or precise distinction between the members and nonmembers of a set. For examples, the boundaries of the sets of beautiful womens, highly contaminated water, deep wells, or numbers much greeater than 1.0 are fuzzy. Since the transition from members to nonmembers appears gradual rather than abrupt. The fuzzy set introduces vagueness (with the aim of reducing complexity) by eliminating the sharp boundary dividing members of the set from nonmembers (Klir and Folger, 1988). Thus, if an element is a member of a fuzzy set to some degree, the value of its membership function can be between 0 and 1. When the membership function of an element can only have values 0 or 1, the fuzzy-set theory reverts to the classical-set theory. Figure 3.2: Membership function for certain fuzzy

56 Membership Value All membership function is a data which obtained and identified from fuzzy set with assume value 0 and 1. Value of 0 represent unreal set and value 1 represent the real value of certain set Membership Function Membership function is a curve that shows the space which includes membership value between value 0 and 1. As example fuzzy set for a river is all uncertainty with river flow. All data is consider to count and will present as a curve which include from any shape such as linear line, triangle, trapezium, exponent, quadratic and cubic Ordinarily membership function present into graphic form which it showed the representative value. That s mean the value in the curve show the probability to happen is exist in triangle shape or other shape. The example of equation below is derived for triangle shape of membership function (Anand, 1997).

57 39 All obtained data also can be formula into matrix form as shown below: Matrix schedule technique will make the process of calculation easier. From the collection of data R dan R k, calculation of fuzzy weight for each alternative will apply the equation below (Anand, 1997): ( ~ w i ; i = 1,2..., m) ~ [( m n ) + ( m n ) +... ( m n )] w i = ( 1/ KL) i1 i1 i 2 i2 + ik ik Where: K = number of criteria that to be consider L = value that analyst used (0 to infinity) m i1 = represent value in criteria R k. matrix n i1 = represent value in criteria R. matrix. The weight value for certain fuzzy set can be simplified as shown below (Anand, 1997):

58 40 After through the number of process above, value for membership function can be draw as result from the calculation that have been made(anand, 1997). From the drawn membership function figure for each alternative, the determination of utility value can be determine from maximum membership function µ M (x) and minimum µ m (x) after maximum and minimum x value is identified(anand, 1997) Fuzzy Composite Programming Format Composite programming (CP) organizes a problem into the following format: 1. define management alternatives 2. define basic indicators 3. group basic indicators into progressively smaller, more general group 4. define weights, balancing factors and worst and best values for the indicators and 5. evaluate and rank the alternatives

59 41 The balancing factors are assigned for each group of indicators. Balancing factors indicate the importance of the maximal deviations of the indicators and limit the ability of one indicator to substitute for another. In other words, with an appropriately high balancing factor, an indicator that must not be compromised such as endanger species will not be replaced byu one such as risk an abundant, nonendanger species.the larger the value of balancing factor, the greater the concern with respect to the maximal deviation. For p = 1, all deviations are weighted equally. For p = 2, each deviation is weighted in proportion to ist magnitude. As p becomes larger and larger, the largest deviation receives more and more weight, until finally at p = we observed that the distance corresponds to the maximum deviation: L j = max s ij. Clearly, the nature of the investigated system should govern the choice of the two types of weighting factors. As a general rule it can be stated that p = 3 or greater is to be used for ecological related indicators wheneverthe limiting factor priciple, such as limiting nutrient, holds for the system. In other cases, p = 1 or 2 seem to be good choice (Goicoechea et al.1982). The best and worst possible values for the basic indicators for the particularly study are then assessed. Evidently, in a number of cases the minimum values represents the best situation. Water quality indicators feferring to chemical concentration are typical examples in this respect. The best value would mean an ideal value representing ideal condition in the region considered. Whil some of the indicators can be characterized by overall best value, regardless of the region considered, such as some water quality indicators as dissolved oxygen, the best value for other indicators depends very much on the region investigated. For example, the best value of percapita income an important basic indicator and would be different in developing country from the one representing an industrialized country. Similarly, the worst value of the basic indicators corresponds to the absolute adverse situation in the region considered.

60 42 Evaluation of the various management alternatives proceeds by computing composite distance using CP. The first step is normalization (plazing into the 0-1 interval) of the basic indicator value (Z i ). This is necessary since the units of the various basic indicators can be quite different (e.g probability, ringgit Malaysia) and difficult to compare directly. The ability to compare and evaluate variables having different units is one of the primary benefit of using CP. Using the maximum (Z + i ) and the minimum (Z i ), the normalized value (S i ) of Z i can be computed as: S i Z = ( Z i + i ( Z ) ( Z i ) i ) (1) Next, second level composite indices are calculated for every second-level group of basic indices, using the following equation : L j n j = α ij S i= 1 pj ij 1 pj (2) where S ij L ij Nj α ij Pj the actual value of basic index in second-level group j of basic indicators the composite index for second-level group j of basic indicators the number of basic indicators in group j the weight expressing the relative importance of basic indicators in group j the balancing factor among indicators for group j To obtain the optimal solution or to compare between alternatives, the decision-maker must provide a complete set of weights as required by Equation (2). These weight parameters are established based on the degree of importance for each indicator possesses relative to other indicators of the same group. The tentative

61 43 weight used in Equation. 2 range between 0 and 1.0. The preferences were identified from the highest ordered sequence value, N which was computed from: N j β j α j = 2 + α j (3) Where β and α were obtained from the interaction line of maximizing and minimizing membership function (Chen, 1985). The balancing factor, p reflects the maximal deviations between indicators of the same group. The normal values used for balancing factors in Equation. 2 are 1.0 and 2.0. By increasing the p value in Equation (2) the influence of the maximum deviations from the ideal point on the value of L j is increased. In other words, when the decision-maker uses a high value of p, those alternatives that have a poor performance will be penalized severely. This allows the decision-maker to impose different values of p to different groups of objectives. The uncertainty in the determination of the distance from the ideal is the consequence of the uncertainty inherent in the information that fed the multiobjective decision process. The calculated fuzzy distances for all alternatives were then used to determine the closest distance to the ideal solution. The alternative that minimizes Equation (2) will be the optimal solution to the problem. If the problem involves only a few alternatives, it is possible to achieve an order of preference in the alternatives by visual inspection.

62 Analytic Hierarchy Process Introduction AHP was developed in the 1970 s by Dr. Thomas Saaty.The Analytic Hierarchy Process is a systematic procedure for representing the element of any problem, hierarchically. It organizes the basic rationality by breaking down a problem into its smaller and smaller constituent parts and then guides decision makers through a series of pairwise comparison judgements to express the relative strength or intensity of impact of the elements in the hierarchy. These judgements are then translated to numbers. The Analytic Hierarchy Process is a powerful and flexible decision making process to help people set priorities and make the best decision when both quantitative and qualitative aspects of decision need to be considered. By reducing complex decisions to a series of one-on-one comparisons, then synthesizing the results, AHP not only help decision makers arrives the best decision but also provide a clear rationale that is the best. This hierarchy is then manipulated analytically to produce a final matrix representing the overall priorities of the alternatives relative to each other. One can then make a logical decision based on the pairwise comparisons made between the alternatives and the criteria being used in the decision.

63 The Principle of Identity and Decomposition The principle of identity and decomposition calls for structuring problems hierarchically which is the first step one must be complete when using the AHP. In its most elementary form, a hierarchy is structured from the top, through intermediate levels to the lowest level. There are several kind of hierarchies. The simplest are dominance hierarchies, which descend like an inverted tree with the boss at the top, followed by successive level of bossing. Holarchies are essentially dominance hierarchies with feedback. Chinese box (or modular) hierarchies grow in size from the simplest elements or components (the inner boxes) to the larger and larger aggregates (the outer boxes). In biology, neogenetic hierarchies are of interest because of their new top levels that emerge successively through evolution. A hierarchy is said to be complete when every element of a given level functions as a criterion for all elements of the level below. Otherwise, it is incomplete. There is no problem with the priorities of the appropriate element with respect to which the evaluation is made. That is, the hierarchy can be divided into subhierarchies sharing only a common topmost element Pairwise Comparisons In the AHP, elements of a problem are compared in pairs with respect to their relative impact ( weight or intensity ) on a property they share in common. We reduce the pairwise comparisons to a matrix form-a square form in which an array of numbers is arranged as in the following example:

64 The brackets that enclose this four-by-four matrix are used to identify such a group of numbers in its standard form. When we compare a set of elements of a problems with each other a square matrix is produced that resembles the following: α 11 α 12 α 13 α 1n α 21 α 22 α 23 α 2n α 31 α 32 α 33 α 3n α n1 α n2 α n3 α nn This matrix has reciprocal properties; that is: α ji 1 = α ij (4) Where the subscripts i and j refer to the row and column. Let A 1, A 2, A 3,, A n be any set off elements and w 1, w 2,w 3,.,w n their corresponding weights or intensities. Using the AHP, we want to compare the corresponding weight or intensities of each element in the set with respect to a property or goal that they have in common. The comparison of weights can be represented as follows:

65 47 Know as a vector of the matrix Figure 3.3: A matrix that represent the comparison of weights A matrix may consist of only one row or one column in which case it is called a vector. The square matrix obviously has an equal number of rows and columns, but it has other useful properties, such as eigenvectors and eigenvalues. These will be developed further to solve this reciprocal matrix. The reason for this computation is that it gives us a way to determine quantitatively the relative importance of factors or issue in a problem situation. The factors with the highest values are the ones that we should concentrate on in solving a problem or developing a plan of action. It is important to see that if w 1, w 2,w 3,.w n are not known in advance, then we perform pairwise comparisons on the elements by using subjective judgements estimated numerically from a scale of number and then solve the number to find w s.

66 The Need for a Scale of Comparison In the matrix, on begins with an element on the left and asks how much more important is it than an element listed on the top. When compared with itself the ratio is one. When compared with another element, both it is more important than that element, and then an integer value, from the scale given it s reciprocal in the opposite case. In either case the reciprocal ratio is entered in transpose position of the matrix. Thus we are always dealing with positive reciprocal matrices and need only elicit n(n-1)/2 judgement where n is the total number of elements being compared. Table 3.1: Intensity of relative importance (Saaty T.L dan Kearns K.P, 1980).

67 Consistency A key step is the establishment of priorities through the use of the pairwise comparison procedure above. An important consideration is the consistence of the judgements made by the decision-maker. Perfect consistency is practically impossible to achieve. We need a method to measure the degree of consistency among the pairwise judgements provided by the decision-maker. If the degree is acceptable then the decision process continues or else the decision-maker should reconsider and possibly revise the pairwise judgements before proceeding with the analysis. A measure of consistency used by the AHP that can be computed is known as the Consistency Ratio. This ratio is designed so that values of the ratio exceeding 0.1 are indicative of inconsistent judgements indicating that the decision maker would probably want to revise the original values in the pairwise comparison matrix. Again we will use an approximation to the exact mathematical calculations which are again very complex To Estimate the Consistence Ratio 1. Multiply each value in the first column of the pairwise comparison matrix by the relative priority of the first item considered. Multiply each value in the second column of the pairwise comparison matrix by the relative priority of the second item considered. Multiply each value in the third column of the pairwise comparison matrix by the relative priority of the third item considered. Sum the values across the rows to obtain a vector of "weighted sums".

68 50 2. Divide the elements of the vector of weighted sums obtained in step 1 by the corresponding priority value. 3. Computer the average of the values in step 2 and denote it by λ max. 4. Compute the Consistency Index which is defined as follows: CI = λ max - n n - 1 where n is the number of items being compared. 5. Compute the Consistency Ration (CR) which is defined as: CR = CI RI Where RI, is the random index and CI is the consistency index of a randomly generated pairwise comparison matrix.

69 CHAPTER IV RESULTS AND ANALYSIS 4.1 Data Input Data for certain parameters of water quality and quantity were obtained from river samples and test at three selected stations. Rainfall data was obtained from the Department Irrigation and Drainage (DID) and all samples were tested at Environmental Laboratory, UTM. Data for economy was obtained based on researcher s own judgements using rank numbers. Rank number was used as the real values for revenue, construction and maintenance cost considered private and confidential by the factory or industries management team. All the 36 samples results were classified into ranges. Histograms were plotted based on that most likely range and maximum possible range for all the basic indicators. However, the rank numbers were established by assuming 5 categories a number were assigned to each category as follows

70 52 a) 1 is for than RM 30, 000 b) 2 is for RM30, 001 to RM80, 000 c) 3 is for RM80, 001 to RM120, 000, d) 4 is for RM120, 001 to RM400, 000 e) 5 is for RM400, 001 and above Data Input for FCP The manual selection of ideal and the worst point for the all basic indicators were not utilized. Instead, automatic selection was used by inputting -1 if minimum value is optimum and 1 if otherwise. The following step would be the determination of the weight and balancing factor for each basic indicator. 3 trials were conducted for sensitivity analysis. Weights for all 3 trials were determined by the researcher himself and agreed by the supervisor. Weights (w) represent the relative importance between indicators in a group. The greater the importance of an indicator, the greater weight is assigned to it. The sets of balancing factor used for the study reflect the importance of maximal deviations of the indicators. This maximal deviation is the maximum difference between an indicator value and the best value for that indicator. In this research, the balancing factor values were determined by the researcher and agreed by the supervisor.

71 53 Table 4.1: Data Input for subwatershed heavy industry Basic Indicator: Maximum Possible Value: Most Likely Value: Low High Low High Maintenance Cost Construction Cost Industrial Commercial Rainfall Flowrate DO BOD COD TSS Nitrate Phosphorus Ammonia Table 4.2: Data Input for subwatershed residential Basic Indicator: Maximum Possible Value: Most Likely Value: Low High Low High Maintenance Cost Construction Cost Industrial Commercial Rainfall Flow rate DO BOD COD TSS Nitrate Phosphorus Ammonia Table 4.3: Data Input for subwatershed light industry Basic Indicator: Maximum Possible Value: Most Likely Value: Low High Low High Maintenance Cost Construction Cost Industrial Commercial Rainfall

72 54 Flow rate DO BOD COD TSS Nitrate Phosphorus Ammonia Table 4.4: Weight used in FCP for the 3 trials: Indicators: Weights: Trial 1 Trial 2 Trial 3 Level 1 Maintenance Cost Construction Cost Industrial Commercial Rainfall Flow rate DO BOD COD TSS Nitrate Phosphorus Ammonia Level 2 Investment Revenue Quantity Oxygen Based Contaminant Level 3 Economy Water Quantity Water Quality

73 55 Table 4.5: Balancing Factor used in FCP for the 3 trials Indicator Balancing Factor: Trial 1 Trial 2 Trial 3 Investment Revenue Quantity Oxygen Based Contaminant Economy Water Quantity Water Quality System All the values in the tables would be input into Fuzzy Composite programming for evaluation. The output from the program s computation would be in numerical data and graphical form. However, the researcher replot the graph using Excel spreadsheet, based on the numeral data output. Before data for all input parameter we obtained, we need to build a histogram for each parameter and subwatershed. The data inputs for Fuzzy Composite Programming which is maximum possible value and most likely range were obtained from the histogram and membership function Figure 4.1 showed the histogram of DO (mg/l) for subwatershed heavy industry. Maximum possible value is 1.73 to 2.22 mg/l and most likely range value is 2.00 to 2.05 mg/l. Figure 4.1 to figure 4.3 showed histograms of DO (mg/l) parameter and other subwatershed. Other figures were shown in Appendix A.

74 56 Figure 4.1: DO (mg/l) for subwatershed heavy industry Figure 4.2: DO (mg/l) for subwatershed residential

75 57 Figure 4.3: DO (mg/l) for subwatershed light industry Data Input for AHP Table 4.6: Average values for each parameter of the subwatershed Basic Indicator Unit Average Value Heavy Industry Residential Light Industry DO mg/l COD mg/l BOD mg/l TSS mg/l Ammonia mg/l Phosphorus mg/l Nitrate mg/l Flowrate m 3 /s Rainfall mm

76 58 Table 4.6 showed the average values for each parameter and each subwatershed. To see clearly the average values for overall indicator, histograms of parameter distribution was created for each parameter and subwatershed. Figure 4.4 to Figure 4.6 showed the parameter distribution for average values for DO (mg/l), COD (mg.l), and BOD (mg/l) from the three subwatersheds. Other parameters were presented in Appendix B. AVERAGE VALUE FOR DO(mg/l) DO (mg/l) Heavy Industry Residential Light Industry Subwatershed Figure 4.4: Average value of DO (mg/l) for each subwatershed

77 59 AVERAGE VALUE FOR COD (mg/l) COD(mg/l) Heavy Industry Residential Light Industry Subwatershed Figure 4.5: Average value of COD (mg/l) for each subwatershed AVERAGE VALUE FOR BOD (mg/l) BOD (mg/l) Heavy Industry Residential Light Industry Subwatershed Figure 4.6: Average value of BOD (mg/l) for each subwatershed

78 Hierarchies Structure Analysis for this study started with process of decomposing problem into hierarchy structure. Hierarchy is a method to represent complex problem into storey shape according to their prioritisation. Figure 4.7: Fuzzy composite programming (FCP) composite structure From left to right, basic indicators are grouped into progressively smaller, and more general group. The overall result for the system is obtained by considering basic indicators, level one indicators and level two indicators.

79 61 Figure 4.8: Analytic hierarchy process (AHP) hierarchy structure Top stage is the objective of the study which is to give attention to the worst subwatershed. The second stage is first level indicators, third stage is second level indicator and the forth stage is basic indicators. The bottom stage is the alternative of subwatershed.

80 Result of Fuzzy Composite Programming Fuzzy composite programming gave numerical and graphical output as an output of analysis Numerical Output The Fuzzy Composite programme is also capable of evaluating problems and produced numerical outputs. The analysis involved three options which is residential, light industry and heavy industry. Three trials were carried out during the analysis using Fuzzy Composite programme is shown below. Table 4.7: Systems Sequence Order Trials: Systems Sequence Order: Option Values Residential Light Industry Heavy Industry 2 Residential Light Industry Heavy Industry 3 Residential Light Industry Heavy Industry From the analysis, heavy industry is the worst subwatershed and need to be given attention and taken corrective action especially by the government agencies.

81 Graphical Output Graphical output was obtained from numerical output of analysis. The ordered sequence value calculated by the fuzzy composite programme revealed that heavy industry is the most critical subwatershed. Figure 4.9 to 4.11 showed graphical charts of overall analysis for trial 1. Other graphical chart for trial 2 and 3 were shown in Appendix C. Figure 4.9: Water Quality versus Economy (Trial 1) Ideal point

82 64 Ideal point Figure 4.10: Water Quantity versus Economy (Trial 1) Ideal point Figure 4.11: Water Quantity versus Water Quality (Trial 1)

83 Result of Analytic Hierarchy Process For analytic hierarchy process, there are three syntheses being carried out using different weight to obtain the variation of results of the study. Three sensitive analyses were carried out to obtain the robustness of the subwatershed selected. Table 4.8: Priority index values from AHP Analysis Weight Alternative of subwatershed Inconsis Economic Quantity Quality Industry -tial Industry Water Water Heavy Residen Light -tency Figure 4.12: Priority index of first analysis

84 66 Figure 4.13: Priority index of second analysis Figure 4.14: Priority index of third analysis The result obtained was shown in table 4.8, the highest priority index is heavy industry with values about to 3.73 followed by light industry and residential alternative. The highest value of priority index means that the alternative need more attention and action by government authority.

85 Sensitivity Analysis Sensitivity analyses from the goal node will show the sensitivity of the alternatives with respect to all the objectives below the goal. It can be performed from the nodes under the goal if the model has more than three levels of nodes with respect to an objective or sub-objective. When performing a sensitivity analysis the priorities of the objectives may be varied and observe how the priorities of the alternatives would change. There are five types of sensitivity analysis. Dynamic Performance Gradient Head to Head Two-Dimensional (2D Plot) In this study, performace type is used to see how strong the result was which obtained from the previous analysis by rearranging the importance of objective listed below the goal. This type of sensitivity analysis was selected because this type can show the overall criteria and objective in one graph. So we can easily see the changes of result if we input the different weight for each objective. Table 4.9: Sensitivity analysis trial on model Synthesis Trial Weight Priority Water Water Economic Index Quantity Quality Result Heavy Industry Heavy Industry Light Industry Light Industry Heavy Industry Heavy Industry Heavy Industry Heavy Industry Light Industry

86 68 Figure 4.15: Sensitivity analysis for synthesis 1 (trial 1) Figure 4.16: Sensitivity analysis for synthesis 1 (trial 2)

87 69 Figure 4.17: Sensitivity analysis for synthesis 1 (trial 3) Figure 4.15 to 4.17 showed sensivity analysis for synthesis 1 for each trials. Other figures were at Appendix D.

88 Discussion Fuzzy Composite Programming, is an extension of compromise programming This programming is a distance based technique that depends on the point of reference or ideal point. This technique attempts to minimize the distance from the ideal solution for a satisfying solution. The closest one to the ideal across all criteria is the compromise solution or compromise set. Fuzzy compromise programming is a multiobjective decision analysis tool powerful enough to accurately model subjective information while keeping interpretation of results relatively straightforward. Fuzzy composite programming (FCP) results showed that, the shortest distance to the ideal point was subwatershed residential, followed by light industry and heavy industry. After many trials, it is confirmed that heavy industry is the the worst subwatershed. From numerical output, the highest sequence order value was residential with Second one was light industry with and lastly was heavy industry with The highest sequence order value shows the best subwatershed based on three aspects which are economic, water quality and water quantity. The lowest sequence order value shows the worst subwatershed based on three aspects which are economic, water quality and water quantity. Analytic hierarchy process (AHP) also gave the same result even though some analysis showed light industry was the worst subwatershed through synthesis analysis and sensitivity analysis. From synthesis analysis and sensitivity analysis, the highest priority index value means the worst subwatershed and need to be given priority and attention by government authority. Corrective action should be taken by government and associated agencies. Immediate controls should be taken to increase economic value and to reduce pollution for heavy industry. Since, the location of heavy industry is at the up-stream

89 71 of Melana watershed, this can cause more serious problem, especially pollution, in the down-stream if no controls measures are imposed. It is necessary to focus on light industry also because the results from AHP and FCP such as value of priority index and sequence order are almost similar with each other.

90 CHAPTER V CONCLUSION AND RECOMMENDATION 5.1 Conclusion Fuzzy Composite Programming (FCP) produced the ordered sequence value and trade-off boxes between economy, water quantity and water quality. The width of the boxes represents the uncertainty and fuzzinest in the trade-off. The shortest distance between the fuzzy box and ideal point was the highest ranking point. Residential area is the highest ranking subwatershed with highest ordered sequence value followed by light and heavy industry. Residential area is the best in term of trade-off between economy, water quantity and quality. The future management and planning should concentrate at heavy industry because that area was the overall most critical subwatershed. Analytic Hierarchy Process (AHP) result also showed that heavy industry is the critical one compare to othersubwatershed. Analytic Hierarchy process result showed that the highest ranking watershed is heavy industry with value followed by light industry and residential area. Many diagrams of priority index were obtained with the

91 73 different weight for economy, water quantity and quality, but heavy industry still the most critical subwatershed. Therefore heavy industry is considered the most robust option. Multicriteria decision making approach specifically Fuzzy Composite Programming and Analytic Hierachy Process were applied to rank subwatershed in Melana watershed system which includes residential, heavy industry and light industry. The study showed that residential was the best point in terms of trade-off between economy, water quantity and quality. Heavy industry is the most critical subwatershed and this subwatershed need proper management and attention by related authority agencies to improve the economy, water quality and quantity aspects. The differences between FCP and AHP are: 1. FCP use actual data for the analysis. 2. AHP was based on priority value that the decision maker has to setup or input 3. AHP was also based on human judgement. 5.2 Recommendation Due to time constrain and difficulty in obtaining relevant data, a simpler version of composite structure was built and used. In order to improve the results, the following recommendations are suggested:

92 74 1. The data for water quality should be observed form time to time for a period of one year. This is because concentration of pollutants in the water may be reduced and flow rate increases during rainy days and vice versa during dry seasons. 2. Rain gauge and evaporation pan should be built and used at interested subwatersheds so that more reliable data can be obtained. 3. More water quality basic indicators should be used when ranking alternatives as the composite structure used for this analysis is a simplified version of composite structure.

93 75 REFERENCES Anand Raj (1995), Multicriteria Methods In River Basin Planning-A Case Study, Department of Civil Engineering, Regional Engineering College, WARANGAL. Anand Raj and D. Kumar (1995), Ranking multi-criterion river basin planning alternatives using fuzzy numbers, Department of Civil Engineering, Regional Engineering College, WARANGAL. Ayob Katimon, and Supiah Shamsudin (2005), Watershed Protection: A Key Factor Towards Sustainable Reservoir Yield, Faculty of Civil Engineering, University of Technology Malaysia. B. Srdjevic, Y. Medeiros, Z. Srdjevic and M. Schaer, Evaluating Management Strategies in Paraguacu River Basin by Analytic Hierarchy Process, Faculty of Agriculture, University of Novi Sad, Yugoslavia. Bender, and M.Simonovic, (2000), a fuzzy compromise approach to water resources planning under uncertainty fuzzy set and system. Bogardi.L and Bardossy, A. (1983),. Aplication of MCDM to geological exploration Essays and Surveys on Multiple Criterion Decision Making, P. Hansen. Ed. Springler-verlag, New York, N.Y. Chankong and Haimes (1983), Multiobjective decision making: Theory and Methodology. North-Holland: New York.

94 76 Chen, S.H. (1985) Ranking fuzzy numbers with maximizing set and minimizing set, Fuzzy Sets and Systems, 17: Davis, M. L. and Cornwell, D. A., Environmental Engineering, McGraw-Hill, New York, 1961 Goicoechea, A. Hansen, D.R, and Duckstein, L (1982), Multiobjective decision analysis with engineering and business applications. John Wiley & Sons, Inc N.Y. Keeney, R.L, and Wood, E.F (1977). An illustrative example of the use of multiattribute utility theory for water resources planning. Lee, S.M, Laurence, J.M, and Taylor III, B.W (1985), Management science. Wm. C.Brown Publisher, Dubuque, Iowa. Michael Hagemeister, David and Woldt (1996), Hazard ranking of landfill using fuzzy composite programming. Predrag Prodanovic and Slobadan P.Simonovic, Comparison of fuzzy set ranking methods for implementation in water resources decision making. Department of Civil and Environmental Engineering, The University of Western Ontario, London. Thomas L. Saaty, Kevin P.Kearns (1980) Analytic planning the organization of systems. Zeleny, M. (1982) Multiple criteria decision making, McGraw-Hill, New York,

95 APPENDIX A DATA INPUT FOR FCP

96 78 Figure A1: BOD (mg/l) for subwatershed heavy industry Figure A2: BOD (mg/l) for subwatershed residential

97 79 Figure A3: BOD (mg/l) for subwatershed light industry Figure A4: COD (mg/l) for subwatershed heavy industry

98 80 Figure A5: COD (mg/l) for subwatershed residential Figure A6: COD (mg/l) for subwatershed light industry

99 81 Figure A7: TSS (mg/l) for subwatershed heavy industry Figure A8: TSS (mg/l) for subwatershed residential

100 82 Figure A9: TSS (mg/l) for subwatershed light industry Figure A10: Ammonia (mg/l) for subwatershed heavy industry

101 83 Figure A11: Ammonia (mg/l) for subwatershed residential Figure A12: Ammonia (mg/l) for subwatershed light industry

102 84 Figure A13: Phosphorus (mg/l) for subwatershed heavy industry Figure A14: Phosphorus (mg/l)for subwatershed residential

103 85 Figure A15: Phosphorus (mg/l) for subwatershed light industry Figure A16: Nitrate (mg/l) for subwatershed heavy industry

104 86 Figure A17: Nitrate (mg/l) for subwatershed residential Figure A18: Nitrate (mg/l) range for subwatershed light industry

105 87 Figure A19: Flowrate (m 3 /s) for subwatershed heavy industry Figure A20: Flowrate (m 3 /s) for subwatershed residential

106 88 Figure A21: Flowrate (m 3 /s) for subwatershed light industry Figure A22: Rainfall (mm) for all subwatershed