URBAN LAND COVER CHANGE DETECTION ANALYSIS AND MODELLING SPATIO-TEMPORAL GROWTH DYNAMICS USING REMOTE SENSING AND GIS TECHNIQUES “A CASE STUDY OF DHAKA, BANGLADESH” BAYES AHMED URBAN LAND COVER CHANGE DETECTION ANALYSIS AND MODELLING SPATIO-TEMPORAL GROWTH DYNAMICS USING REMOTE SENSING AND GIS TECHNIQUES “A CASE STUDY OF DHAKA, BANGLADESH” Dissertation Supervised by Dr. Pedro Latorre Carmona Professor, Institute of New Imaging Technologies (INIT) Universitat Jaume I (UJI), Castellón, Spain Dissertation Co-Supervised by Dr. Mário Caetano Professor, Instituto Superior de Estatística e Gestão de Informação (ISEGI) Universidade Nova de Lisboa (UNL), Lisbon, Portugal Dr. Edzer Pebesma Professor, Institute for Geoinformatics (ifgi) Westfälische Wilhelms-Universität (WWU), Münster, Germany Dr. Nilanchal Patel Professor, Department of Remote Sensing Birla Institute of Technology Mesra, Jharkhand, India February 2011 Candidate’s Declaration This is to certify that this research work is entirely my own and not of any other person, unless explicitly acknowledged (including citation of published and unpublished sources). All views and opinions expressed therein remain the sole responsibility of the author, and do not necessarily represent those of the institutes. It is hereby also declared that this dissertation or any part of it has not been submitted elsewhere for the award of any degree or diploma. _____________________________ BAYES AHMED Date: February 28, 2011 Dedication I want to dedicate this research work to My beloved Parents Md. Abdus Sattar and Mrs. Suariya Begum Abstract Dhaka, the capital of Bangladesh, has undergone radical changes in its physical form, not only in its vast territorial expansion, but also through internal physical transformations over the last decades. In the process of urbanization, the physical characteristic of Dhaka is gradually changing as open spaces have been transformed into building areas, low land and water bodies into reclaimed builtup lands etc. This new urban fabric should be analyzed to understand the changes that have led to its creation. The primary objective of this research is to predict and analyze the future urban growth of Dhaka City. Another objective is to quantify and investigate the characteristics of urban land cover changes (1989-2009) using the Landsat satellite images of 1989, 1999 and 2009. Dhaka City Corporation (DCC) and its surrounding impact areas have been selected as the study area. A fisher supervised classification method has been applied to prepare the base maps with five land cover classes. To observe the change detection, different spatial metrics have been used for quantitative analysis. Moreover, some post- classification change detection techniques have also been implemented. Then it is found that the ‘builtup area’ land cover type is increasing in high rate over the years. The major contributors to this change are ‘fallow land’ and ‘water body’ land cover types. In the next stage, three different models have been implemented to simulate the land cover map of Dhaka city of 2009. These are named as ‘Stochastic Markov (St_Markov)’ Model, ‘Cellular Automata Markov (CA_Markov)’ Model and ‘Multi Layer Perceptron Markov (MLP_Markov)’ Model. Then the best-fitted model has been selected based on various Kappa statistics values and also by implementing other model validation techniques. This is how the ‘Multi Layer Perceptron Markov (MLP_Markov)’ Model has been qualified as the most suitable model for this research. Later, using the MLP_Markov model, the land cover map of 2019 has been predicted. The MLP_Markov model shows that 58% of the total study area will be converted into builtup area cover type in 2019. The interpretation of depicting the future scenario in quantitative accounts, as demonstrated in this research, will be of great value to the urban planners and decision makers, for the future planning of modern Dhaka City. Key Words: Remote Sensing, Land Cover, Markov Chain, Cellular Automata, Multi Layer Perceptron Neural Network, Change Detection, Supervised Classification, GIS. i Acknowledgement At the outset, all praises belong to the almighty ‘Allah’, the most merciful, the most beneficent to all the creatures and their dealings. First of all, I would like to express my gratitude to the European Commission and Erasmus Mundus Consortium (Universitat Jaume I, Castellón, Spain; Westfälische Wilhelms-Universität, Münster, Germany and Universidade Nova de Lisboa, Portugal) for awarding me the Erasmus Mundus scholarship in Master of Science in Geospatial Technologies. It is a great opportunity for my lifetime experiences to study in the reputed universities of Europe. It is a great pleasure to acknowledge my sincere and greatest gratitude to my dissertation supervisor, Dr. Pedro Latorre Carmona, Professor, Institute of New Imaging Technologies, University Jaume I, Spain; for his untiring effort, careful supervision, thoughtful suggestions and enduring guidance at every stage of this research. This thesis would not be in its current shape without his continuous exertion and support. I am very grateful to my dissertation co-supervisors Dr. Mário Caetano, Dr. Edzer Pebesma and Dr. Nilanchal Patel; for accepting my thesis proposal at the very early stage and also for their valuable time and effort in contributing information and practical suggestions on numerous occasions. My heartfelt thanks goes to Dr. Raquib Ahmed, Professor and Director, Institute of Environmental Science, University of Rajshahi, Bangladesh; for his initial encouragement for conducting this research and helping me developing the research proposal. I also want to thank Dr. Filiberto Pla Bañón, Professor, Institute of New Imaging Technologies, Universitat Jaume I, Spain; for his time and some comments. I am pleased to extend my gratitude to Prof. Dr. Joaquín Huerta Guijarro, Dolores C. Apanewicz, Prof. Dr. Jorge Mateu and Dr. Christoph Brox; for their support and hospitality during my stay in Spain and Germany. My thanks and best wishes also conveyed to my classmates and lovely friends, from all over the world, for sharing their knowledge and giving me inspirations during the last eighteen months in Europe. Special thanks goes to Dipu da, Anik vai, Shiuli vabi, Irene, Mauri, Carlos, Sherzod, Neba, Shahin vai, Diyan vai, Pathak, Freska and Pearl for their patronage and helping me coping with this new and challenging European environment. Finally yet importantly, I want to express deep gratitude and indebtedness to my beloved parents for a life-long love and affection. I would also like to thank them for their continuous inspiration and encouragement, regarding the completion of this thesis and for their overall support throughout the years of my studies. ii Index of the Text Content Page No Abstract i Acknowledgement ii Index of the Text iii Index of Figures viii Index of Tables x Abbreviations and Acronyms xii Chapter 1: Introduction 1.1 Background of the Research 1 1.2 Statement of the Problem 1 1.3 Study Area Profile 5 1.4 Objectives of the Research 9 1.5 Research Hypotheses 9 1.5.1 Related to Objective 1 9 1.5.2 Related to Objective 2 9 1.6 Limitations of the Research 10 1.6.1 Collection of Satellite Images 10 1.6.2 Seasonal Variation 10 1.6.3 Collection of Reference Data 10 Chapter 2: Theoretical Framework and Methodology 2.1 Basic Terminologies 11 2.1.1 Remote Sensing (RS) 11 2.1.2 Geographic Information System (GIS) 11 2.1.3 Land 12 2.2 Literature Review 12 2.2.1 Examples Related to Land Cover Change Detection 13 2.2.2 Examples Related to Future Land Cover Prediction 14 2.3 Methodology of the Research 14 iii 2.3.1 Selection of Study Area 14 2.3.2 Problem Identification and Research Objectives 14 2.3.3 Data Collection 16 2.3.3.1 Satellite Images 16 2.3.3.2 Reference Data 18 2.3.3.3 Literature Review 18 2.3.4 Base Map Preparation and Accuracy Assessment 18 2.3.5 Change Detection Analysis 18 2.3.6 Model Calibration/ Simulation 18 2.3.7 Model Validation and Selection 19 2.3.8 Future Prediction 19 2.3.9 Directions for Future Planning 19 2.3.10 Report Writing 19 2.4 Tools Used for this Research 19 Chapter 3: Base Map Preparation and Accuracy Assessment 3.1 Image Enhancement 20 3.2 Composite Generation 20 3.3 Image Classification 22 3.3.1 Training Site Development 22 3.3.2 Signature Development 24 3.3.3 Classification 24 3.3.4 Generalization 24 3.4 Accuracy Assessment 26 3.4.1 Assessment Procedure 27 3.4.2 Results and Discussion 30 Chapter 4: Land Cover Change Detection Analysis 4.1 Change Detection 31 4.2 Terminologies 31 4.3 Analysis and Interpretation of the Change Detection Techniques 32 4.3.1 Number of Patches and Largest Patch Index 32 4.3.2 Edge Density and Mean Fractal Dimension Index 33 4.3.3 Mean Euclidean Nearest-Neighbour Distance and CPLAND 34 iv 4.3.4 Change in Area 35 4.3.5 Gains and Losses by Category 35 4.3.6 Contributors to Net Change Experienced by Builtup Area 36 4.3.7 Transition to Builtup Area 36 4.3.8 Gains and Losses in Land Cover Types 38 4.4 Summary of Land Cover Change Detection Analysis 38 Chapter 5: Stochastic Markov Model 5.1 Stochastic Process 40 5.2 Markov Chain 40 5.2.1 Markov Property 41 5.2.2 Transition Matrix for a Markov Chain 41 5.2.3 Example of Markov Chain 42 5.2.3.1 Weather Prediction 43 5.3 Stochastic Markov Model 43 Chapter 6: Cellular Automata Markov Model 6.1 Cellular Automata 48 6.1.1 What are Cellular Automata? 48 6.1.2 The Elements of Cellular Automata 48 6.1.3 The Cell Space 49 6.1.4 The Cell States 50 6.1.5 The Cell Neighbourhood 50 6.1.6 The Transition Rules 51 6.1.7 The Temporal Space 52 6.1.8 Mathematical Notation of Cellular Automata 52 6.1.9 Running a Simulation 53 6.2 Cellular Automata Markov Model 54 6.3 How CA_Markov Model Works? 54 6.3.1 Suitability Maps for Land Cover Classes 56 6.3.2 Preparing Suitability Maps 57 6.3.4 Future Prediction 63 v Chapter 7: Multi Layer Perceptron Markov Model 7.1 Artificial Neural Network 64 7.2 Basic Concept of Artificial Neural Network (ANN) 64 7.2.1 Types of Artificial Neural Network 65 7.3 Multi Layer Perceptron (MLP) 65 7.3.1 Input Layer 66 7.3.2 Hidden Layer 66 7.3.3 Output Layer 66 7.3.4 The Feed-Forward Concept of MLP Neural Network 67 7.3.5 Number of Nodes 68 7.3.6 Number of Training Samples and Iterations 68 7.4 Multi Layer Perceptron Markov Modelling 69 7.4.1 Testing Potential Explanatory Power 72 7.4.2 Transition Potential Modelling 73 7.4.3 Future Prediction 75 Chapter 8: Model Validation and Selection 8.1 Model Validation 76 8.1.1 Per Category Method 77 8.1.2 Fraction Correct 78 8.1.3 Fuzzy Sets 78 8.1.4 Fuzzy Kappa 79 8.2 Actual Base Maps vs. Simulated Maps 81 8.2.1 Base Map (2009) vs. St_Markov (2009) 81 8.2.1.1 Analysis of the Results of St_Markov 84 8.2.2 Base Map (2009) vs. CA_Markov (2009) 85 8.2.2.1 Analysis of the Results of CA_Markov 88 8.2.3 Base Map (2009) vs. MLP_Markov (2009) 89 8.3 Model Selection 91 Chapter 9: Future Prediction and Analysis 9.1 Future Prediction 92 9.1.1 Creating Boolean Images (2009) 92 9.1.2 Creating Distance Images (2009) 93 vi