An Informed System Development Approach to Tropical Cyclone Track and Intensity Forecasting.

• Superior document: Linköping Studies in Science and Technology. Dissertations Series ; v.1734
• Author: Roy, Chandan.
• Place / Publishing House: Linköping : : Linkopings Universitet,, 2015.; {copy}2016.
• Year of Publication: 2015
• Edition: 1st ed.
• Language: English
• Series: Linköping Studies in Science and Technology. Dissertations Series
• Physical Description: 1 online resource (166 pages)

Table of Contents

• Intro
• Abstract
• Populärvetenskaplig sammanfattning
• Acknowledgements
• Contents
• Chapter 1 Introduction
• 1.1 Problem description
• 1.2 Research objective
• 1.3 Studies conducted in this research
• 1.4 Contributions
• 1.5 Organization of the thesis
• Chapter 2 Background
• 2.1 TCs in the Bay of Bengal
• 2.2 Factors governing TC track and intensity development
• 2.3 Technical processes involved in TC forecasting and warning
• 2.3.1 Data collection
• 2.3.2 TC track forecasting
• 2.3.3 TC Intensity forecasting
• 2.3.4 Warning message formulation and dissemination
• 2.4 Community response to warning
• 2.4.1 Incorporation of human perception into cyclone warning
• 2.5 Cyclone early warning system in Bangladesh
• 2.6 TC track and intensity forecasting using ANN
• 2.7 Biologically based ANN techniques for image processing
• 2.7.1 Neocognitron
• 2.7.2 Convolutional Neural Networks
• 2.7.3 Saliency based visual attention models
• 2.7.4 ANNs used in this research
• 2.8 Exploratory study on TC movement direction prediction
• 2.8.1 Data and method
• 2.8.2 Training and testing of the network
• Chapter 3 Methodological considerations
• 3.1 Central and peripheral parts of this research
• 3.2 Technological paradigm of this research
• Chapter 4 Method
• 4.1 Study area
• 4.2 Eliciting providers' and receivers' views
• 4.2.1 Interview among the meteorologists
• 4.2.2 Questionnaire survey among the residents in the coastal areas
• 4.3 TC track and intensity forecasting technique development
• 4.3.1 Efficiency of the simulation tool with respect to 2D image processing
• 4.3.2 Testing assumptions
• 4.3.3 Match between network-generated and expected outputs
• 4.3.4 Datasets used for training and testing
• 4.3.5 Network structure
• 4.3.6 Information processing in the network
• 4.3.7 Training of the networks.
• 4.3.8 Training performance
• 4.3.9 Activation-based receptive field analysis
• 4.3.10 Testing of the networks
• Chapter 5 Results
• 5.1 Results elicited from warning providers
• 5.1.1 TC forecasting
• 5.1.2 Warning message formulation and dissemination
• 5.1.3 Limitations and future development plans
• 5.2 Results elicited from warning receivers
• 5.2.1 Warning message reception
• 5.2.2 Warning message interpretation
• 5.2.3 Response to evacuation orders
• 5.2.4 Key reasons for non-evacuation
• 5.2.5 Satisfaction with the warnings and suggestion for improvement
• 5.3 Results obtained during systematic parameter testing
• 5.4 TC intensity forecasting performance
• 5.4.1 Pattern of intensity forecasting error
• 5.5 Combined TC track and intensity forecasting performance
• Chapter 6 Discussion
• 6.1 Problems associated with cyclone early warning in Bangladesh
• 6.2 TC track and intensity forecasting using biologically based ANNs
• 6.2.1 TC Intensity forecasting
• 6.2.2 Graphically-presented TC track and intensity forecasting-ongoing work
• 6.2.3 Prediction performance improvement
• Chapter 7 Conclusions and future work
• Chapter 8 References
• Chapter 9 Articles
• Appendix A Cyclone track forecasting using biologically based ANNs
• Appendix B TC signaling system for the maritime ports
• Appendix C TC signaling system for the river ports
• Appendix D Questions used for the in-depth interview
• Appendix E Questionnaire used for the survey
• Appendix F Per-recorded warning messages
• Appendix G Generalized TC signaling system.