THESIS TITLE GOES HERE IF NECESSARY SECOND LINE IF NECESSARY THIRDLINE. M.Sc. Thesis by Student Name SURNAME

Similar documents
You can see w h at gold is u sed f or dow n below : )

Linear Computational Structural Mechanics for Wind Energy Systems

Handbook of Territorial Diversity

DEVELOPMENT OF LAND SUITABILITY EVALUATION SYSTEM FOR COASTAL AQUACULTURE USING ARTIFICIAL NEURAL NETWORK AND GEOGRAPHICAL INFORMATION SYSTEMS

[THESIS TITLE] - USE ONLY UPPERCASE LETTERS

UNCERTAINTY ANALYSIS OF TWO-SHAFT GAS TURBINE PARAMETER OF ARTIFICIAL NEURAL NETWORK (ANN) APPROXIMATED FUNCTION USING SEQUENTIAL PERTURBATION METHOD

I n Bl ack & Wh i te

Introduction to Eco n o m et rics

Application of an artificial neural network to typhoon rainfall forecasting

TIME SERIES DATA ANALYSIS USING EVIEWS

Elements of Multivariate Time Series Analysis

G. S. Maddala Kajal Lahiri. WILEY A John Wiley and Sons, Ltd., Publication

UNIVERSITI TEKNOLOGI MARA

Undergraduate June 2001, Ankara University Faculty of Engineering Department of Physics Engineering

Biological Forum An International Journal 7(1): (2015) ISSN No. (Print): ISSN No. (Online):

NON-STATIONARY & NON-LINEAR ANALYSIS, & PREDICTION OF HYDROCLIMATIC VARIABLES OF AFRICA

EFFECT OF NOZZLE ANGLE ON JET IMPINGEMENT COOLING SYSTEM KHAIDER BIN ABU BAKAR

Colorado Springs, CO B.A., Chemistry GPA: 3.73 May 2009 Research Advisor: Dr. Rongson Pongdee

IBAC 2012 vol.2. Assist. Prof. Dr. Beyza USTAOĞLU

5 Autoregressive-Moving-Average Modeling

Dorset Coastal Explorer Planning

Welcome Note. Your local organizing team

ASSESSING THE POTENTIAL IMPACT OF RAINFALL CHANGE ON SOUTHERN AFRICAN RIVER FLOWS RAKHEE LAKHRAJ-GOVENDER SUPERVISOR: PROF STEFAN GRAB

Optimal Artificial Neural Network Modeling of Sedimentation yield and Runoff in high flow season of Indus River at Besham Qila for Terbela Dam

Summer Review Packet. for students entering. AP Calculus BC

TEXAS STATE VITA. A. Name: Niem Tu HUYNH Title: Assistant Professor

Forecasting of Rain Fall in Mirzapur District, Uttar Pradesh, India Using Feed-Forward Artificial Neural Network

2007 Area Source Emissions Inventory Methodology 670 RANGE IMPROVEMENT

TIME SERIES ANALYSIS. Forecasting and Control. Wiley. Fifth Edition GWILYM M. JENKINS GEORGE E. P. BOX GREGORY C. REINSEL GRETA M.

Stochastic Hydrology. a) Data Mining for Evolution of Association Rules for Droughts and Floods in India using Climate Inputs

Cross validation of prediction models for seasonal time series by parametric bootstrapping

Time Series Analysis Model for Rainfall Data in Jordan: Case Study for Using Time Series Analysis

Estimation and application of best ARIMA model for forecasting the uranium price.

The performance of estimation methods for generalized linear mixed models

Peter Haggett Professor of Urban and Regional Geography, University of Bristol

Neuroevolution methodologies applied to sediment forecasting

CLEAR SPACE AND MINIMUM SIZE. A clear space area free of competing visual elements should be maintained.

Summer Review Packet. for students entering. IB Math SL

Study of Time Series and Development of System Identification Model for Agarwada Raingauge Station

Forecasting River Flow in the USA: A Comparison between Auto-Regression and Neural Network Non-Parametric Models

Comparison learning algorithms for artificial neural network model for flood forecasting, Chiang Mai, Thailand

Measuring Water Resources Unit 3: Groundwater withdrawal and fluctuations in the height of Earth's surface student exercise

Data and prognosis for renewable energy

Mayan Math. Extra Credit Grade 6 mathematics Mr. Livingood and Mrs. Dettlinger. Extra Credit: Part 1 MAYAN AND ROMAN MATH SYMBOLS

AN ASSESSMENT OF THE RELATIONSHIP BETWEEN RAINFALL AND LAKE VICTORIA LEVELS IN UGANDA

Forecasting of meteorological drought using ARIMA model

Recurrent Neural Networks. COMP-550 Oct 5, 2017

DEFRA FLOOD RISK MANAGEMENT AND MODELLING COMPETITION

Earth System Modeling EAS 4610 / 6130 Fall Semester 2005

FORMAT FOR MASTER S THESES IN GEOLOGY AT SAN JOSE STATE UNIVERSITY APRIL 2009

Combining Geospatial and Statistical Data for Analysis & Dissemination

Application of Time Sequence Model Based on Excluded Seasonality in Daily Runoff Prediction

Forecasting Drought in Tel River Basin using Feed-forward Recursive Neural Network

The Fitting of a SARIMA model to Monthly Naira-Euro Exchange Rates

The student will plan and conduct investigations in which

Recommended Grade Level: 8 Earth/Environmental Science Weather vs. Climate

CURRICULUM VITAE PANAGIOTIS N. GAVRILIADIS

Worksheet: The Climate in Numbers and Graphs

Geographical General Regression Neural Network (GGRNN) Tool For Geographically Weighted Regression Analysis

1. Introduction. Karen A. Latínez Sotomayor 1 Instituto Geofísico del Perú, Lima, Perú

New Introduction to Multiple Time Series Analysis

Model Implementation Details for the Final Project in EEOS 465/

Lab: Box-Jenkins Methodology - US Wholesale Price Indicator

ESTIMATION OF GLOBAL SOLAR RADIATION AT CALABAR USING TWO MODELS

PLANNED UPGRADE OF NIWA S HIGH INTENSITY RAINFALL DESIGN SYSTEM (HIRDS)

STATE UNIVERSITY OF NEW YORK COLLEGE OF TECHNOLOGY CANTON, NEW YORK

Practical Statistics for Geographers and Earth Scientists

Biostatistics 301A. Repeated measurement analysis (mixed models)

Statistical Methods. for Forecasting

ANALYSIS OF DEPTH-AREA-DURATION CURVES OF RAINFALL IN SEMIARID AND ARID REGIONS USING GEOSTATISTICAL METHODS: SIRJAN KAFEH NAMAK WATERSHED, IRAN

PROJECT TITLE. B.Tech. BRANCH NAME NAME 2 (ROLL NUMBER)

Time Series Analysis. James D. Hamilton PRINCETON UNIVERSITY PRESS PRINCETON, NEW JERSEY

EE-588 ADVANCED TOPICS IN NEURAL NETWORK

Study Guide. Earth Systems 1109 Weather Dynamics. Adult Basic Education Science. Prerequisites: Credit Value: 1

Science 2015 Earth Science. Curriculum Guide

DR. ÖVGÜ ÇIDAR İYİKAL

Statistical Methods for Forecasting

Optimum Neural Network Architecture for Precipitation Prediction of Myanmar

Lab # - Ocean Bottom Topography. Background Information:

Analysis. Components of a Time Series

VALIDATION OF SPATIAL INTERPOLATION TECHNIQUES IN GIS

Statistics 349(02) Review Questions

Cover Sheet: Request 10675

Empirical Approach to Modelling and Forecasting Inflation in Ghana

ENSO effects on mean temperature in Turkey

Inflow forecasting for lakes using Artificial Neural Networks

CLIMATE CHANGE IMPACTS ON HYDROMETEOROLOGICAL VARIABLES AT LAKE KARLA WATERSHED

Chapter-1 Introduction

FRACTAL RIVER BASINS

Comparison of Mann-Kendall and innovative trend method (Şen trend) for monthly total precipitation (Middle Black Sea Region, Turkey)

An Example of FFGS Implementation: Black Sea and Middle East FFG System. WMO; Name of Department (ND)

MODELING MAXIMUM MONTHLY TEMPERATURE IN KATUNAYAKE REGION, SRI LANKA: A SARIMA APPROACH

Prediction of Grain Products in Turkey

A Joint European GIS Under Construction: The 1:5 Million International Geological Map of Europe and Adjacent Areas (IGME 5000)

Temporal and Spatial Distribution of Tourism Climate Comfort in Isfahan Province

Call for Hosting the 16 th World (ISSP) Congress in Sport Psychology in 2025

Requisite Approval must be attached

STOCHASTIC MODELING OF MONTHLY RAINFALL AT KOTA REGION

Modeling of peak inflow dates for a snowmelt dominated basin Evan Heisman. CVEN 6833: Advanced Data Analysis Fall 2012 Prof. Balaji Rajagopalan

Design of Time Series Model for Road Accident Fatal Death in Tamilnadu

Transcription:

İSTANBUL TECHNICAL UNIVERSITY INSTITUTE OF SCIENCE AND TECHNOLOGY Don t write your degree. Write only Name SURNAME THESIS TITLE GOES HERE IF NECESSARY SECOND LINE IF NECESSARY THIRDLINE Delete this note before You print the thesis. Thesis supervisor text box exists in draft thesis (white cover) but not in final version (blue or black cover). Delete this note before You print the thesis. M.Sc. Thesis by Student Name SURNAME Department : Whatever Engineering or Science Programme : Whatever Engineering Thesis Supervisor: Prof. Dr. Name SURNAME THESIS SUBMISSION MONTH AND YEAR

İSTANBUL TECHNICAL UNIVERSITY INSTITUTE OF SCIENCE AND TECHNOLOGY This date should be the day when you submit draft thesis (white cover) to the department (not institute) and doesnt change in the final version (blue/black cover). Delete this note before You print the thesis. THESIS TITLE GOES HERE IF NECESSARY SECOND LINE IF NECESSARY THIRDLINE M.Sc. Thesis by Student Name SURNAME (Student Number) Date of submission : 06 February 2008 Date of defence examination: 27 February 2008 Supervisor (Chairman) : Prof. Dr. Name SURNAME (ITU) Members of the Examining Committee : Prof. Dr. Name SURNAME (METU) White cover: the Assis. Prof. Dr. Name SURNAME (BU) date when you submit thesis to department Blue/black cover: the date when you submit final THESIS SUBMISSION MONTH AND YEAR version to institute. Delete this note before You print the thesis.

İSTANBUL TEKNİK ÜNİVERSİTESİ FEN BİLİMLERİ ENSTİTÜSÜ TURKISH THESIS TITLE GOES HERE IF NECESSARY SECOND LINE IF NECESSARY THIRDLINE YÜKSEK LİSANS TEZİ Student Name SURNAME (Student Number) Tezin Enstitüye Verildiği Tarih : 06 Şubat 2008 Tezin Savunulduğu Tarih : 27 Şubat 2008 Tez Danışmanı : Prof. Dr. Ad SOYAD (İTÜ) Diğer Jüri Üyeleri : Doç. Dr. Ad SOYAD (ODTÜ) Yrd. Doç. Dr. Ad SOYAD (BÜ) THESIS SUBMISSION MONTH AND YEAR ( month should beturkish )

FOREWORD Line spacing: 1 Delete this note before You print the thesis. FOREWORD I would like to express my deep appreciation and thanks for my advisor. This work is supported by ITU Institute of Science and Technology. February 2008 Thesis Author Name Surname Whatever Engineer or Science Margins of the document: Setup the margins as Mirrored Margins and 2.5 cm for Top, bottom, and outside margins and 4 cm for outside margin. Delete this note before You print the thesis. This page starts at v because counting starts at the first page (but page numbers are not shown at the first pages). Delete this note before You print the thesis. v

vi

Page title right sided. Before :6 pt After : 12 pt Delete this note before You print the thesis. TABLE OF CONTENTS TABLE OF CONTENTS... vii ABBREVIATIONS... ix LIST OF TABLES... xi LIST OF FIGURES... xiii SUMMARY... xv ÖZET... xvii 1. INTRODUCTION... 1 1.1 Purpose of the Thesis... 1 1.2 Background... 1 1.3 Hypothesis... 2 2. INTEGRATED DATA... 3 2.1 Objectives... 3 2.2 Internet Server-Client Technology... 3 2.3 Web based Project Development... 4 3. FLOW FORECASTING... 5 3.1 Objectives... 5 3.2 Methodology... 5 3.2.1 Artificial neural networks... 5 3.2.2 Autoregressive-moving average models... 6 3.2.3 A process-based model: SWAT... 7 3.2.4 Cluster analysis... 8 3.3 Description Of Study Area... 10 3.4 Data... 10 3.5 Data Pre-processing... 13 4. CONCLUSION AND RECOMMENDATIONS... 15 4.1 Application of The Work... 15 REFERENCES... 17 APPENDICES... 21 CURRICULUM VITAE... 25 Page vii

viii

ABBREVIATIONS Line spacing: 1 Delete this note before You print the thesis. ABBREVIATIONS AIC ANN App BP CGI ESS GARCH GIS HCA Mbps St SWAT UMN : Akaike Information Criteria : Artificial Neural Network : Appendix : Backpropagation : Common Gateway Interface : Error sum-of-squares : Generalized Autoregressive Conditional Heteroskedasticity : Geographic Information Systems : Hierarchical Cluster Analysis : Megabits per second : Station : Soil and Water Assessment Tool : University of Minnesota ix

x

LIST OF TABLES Table 2.1: One-line table title with right sided columns... 4 Table 3.1: Gauging stations used in this study and their locations; a sample for twoline title, second line indented, column centred.... 11 Table 3.2: One-line table title with right sided columns... 11 Table 3.3: Full page table, centred columns... 12 Table 3.4: One-line table title with centred columns... 14 Table 3.5: One-line table title with sublevel columns, centred... 14 Table A.1 : Sample table in appendix.... 23 Page LIST OF TABLES Line spacing: 1 Second line indented at the same line with the beginning of the table title (shown by red line) Delete this note before You print the thesis. xi

xii

LIST OF FIGURES Figure 1.1 : The system types in modelling field.... 2 Figure 2.1 : Server-client architecture... 3 Figure 3.1 : Elements of a neuron cell, adapted from Cetin (2003).... 5 Figure 3.2 : Conceptual diagram of three-layer neural network model, a sample for two line title... 7 Figure 3.3 : Single linkage distance.... 8 Figure 3.4 : Flow-chart for compiling.... 9 Figure 3.5 : Location map for the study area, adapted from Url-4.... 10 Figure 3.6 : Experiment II: fluctuation in model performance (MSE).... 13 Figure A.1 : Environmental characteristics of Algarve basin.... 22 Page LIST OF FIGURES Line spacing: 1 Second line indented at the same line with the beginning of the figure title (shown by red line) Delete this note before You print the thesis. xiii

xiv

ENGLISH THESIS TITLE GOES HERE SUMMARY In last two decades Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. SUMMARY Line spacing: 1 This is an extended summary which is min 200 words to max 5 pages Delete this note before You print the thesis. xv

xvi

TURKISH THESIS TITLE ÖZET Son yıllarda Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed ÖZET Line spacing: 1 This is an extended summary in Turkish. Min 200 words to max 5 pages Delete this note before You print the thesis. xvii

xviii

1. INTRODUCTION Text goes here Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed 1.1 Purpose of the Thesis The two main objectives of this study are Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. 1.2 Background Following developments in computer network technology, people can manage and disseminate vast quantities of data quickly. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. 1

SAMPLE FIGURE Figure 2.1 : The system types in modelling field. 1.3 Hypothesis This study provided a unique opportunity to look at forecast issue with a wide range of model types and new clustering approach for model evaluation. eos et accusam et justo duo dolores et ea rebum. 2

2. INTEGRATED DATA 2.1 Objectives 2.2 Internet Server-Client Technology SAMPLE FIGURE Figure 2.2 : Server-client architecture. 3

2.3 Web based Project Development Table 2.1: One-line table title with right sided columns Parameter Column 2 Column 3 Column 3 Column 4 Latitude (ºN) 37.1743 37.1619 0.0 0.0 Longitude (ºW) -7.6804 7.6988 0.0 0.0 Altitude (m) 140-0.0 0.0 4

3. FLOW FORECASTING 3.1 Objectives The main focus of this chapter is to develop an Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. 3.2 Methodology The general definition of Neural Networks and Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. 3.2.1 Artificial neural networks The idea of NNs is based on biological neuron cells model of. SAMPLE FIGURE Figure 3.1 : Elements of a neuron cell, adapted from Cetin (2003). 5

3.2.2 Autoregressive-moving average models y t = φ. 1 y t 1 + εt (3.1) ε t is from a white-noise (stochastic process) which is random and normally distributed, y t is data value in year t, y t 1 is the predictor series. φ 1 is the AR(1) coefficient. It is estimated by the following formula which includes lag 1 and lag 2 autocorrelation coefficients of the residuals for AR model. φ = 1 ρ1 ρ. ρ 1 ρ 1 2 2 1 (3.2) This form is in the similar form of classical regression model in which y t is estimated by previous values and ε t is the regression residuals. Even the name autoregressive expresses the regression on self. Higher order AR models contain more lagged values inside the equation. Equations are right aligned and bold. First chapter #, second eq. # Helpful source: http://support.microsoft.com/kb/313017 Delete this note before You print 6 the thesis.

3.2.3 A process-based model: SWAT Figures are left aligned or centered. Line spacing: 1 Second line indented at the same line with the beginning of the figure title (shown by red line) Delete this note before You print the thesis. SAMPLE FIGURE Figure 3.2 : Conceptual diagram of three-layer neural network model, a sample for two line title. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. 7

Fitting error is defined as, Error = N 1 2 ( y i d i ) 2 (3.3) i= 1 where, N= number of output nodes, y i=calculated output, and d i = desired data value. Minimization is performed by calculating the gradient for each node at the output layer ( y d ) δ i = dσ i. i i (3.4) dσ i is the derivative of the sigmoid function applied at y i which is defined for each 3.2.4 Cluster analysis Cluster analysis (CA) is widely used to organize the observed data into meaningful structures and to find patterns. It has been widely used in atmospheric sciences. SAMPLE FIGURE dis tan ce Figure 3.3 : Single linkage distance. ( C, C ) min X C, X C d( X, X ) A B = A A B B A B. tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. 8

SAMPLE FIGURE Figure 3.4 : Flow-chart for compiling. 9

3.3 Description Of Study Area Alportel River, which is highlighted in Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. SAMPLE FIGURE Figure 3.5 : Location map for the study area, adapted from Url-4. 3.4 Data tempor invidunt ut labore et dolores. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. 10

Tables are left aligned or centered. Line spacing: 1 Second line indented at the same line with the beginning of the table title (shown by red line) Delete this note before You print the thesis. Reference at the end of sentence. Reference at the beginning of sentence. Numbered reference. Delete this note before You print the thesis. Table 3.1: Gauging stations used in this study and their locations; a sample for twoline title, second line indented, column centred. Parameter Column 2 Column 3 Column 4 Column 5 Column 6 Latitude (ºN) 37.174357 37.16192697 0.0 0.0 0.0 Longitude (ºW) -7.680442 7.6986348 0.0 0.0 0.0 Altitude (m) 140-0.0 0.0 0.0 diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna (Nelson, 1988). eos et accusam et justo duo dolores et ea rebum. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum [1]. Nelson (1988) analized lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Table 3.2: One-line table title with right sided columns Parameter Column 2 Column 3 Column 3 Column 4 Latitude (ºN) 37.174357 37.16192697 0.0 0.0 Longitude (ºW) -7.680442 7.6986348 0.0 0.0 Altitude (m) 140-0.0 0.0 tempor invidunt ut lab ore sit et dolore magna. 11

Table 3.3: Full page table, centred columns Parametre Kolon 2 Kolon 3 Alt kolon Kolon 4 Kolon 5 Alt Alt Alt Alt kolon kolon kolon kolon Satır 1-7.680442 7.6986348 0.00 0.00 0.00 12 12 Satır 2 140-0.50 0.00 0.00 0 0 Satır 3 37.174357 37.16192697 0.00 0.00 0.00 0 24 Satır 4 140-0.50 0.00 0.00 0 0 Satır 5 37.174357 37.16192697 0.00 0.00 0.00 0 24 Satır 6 140-0.50 0.00 0.00 0 0 Satır 7 37.174357 37.16192697 0.00 0.00 0.00 0 24 Satır 8 140-0.50 0.00 0.00 0 0 Satır 9 37.174357 37.16192697 0.00 0.00 0.00 0 24 Satır 10 140-0.50 0.00 0.00 0 0 Satır 11 37.174357 37.16192697 0.00 0.00 0.00 0 24 Satır 12 140-0.50 0.00 0.00 0 0 Satır 13 37.174357 37.16192697 0.00 0.00 0.00 0 24 Satır 14 140-0.50 0.00 0.00 0 0 Satır 15 37.174357 37.16192697 0.00 0.00 0.00 0 24 12

3.5 Data Pre-processing Data used in this study such as precipitation, humidity, temperature, streamflow, are Z initially normalized to [0.1-0.9] interval. equation (Demirel, 2004) it scores are computed by the following Z it i _ Sit S = σ i (3.5) _ S where it = input data for variable i at time t; S i S = arithmetic average of it ; standard S deviation of it. SAMPLE FIGURE Figure 3.6 : Experiment II: fluctuation in model performance (MSE). 13

Table 3.4: One-line table title with centred columns Parameter Column 2 Column 3 Column 3 Latitude (ºN) 37.174357 37.16192697 0.0 Longitude (ºW) -7.680442 7.6986348 0.0 Altitude (m) 140-0.0 River Type of soil RIBEIRA DE ALPORTEL Cambissolos and Litossolos RIBEIRA DE ALPORTEL Cambissolos and Litossolos 0.0 0.0 Table 3.5: One-line table title with sublevel columns, centred Parameter Column 2 Column 3 Latitude (ºN) 37.174357 Sublevel 1 Sublevel 2 Longitude (ºW) -7.680442 0.0 0.0 Altitude (m) 140 0.0 0.0 River RIBEIRA DE 0.0 0.0 ALPORTEL 14

4. CONCLUSION AND RECOMMENDATIONS The major purpose of this research was to achieve the two hypotheses mentioned in the first section. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed 4.1 Application of The Work In this thesis, the necessary steps for constructing an end-to-end streamflow forecasting system were discussed. These steps include the use Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est tempor invidunt ut lab ore sit et dolore magna. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et justo duo dolores et ea rebum. Stet clita kasd gub rgren, no sea takimata sanctus est Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut lab ore sit et dolore magna. Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. 15

16

REFERENCES REFERENCES Line spacing: 1 Author Surname, hint: sorted A to Z. Delete this note before You print the thesis. Abrahart, R. J., and See, L., 1998: Neural Network vs. ARMA Modelling: Constructing Benchmark Case Studies of River Flow Prediction. In GeoComputation 98. Proceedings of the Third International Conference on GeoComputation, University of Bristol, United Kingdom, 17 19 September (CD-ROM). Abrahart, R. J., and See, L., 2000: Comparing neural network and autoregressive moving average techniques for the provision of continuous river flow forecasts in two contrasting catchments, Hydrolog. Process., 14, 2157 2172. Acar, M. H. and Yılmaz, P., 1997. Effect of tetramethylthiuramdisulfide on the cationic poymerization of cylohexeneoxide, The 2 nd International Conferences on Advanced Polymers via Macromolecular Engineering, Orlando, Florida, USA, April 19-23. Box, G. E. P., and Jenkins, J. M., 1976: Time Series Analysis: Forecasting and Control. Holden-Day, San Francisco, CA. Burke, W.F. and Uğurtaş, G., 1974. Seismic interpretation of Thrace basin, TPAO internal report, Ankara, Turkey. Burlando, P., Rosso, R., Cadavid, L. G., and Salas, J. D., 1993: Forecasting of Short-term Rainfall Using ARMA Models. Journal of Hydrology. Vol. 144, no. 1-4, pp. 193-211. IOC-UNESCO, 1981. International bathymetric chart of the Mediterranean, Scale 1:1,000,000, 10 sheets, Ministry of Defence, Leningrad. LePichon, X., 1997. Kişisel görüşme. McCaffrey, R. and Abers, G., 1988. SYN3: A program for inversion of teleseismic body wave forms on microcomputers, Air Force Geophysics Laboratory Technical Report, AFGL-TR-88-0099, Hanscomb Air Force Base, MA. Nelson, M.R., 1988. Constraints on the seismic velocity structure of the crust and upper mantle beneath the eastern Tien Shan, Central Asia, PhD Thesis, MIT, Cambridge, MA. Roberts. S. and Jackson, J.A., 1991. Active normal faulting in central Greece: An overview, in The Geometry of Normal Faults, Spec. Publ. Geol. Soc. Lond., 56, p. 125-142, Eds. Roberts, A.M., Yielding, G. and Freeman, B., Blackwell Scientific Publications,Oxford. Sisaky, A., Golab, F. and Myer, B., 1989. Rust resistant potatoes, United Kingdom Patent, No: 2394783 dated 23.1.1989. 17

TS-40561, 1985. Çelik yapıların plastik teoriye göre hesap kuralları, Türk Standartları Enstitüsü, Ankara. Url-1 <http://www.mohid.com>, accessed at 29.06.2006. Url-2 <http://www.elet.polimi.it/ >, accessed at 10.01.2007. Vanden, G., Knapp, S., & Doe, J. (2001). Role of referenceelements in the selection of resources by psychology undergraduates. Journal of Bibliographic Research, 5, 117-123. Retrieved October 13, 2001, from http://jbr.org/articles.html In case, the internet document reference has very limited information (i.e., anonymous text), please indicate the full link and the date you retrieved the source. Delete this note before You print the thesis. 18

[1] Abrahart, R. J., and See, L., 1998: Neural Network vs. ARMA Modelling: Constructing Benchmark Case Studies of River Flow Prediction. In GeoComputation 98. Proceedings of the Third International Conference on GeoComputation, University of Bristol, United Kingdom, 17 19 September (CD-ROM). [2] IOC-UNESCO, 1981. International bathymetric chart of the Mediterranean, Scale 1:1,000,000, 10 sheets, Ministry of Defence, Leningrad. REFERENCES Line spacing: 1 Numbered reference style, hint: numbered as the sequence of source usage. Delete this note before You print the thesis. 19

20

APPENDICES APPENDIX A.1 : Regional Thematic Maps of Algarve, Adapted from Url-4 21

APPENDIX A.1 Figure A.1 : Environmental characteristics of Algarve basin. Multiple figures are numbered by small letters (a, b etc.) if each figure will be referenced otherwise a group name will be enough as used here. Pls check the Turkish template for an example in appendix part. Delete this note before You print the thesis. 22

Table A.1 : Sample table in appendix. Parameter Column 2 Column 3 Latitude (ºN) 37.174357 Sublevel 1 Sublevel 2 Longitude (ºW) -7.680442 0.0 0.0 Altitude (m) 140 0.0 0.0 River RIBEIRA DE 0.0 0.0 ALPORTEL 23

24

CURRICULUM VITAE Candidate s full name: Place and date of birth: Permanent Address: Universities and Colleges attended: Publications: Ganapuram S., Hamidov A., Demirel, M. C., Bozkurt E., Kındap U., and Newton A., 2007: Erasmus Mundus Scholar's Perspective On Water And Coastal Management Education In Europe. International Congress - River Basin Management, March 22-24, 2007 Antalya, Turkey. Cv with photo is recommended. Publications should be listed. Line spacing: 1 Delete this note before You print the thesis. 25