A systematic review of tropical cyclone disaster management research using remote sensing and spatial analysis

Ocean & Coastal Management 146 (2017) 109e120

Contents lists available at ScienceDirect

Ocean & Coastal Management journal homepage: www.elsevier.com/locate/ocecoaman

A systematic review of tropical cyclone disaster management research using remote sensing and spatial analysis Muhammad Al-Amin Hoque a, b, *, Stuart Phinn a, Chris Roelfsema a a b

Remote Sensing Research Centre, School of Earth and Environmental Sciences, The University of Queensland, Brisbane, QLD, 4072, Australia Department of Geography and Environment, Jagannath University, Dhaka, 1100, Bangladesh

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 March 2017 Received in revised form 24 June 2017 Accepted 2 July 2017

Tropical cyclones are among the most dangerous and devastating natural disasters affecting life, property and environment. The use of remote sensing and spatial analysis has significantly increased to manage the on-ground impacts of these disasters with rapid advances in a wide range of data availability and processing techniques. This paper reviews recent studies of on-ground cyclone disaster management using remote sensing and spatial analysis in the context of response, recovery, prevention/reduction and preparedness to find out the key knowledge gaps for future research. The study used a systematic quantitative literature review technique to assess the past 21 years of research. Following the systematic search and developed selection criteria, the relevant original published articles on cyclone disaster management using remote sensing and spatial analysis were selected. The selected literature was then categorised and analysed based on the particular research focus. Our findings showed that most of the studies were concentrated in Asia (55%) and North and Central America (40%). The extensive use of remote sensing and spatial analysis started after 2004 and largely focused on the preparedness (34%) and prevention/reduction (32%) phases. Nearly all studies used the optical imagery, and the use of SAR imagery was limited. The object-based classification approach was rarely used under post-classification comparison techniques for overall tropical cyclone impact assessment and recovery. Very limited studies examined tropical cyclone risk assessment incorporating mitigation capacity and spatial multicriteria using the analytical hierarchy process (AHP). A simple modelling approach is required for producing detailed cyclone risk models. Most of the studies were conducted at the regional scale without validation of results. Cyclone risk mapping and modelling should consider future climate changes scenarios at the local scale. Future research is needed to cover reported knowledge gaps for improving cyclone disaster management. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Tropical cyclone Disaster management Remote sensing Spatial analysis Risk Climate change

Contents 1. 2.

3.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 2.1. Document search . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 2.2. Criteria for document selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 2.3. Document categorization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 2.4. Data interpretation and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Result and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 3.1. Spatio-temporal scales of research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 3.1.1. Spatial distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

* Corresponding author. Remote Sensing Research Centre, School of Earth and Environmental Sciences, The University of Queensland, Brisbane, QLD, 4072, Australia. E-mail addresses: [email protected] (M.A.-A. Hoque), [email protected] (S. Phinn), [email protected] (C. Roelfsema). http://dx.doi.org/10.1016/j.ocecoaman.2017.07.001 0964-5691/© 2017 Elsevier Ltd. All rights reserved.

110

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120

3.1.2. Identification of temporal pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Methods used in cyclone disaster management studies using remote sensing and spatial analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 3.2.1. Response and recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 3.2.2. Prevention/reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 3.2.3. Preparedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 3.3. Climate change scenarios and cyclone disaster management research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 3.4. Challenges in cyclone disaster management studies using remote sensing and spatial analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 Conclusions and future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Appendix 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

3.2.

4.

1. Introduction Tropical cyclones, hurricanes or typhoons, are major natural disturbances affecting life, property and the environment in many coastal areas across the world (Poulos, 2010; Gallina et al., 2016). These hydro-meteorological natural disasters generally form over tropical oceans, and are referred to as cyclones, typhoons and hurricanes based on their particular locations (Puotinen, 2007; Bobby, 2012). The destructive characteristics, for example, sustained high winds, storm surges and intensive rainfall, are typically associated with land falling tropical cyclones (Saxena et al., 2013; Hoque et al., 2017b). Occurrences of tropical cyclones are very common in coastal environments worldwide (Peduzzi et al., 2012). During 1970e2010, 637 major tropical cyclones were recorded globally (Weinkle et al., 2012). It is predicted that the frequency and intensity of tropical cyclones will likely increase in the coming decades under most climate change scenarios (Mendelsohn et al., 2012; Yin et al., 2013; Krishnamohan et al., 2014; Deo and Ganer, 2014). The overall impact of tropical cyclone disasters is very high at the global level compared to any other natural disasters (Li and Li, 2013). These disasters have been responsible for the loss of around 1.9 million lives along with large scale property and environmental damage over the last two centuries globally (Shultz et al., 2005; Hoque et al., 2017a). Tropical cyclones cause around US $5 billion worth of damage per year in the Gulf and east coast of United States (Burroughs, 2007). Mendelsohn et al. (2012) estimated that US $26 billion worth of damage occurs annually due to tropical cyclones worldwide. The impacts of tropical cyclones can be reduced by using appropriate management approaches. An effective cyclone disaster management plan is structured into four phases e response, recovery, prevention/reduction and preparedness (Fig. 1) (Khan, 2008; Joyce et al., 2009b; Islam and Chik, 2011). Response and recovery are considered in the post-disaster phase, while the predisaster phase is covered by prevention/reduction and preparedness. The response phase incorporates evacuation, relief, search and rescue, and the management of natural resources both during and immediately after the cyclone disaster to minimise the impact (Moe and Pathranarakul, 2006; Coppola, 2006). The overall impact assessment of cyclone disasters is an important process to deliver supportive information in the response phase. The restoration and reconstruction of cyclone disaster affected areas, in particular, monitoring the progress of debris removal and vegetation regrowth, and the reconstruction of settlements and structures, are included in the recovery phase (Joyce et al., 2009b; Rathfon et al., 2012). The prevention/reduction and preparedness phases include appropriate measures and planning that reduce the likelihood and impact of tropical cyclone disasters (Islam and Chik, 2011; Van Westen, 2013). Cyclone risk management is an essential process to generate required information for these two phases in the

Fig. 1. The tropical cyclone disaster management cycle (adapted from Khan, 2008).

context of cyclone vulnerability, hazard and mitigation assessment € ck et al., 2008; Li and Li, 2013; Fang et al., and modelling (Taubenbo 2016). The preparedness phase also includes the processes of cyclone tracking and forecasting to produce the required information for warning systems. However, this review is solely focussed on the on-ground impacts and use of remote sensing and spatial analysis to address this. Remote sensing and spatial analysis techniques can provide a valuable source of information in every phase of cyclone disaster management (Hussain et al., 2005; Joyce et al., 2009b; Wang et al., 2010). Repeated high spatial resolution (<5 m pixels) satellite imagery before and after the cyclone event is the most common use of remote sensing for cyclone disaster management (Joyce et al., 2009a; Martino et al., 2009; Klemas, 2009). Advances in remote sensing processing techniques provide methods to use these satellite images for assessing tropical cyclone disaster impacts and monitoring the progress of recovery (Yamazaki and Matsuoka, 2007; Joyce et al., 2009b; Hoque et al., 2016). Accordingly, the information about spatial location, type and intensity, percentage of area and structures affected is derived from impact assessment, while debris removal, vegetation regrowth, and reconstruction information is obtained from recovery assessment (Hoque et al., 2016). Satellite remote sensing and spatial analysis can also be used to assist risk management measures through estimation of cyclone hazard, vulnerability, and mitigation capacity assessment and modelling under likely future climate conditions (Li and Li, 2013; Van Westen, 2013; Appelquist and Balstrøm, 2015; Fang et al., 2016). Additionally, using satellite remote sensing and spatial analysis tools, tropical cyclones can be tracked and forenyi, 2012; Elsberry, 2014), however, this is casted (Roy and Kovorda

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120

not a focus of our paper. The use of remote sensing and spatial analysis for cyclone disaster management is increasing rapidly. Many studies have already been conducted using remote sensing and spatial analysis to provide appropriate information on the different parts of cyclone disaster management. However, a systematic comprehensive assessment of these studies is lacking in the current literature. This study systematically analysed the literature on tropical cyclone disaster management using remote sensing and spatial analysis to find out the knowledge gaps for future research. The study addressed four specific objectives: (1) evaluate the spatio-temporal scales of research; (2) provide an overview of methods used in existing research; (3) assess the inclusion of climate change scenarios in the current research; and (4) identify challenges in the current studies. 2. Materials and methods In this study, we conducted a systematic quantitative literature review method. The relevant literature was systematically searched and articles selected for evaluation following set criteria. The selected literature was categorised and critically assessed based on the study objectives. The materials and systematic processing flow used in this study are explained in the following sections, and presented in Fig. 2. The study used the methods outlined in

111

Pickering and Byrne (2014), following the systematic flow developed by Moher et al. (2009). 2.1. Document search We searched the literature focused on cyclone disaster management using remote sensing and spatial analysis from scholarly electronic databases including Web of Science, Scopus, Google Scholar and Science Direct. These databases were searched between January 2015 and May 2016. The keywords used for the search were ‘tropical cyclone’, ‘typhoon’, ‘hurricane’ and a combination of the following terms; ‘remote sensing’, ‘geographic information system (GIS)’, ‘spatial analysis’, ‘disaster management’, ‘impact assessment’, ‘damage assessment’, ‘recovery assessment’, ‘vulnerability mapping’, ‘hazard mapping’, ‘risk assessment’, ‘risk modelling’, ‘surge modelling’ and ‘climate change’. A total of 155 articles were found using these search terms and additional 16 articles were identified through citation lists and other sources. Out of these total 171 articles, we have selected 53 articles for this study following the criteria of selection those are explained in section 2.2 (Fig. 2). 2.2. Criteria for document selection We considered only peer reviewed papers describing the results

Fig. 2. The flowchart outlined the process and actions used in this study to organise the systematic quantitative literature review. n ¼ number of original research papers.

112

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120

of original research on tropical cyclone disaster management using remote sensing and spatial analysis, and published in academic journals over the last 21 years (May 1995eMay 2016). Review papers, book chapters and conference proceedings were not included (Fig. 2). Papers about cyclone tracking and forecasting using remote sensing were excluded since we focused on non-meteorological and oceanographic remotely sensed data. 2.3. Document categorization We primarily categorised the research papers according to their main focus on the specific phases of cyclone disaster management cycles, e.g. response, recovery, prevention/reduction and preparedness (Fig. 1). However, it was complex to classify based on the cyclone disaster management cycle since they are interrelated with each other. From each research paper, the following information was then recorded in a Microsoft Excel database: authors, year of publication, study location, journal name, management processes, data type and spatial resolution, detail methods, focus of the study, scale of study, validation of results, climate change scenario included or not, and challenges identified in the studies. The management processes were classified as impacts assessment, recovery assessment, risk assessment and risk modelling. Spatial data type (e.g. remote sensing and field data) and resolution of remote sensing data (e.g. high resolution: < 5 m, moderate resolution: 5e30 m and low resolution: > 30 m) were recorded (Hoque et al., 2016). The image classification techniques used were grouped into pixel-based and object-based. The processing methods used were recorded for impact assessment and recovery (e.g. change detection, visual interpretation, data mining and others), risk assessment (e.g. analytical hierarchy process (AHP), equal weight, ranking and rating), and risk modelling (e.g. hydrodynamic model, GIS based model and others). The focus of each study was classified for impacts assessment and recovery (e.g. vegetation, building, crop and others), risk assessment (vulnerability, hazard, mitigation and risk), and risk modelling (hazard modelling, vulnerability modelling and risk modelling). The scale of each study was also recorded as local, regional and global. The information was also recorded about whether the results of each study were validated or not. For studies conducting cyclone risk assessment and risk modelling, we also assessed if climate change scenarios were included or not. If included, then the local scenario or global scenarios were also noted. The challenges reported in each study about data availability, validation of results, processing and spatial resolution of remote sensing data were also recorded. 2.4. Data interpretation and analysis The recorded database was analysed using descriptive and statistical methods to evaluate the spatio-temporal dynamics of research, provide an overview of the methods used in the existing research and identify the challenges in the current research. The statistical graph was produced using Microsoft Excel to identify the temporal pattern of research. We used the ArcGIS 10 software for mapping the geographical distribution of research on tropical cyclone disaster management using remote sensing and analysis.

coastal countries from different continents (Appendix 1). The geographic distributions of the tropical cyclones described in these studies are outlined in Fig. 3. Out of these studies, 29 papers (55%) were from Asia, 21 papers (40%) were from North and Central America and the Caribbean, and 3 papers (7%) were from Australia (Fig. 3). Although some countries (Madagascar, Mozambique, Tanzania, Somalia, Cape Verde and Mauritius) from Africa are highly vulnerable to tropical cyclones, no study was identified from these countries. The studies were predominantly from USA (17 papers, 32%), followed by India (13 papers, 25%), Bangladesh (9 papers, 17%), and Australia (3 papers, 7%) and coincides within countries where the occurrence and intensity of tropical cyclones are very high. 3.1.2. Identification of temporal pattern Our study revealed that the extensive use of remote sensing and spatial analysis for cyclone disaster management started after 2004, although this study commenced after 1995 (Fig. 4). The rapid accessibility of remote sensing and spatial data as well as advancement in their processing techniques accelerated the use (Hoque et al., 2016). The focuses of studies were expanded in all of the phases of cyclone disaster management. According to the cyclone disaster management cycle, the focuses of 53 studies were reported in the preparedness phase (18 papers, 34%), followed by prevention/reduction (17 papers, 32%), response phase (11 papers, 21%), and recovery phase (7 papers, 13%). Only two research papers were identified during 1995e2005, both of these papers focused on recovery assessment of tropical cyclone impact. Since then, the uses of remote sensing and spatial analysis for cyclone disaster management has escalated rapidly, with 21 papers (40%) identified during 2006e2010, and 30 papers (57%) identified in the period of 2010e2016 (Fig. 4). In recent years (2010e2016), most of the studies have focused on prevention/reduction (12 papers) and preparedness (9 papers) compared to response (5 papers) and recovery (4 papers). This trend of focus may be the cause since prevention/reduction and preparedness are considered now as best management approaches to minimise the impacts of tropical cyclone disasters (Van Westen, 2013; Fang et al., 2016). The administration and policymakers of the cyclone affected countries are now giving priority to the generation of risk management and mitigation information. High spatial resolution satellite data, broad access to cyclone disaster spatial information and new assessment as well modelling techniques are making easier now to generate this kind information. 3.2. Methods used in cyclone disaster management studies using remote sensing and spatial analysis The 53 studies reviewed used a variety of methods with different datasets at different scales covering all the cyclone disaster management phases. The overview of these methods including data types and their spatial resolution, processing techniques, component of analysis, scale of study, and validation are discussed analytically and quantitatively to reveal knowledge gaps for future research. The discussion is followed by tropical cyclone disaster management phases: response and recovery, prevention/ reduction and preparedness.

3. Result and discussion 3.1. Spatio-temporal scales of research 3.1.1. Spatial distribution A total of 53 original journal articles were selected out of 171 that used the remote sensing and spatial analysis tools for cyclone disaster management (Appendix 1). These studies covered 30

3.2.1. Response and recovery Response and recovery are two crucial phases of cyclone disaster management. Remote sensing and spatial analysis are used to assess the impact and recovery of cyclone disasters and provide supporting information in these phases of management. The detailed methods used in the reviewed literature (18 papers) are summarised quantitatively in terms of data type, spatial resolution,

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120

113

Fig. 3. Geographic distribution of studies on tropical cyclone disaster management using remote sensing and spatial analysis by country, derived from systematic literature review. Tropical cyclone zones and paths were digitised using International Best Track Archive for Climate Stewardship (IBTrACS) tropical cyclone track data (1842e2015).

Fig. 4. Pattern of research focus and number of studies on cyclone disaster management using remote sensing and spatial analysis between 1995 and 2016.

impact assessment method, image classification approach, focus and scale of study in Table 1, and discussed in the following paragraphs. A variety of remote sensing and ground-based spatial datasets were used with different spatial resolution capabilities for assessing the impacts and recovery of tropical cyclone disasters (Table 1). These included optical and synthetic aperture radar (SAR) satellite images, digital elevation model (DEM) and field data. Our study revealed that most of studies used optical imagery (94%), some of them were combination of optical imagery, DEM and field datasets

(Philpott et al., 2008; Wang and Xu, 2009; Lou et al., 2012; Rathfon et al., 2012; Yuvaraj et al., 2015). Optical imagery is simple to understand and process (Joyce et al., 2009b). However, clouds are major obstacles in relation to obtaining the optical imagery immediately after any tropical cyclone event (Joyce et al., 2009a). In this context, SAR imagery is effective as this imagery is acquired using microwave active sensors that can penetrate the cloud. However, surprisingly only one study used this imagery for tropical cyclone impact assessment and recovery in the current literature (Klemas, 2009). Conversely, the spatial resolution of datasets plays

114

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120

Table 1 Overview of the detailed methods employed for assessing tropical cyclone impact and recovery that are reported in the18 original research papers (see appendix 1 for 18 papers in details). Category

Number of papers

Spatial data type Optical 17 SAR 1 DEM 5 Field data 7 Combined 6 Spatial resolution of dataset Very high (<5 m) 8 Moderate (5e30 m) 11 Low (>30 m) 1 Impact assessment methods Change detection 11 Visual interpretation 2 Data mining approach 1 Others 5 Image classification approach Pixel-based 10 Object-based 0 Focus Vegetation 11 Building/Settlement 7 Crops 2 Others 1 Scale of study Local 3 Regional 15 Global 0 Validation Yes 5 No 13

Percentage of papers 94% 6% 27 39% 33% 44% 61% 6% 61% 11% 6% 27% 56% 0% 61% 38% 11% 6% 17% 83% 0% 28% 72%

a vital role to derive the required information in detail. The uses of moderate to high spatial resolution datasets depend on the availability and budget as most of these datasets are not freely available. Nevertheless, currently, some satellite companies and international organisations freely provide high spatial resolution imagery prior to any disasters. In the current literature, 61% of studies used the moderate spatial resolution dataset, followed by 44% with high spatial resolution, and 6% with low spatial resolution. A range of processing methods e.g. change detection, visual interpretation, data mining etc. and image classification approaches were used for cyclone disaster impact assessment and recovery (Table 1). Change detection techniques were dominant in the current studies (61%). Two mainstream approaches are associated with the satellite-based change detection analysis, one is pre-classification, and another is post-classification. The studies used the pre-classification approaches included image differencing (Wang and Xu, 2010), change vector analysis (Wang and Xu, 2010), principle component analysis (Wang and Xu, 2009), and vegetation index differencing (Lee et al., 2008; Rodgers et al., 2009; Zhang et al., 2013; Bhowmik and Cabral, 2013). However, these approaches were found to be useful for providing a general overview of impact and recovery rather than a detailed overview. Conversely, post-classification change detection is a simple and extensively used technique for natural disaster impacts assessment and recovery due to its detailed change matrix as well as minimum external impacts generated by atmospheric and environmental differences (Wang and Xu, 2010). However, the use of this technique was limited to only three studies (Klemas, 2009; Wang and Xu, 2009, 2010) in the current review. Moreover, two image classification approaches e.g. pixel-based and object-based are mainly associated with the post-classification change detection technique. Pixel-based is a more traditional approach, works are based on per-pixel analysis, and provide the best result for low to moderate

spatial resolution image classification (Hussain et al., 2013). By contrast, object-based image analysis is an emerging approach, where image based mapping of land cover features is based on the hierarchical decomposition of the image into objects, with this approach being particularly suitable for high resolution satellite image classification (Blaschke et al., 2014). The studies followed the post-classification change detection approach; all of them used the pixel-based classification approach. Consequently, researchers can give more emphasis on object-based classification approach since the availability of high spatial resolution datasets are increasing. In contrast, only two studies employed the visual interpretation technique (Rathfon et al., 2012; Yu et al., 2013) and one study employed the data mining approach (Barnes, 2007) for assessing tropical cyclone impact and recovery. The overall impact and recovery information from entire landscapes are required for better cyclone disaster management. Unfortunately, most of the studies focused on single impact and recovery assessment in the landscape. Vegetation was the most dominant feature in these studies (Lee et al., 2008; Wang and Xu, 2009; Rodgers et al., 2009; Zhang et al., 2013; Bhowmik and n-Jua rez et al., 2014), followed by building/ Cabral, 2013; Negro settlement (Barnes, 2007; Rathfon et al., 2012; Lou et al., 2012; Yuvaraj et al., 2015), and crops (Philpott et al., 2008; Lou et al., 2012). One study was reported on change assessment in beach profiles and the sediment properties caused by tropical cyclones (Yu et al., 2013). The scale of the study area and validation of results are two important factors that need to be considered in the application of remote sensing and spatial analysis for cyclone impact assessment and recovery. Remote sensing can provide imagery for mapping cyclone impact and recovery at the local to continental scale (Zhang et al., 2013; Bhowmik and Cabral, 2013). However, the detailed and complete information depends on the selection of the scale of application. The current studies were predominant at the regional scale application (83%), only 17% of the studies were reported at the local scale and no studies were found at the global scale. On the contrary, validation of remote sensing data and their produced results are vital. However, interestingly, only 28% of the studies have validated their results. Though lots of challenges are associated with the validation of results, future studies can focus more on the validation of their results to enhance reliability and accuracy.

3.2.2. Prevention/reduction Prevention and reduction are core management activities to reduce the impacts of tropical cyclones. Risk assessment is an important process in this phase to provide supporting information that includes key infrastructure and areas at risk relative to spatial location, intensity of risk, factors liable for risk and appropriate mitigation options. The detailed methods followed in the reviewed literature (17 papers) for risk assessment are summarised quantitatively focusing on spatial data type, component and criteria of risk, multi-criteria processing, risk equation, scale of study and validation in Table 2 and discussed in the following paragraphs. Researchers used a wide range of spatial datasets for tropical cyclone risk assessment in the current literature. Remote sensing and field data were combined in many of studies (Yin et al., 2010; Darsan et al., 2013; Roy and Blaschke, 2015; Poompavai and Ramalingam, 2013). Most of the remote sensing data were in the form of optical images and DEM. Optical data were used to derive land use and land cover information as well as coastal erosion (Mahendra et al., 2011; Poompavai and Ramalingam, 2013) and DEM data were used to generate elevation, slope, storm surge height etc. (Rafiq et al., 2010; Yin et al., 2013; Darsan et al., 2013). Although airborne LiDAR DEM can provide the maximum amount

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120 Table 2 Overview of the detailed methods used for tropical cyclone risk assessment that were reported in the 17 original research papers (see appendix 1 for 17 papers in details). Category Spatial data type Remote sensing Field data Combined Component of risk Vulnerability Hazard Exposure Mitigation capacity Risk equation Risk ¼ vulnerability  hazard Risk ¼ hazard  vulnerability  exposure/ capacity Others Criteria Multi-criteria Single criteria Multi-criteria processing AHP Equal weight Ranking Rating Scale of study Local Regional Global Validation Yes No

Number of papers

Percentage of papers

15 12 11

88% 71% 65%

14 9 4 2

82% 53% 23% 12%

8 2

47% 12%

6

35%

10 7

59% 41%

3 4 2 1

30% 40% 20% 10%

1 14 2

6% 82% 12%

3 14

18% 82%

of accurate information, the use of these data was limited to only two studies (Shepard et al., 2012; Brakenridge et al., 2013). Additionally, field data were also used to generate different risk component criteria layers including precipitation, cyclone track, wind speed, population, cyclone shelter, sea level rise etc. The fundamental issues in cyclone risk assessment are to select the components of the risk and processing equation. Generally, four risk components e.g. hazard, vulnerability, exposure, and mitigation are used in an effective risk assessment procedure. Our systematic review found that 82% of the studies focused on vulnerability, followed by 53% on hazard, 23% on exposure and 12% on mitigation capacity. Several risk equations are found in the reviewed studies to assess the overall risk integrating these components of risk. The most used risk equation was,

Risk ¼ vulnerability  hazard

(1)

which was employed by eight studies (Rafiq et al., 2010; Gao et al., 2014; Xu et al., 2015). Though recent studies claimed that exposure should be included in the risk assessment procedure, very few studies included this component in the risk assessment procedure (Shepard et al., 2012; Yin et al., 2013). Additionally, in the evaluation of actual risk, it is essential to assess the mitigation capacity of the particular community, environment and resources. Therefore, mitigation capacity should be taken into account in risk assessment procedures. Currently, however, very limited studies have integrated this component in assessing tropical cyclone risk (Bobby, 2012; Poompavai and Ramalingam, 2013). Reliable and detailed tropical cyclone risk information is associated with the selection of sufficient criteria and the scale of study (Yin et al., 2013). Sufficient criteria need to be selected under each of the risk components (e.g. hazard, vulnerability, exposure and

115

mitigation capacity) for accurate risk assessment. Many studies conducted overall risk assessment but employed very limited criteria (Rafiq et al., 2010; Mahendra et al., 2011; Darsan et al., 2013). There are also seven studies where risk assessment was carried out which considering only one form of criteria (Khalid and Babb, 2008; Poulos, 2010; Brakenridge et al., 2013). Our review showed that the studies which focused only on vulnerability assessment employed in most cases sufficient criteria (Kumar and Kunte, 2012; Kunte et al., 2014). Conversely, most of the studies which focused on risk assessment were conducted at the regional scale (82%). While detailed local scale (1e5 km) risk information can help to select the best mitigation options, our review revealed that only one study was conducted at the local scale (Roy and Blaschke, 2015). We also found that two studies assessed the tropical cyclone at the global level (Taramelli et al., 2008; Brakenridge et al., 2013). Since several criteria are used in effective risk assessment procedures, spatial decision making processes such as weighting and ranking are required to integrate these criteria (Roy and Blaschke, 2015). In the current literature, several multi-criterial integrated procedures were used, for example, AHP, equal weight, ranking, and rating (Mahendra et al., 2011; Darsan et al., 2013; Poompavai and Ramalingam, 2013). Though some researchers argue about the usefulness of equal weight to integrate multi-criteria, many studies used equal weight (Mahendra et al., 2011; Kunte et al., 2014; Gao et al., 2014). AHP is considered to be more effective for weighting multi-criteria as this technique was endorsed by the expert in the context of the pairwise comparison matrix (Dewan, 2013b). However, very few studies were identified in this review where AHP was used for weighting the criteria to support the special decision making process (Poompavai and Ramalingam, 2013; Roy and Blaschke, 2015; Yin et al., 2013). Two studies employed ranking (Darsan et al., 2013) and one study used rating (Shepard et al., 2012) in the weighting of multi-criteria. Validation of risk assessment results generally increases the reliability and confidence in the prior of decision making processes. However, unfortunately, most of the studies (82%) skipped the validation procedures of their results. This may be because lots of challenges are associated with the validation process. Only three studies reported in the reviewed literature validated their risk assessment results (Roy and Blaschke, 2015; Poompavai and Ramalingam, 2013; Kunte et al., 2014). 3.2.3. Preparedness The identification of realistic cyclone risk scenarios for the future including level of risk with spatial location, key infrastructure and areas at risk as well as factors liable for risk is essential information in the preparedness phase. This information can be achieved by the modelling of tropical cyclone risk. The detailed methods used in the current literature (18 papers) for cyclone risk modelling are compiled quantitatively cantering on spatial data type, method of surge models, focus and scale of study, and validation in Table 3 and discussed in the following paragraphs. The spatial dataset used for tropical cyclone risk modelling in the current literature can be categorised into remote sensing and field data. All of the studies used the field-based spatial data derived from primary and secondary sources. DEM and optical images were used as a part of the remote sensing data. Terrain data in the context of DEM are vital for GIS based storm surge modelling (Ward et al., 2011; Ozcelik et al., 2012). While high-resolution DEM provides more accurate terrain data, most of studies used the low to moderate resolution DEM (Frazier et al., 2010; Ward et al., 2011; Lewis et al., 2013). A combined use of remote sensing and spatial field data was also found in some of the studies (Puotinen, 2007; Ward et al., 2011; Lewis et al., 2013).

116

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120

Table 3 Overview of detailed methods used for tropical cyclone risk modelling that was reported in the 18 original research papers (see appendix 1 for 18 papers in details). Category Spatial data type Remote sensing Field data Combined Method Hydrodynamic model GIS based model Others Focus Hazard/Storm surge modelling Vulnerability modelling Combined Scale of study Local Regional Global Validation Yes No

Number of papers

Percentage of papers

8 18 7

44% 100% 38%

10 5 3

56% 28% 17%

17 7 4

94% 39% 22%

0 16 2

0% 89% 11%

4 14

22% 78%

Various kinds of modelling techniques are identified in the current reviewed literature for modelling tropical cyclone risk at different scales (Arthur et al., 2008; Puotinen, 2007). Most of these are advanced hydrodynamic models (56%) and these include the CH3D-SSMS model (Condon and Peter Sheng, 2012), the Sea, Lake and Overland Surges from Hurricanes (SLOSH) model (Frazier et al., 2010), the IIT-D model (Lewis et al., 2013) and the ADCIRC model (Irish et al., 2010). Some of these (28%) use simple GIS based processing (Ward et al., 2011; Ozcelik et al., 2012; Saxena et al., 2013). The storm surge model is a vastly used model and output of this model is basically used as a basic input in the cyclone risk modelling process (Rao et al., 2007; Frazier et al., 2010). Most of the storm surge models reported in the current literature are produced using advanced modelling software, and these are at the regional scale (89%) (Jakobsen et al., 2006; Karim and Mimura, 2008; Condon and Peter Sheng, 2012; Lewis et al., 2013). To develop these models, advanced programming skills and lots of hydro-meteorological data are required. While two storm surge models were identified at the global scale (Lewis et al., 2013; Lewis et al., 2014), our study did not find any surge model developed at the local scale. However, the local scale model could deliver the risk information in detail and accurately find the best mitigation options to reduce the future impacts of tropical cyclones. Generally, risk is the product of hazard and vulnerability (Dewan, 2013a). Our systematic review revealed that most of the studies focused on storm surge modelling and identified future cyclone hazard locations and their levels of intensity. Cyclone hazards were modelled in these studies which took into account the 5e100 years return periods (Rao et al., 2007; Ward et al., 2011; Condon and Peter Sheng, 2012). Some of the reviewed studies also focused on vulnerability modelling (39%) (Rao et al., 2007; Irish et al., 2010; Frazier et al., 2010; Ozcelik et al., 2012; Saxena et al., 2013). Vulnerability was assessed in these studies in terms of landscape features and population. However, very few studies combined hazard and vulnerability modelling to assess the overall future risk (Frazier et al., 2010; Li and Li, 2013). While validations of model results influence reliability and confidence, it is a very challenging task to validate the storm surge models (Lewis et al., 2014). Our review found that only 22% of the studies validated their results (Jakobsen et al., 2006; Arthur et al., 2008; Lewis et al., 2013). Conversely, 78% of the studies did not discuss the validation of results in their studies. However, the development of surge models with their validations could provide

reliable future risk information to the decision makers for the purposes of finding proper mitigation measures. 3.3. Climate change scenarios and cyclone disaster management research Climate change induced extreme temperature and rises in sea level have significant impacts on the intensity and frequency of tropical cyclones and this impact is likely to increase in future (Condon and Peter Sheng, 2012; Krishnamohan et al., 2014; Fang et al., 2016). Therefore, cyclone disaster management studies should include future climate changes scenarios in the assessment and modelling of tropical cyclone risk. This section summarised the extent to which climate change scenarios were included in cyclone disaster management studies which focused on tropical cyclone risk assessment and modelling in Table 4. Our systematic review revealed that of the studies which employed cyclone risk assessment and modelling, only 34% of the studies included climate change scenarios in their assessment (Table 4). Out of these only 25% of the studies considered the temperature and sea level rise scenarios together (Karim and Mimura, 2008; Irish et al., 2010; Mousavi et al., 2011); the rest of them focused only on the inclusion of sea level rise scenarios (Frazier et al., 2010; Ward et al., 2011; Mahendra et al., 2011; Condon and Peter Sheng, 2012; Tebaldi et al., 2012; Shepard et al., 2012; Maloney and Preston, 2014). While local/regional climate change scenarios datasets are considered to be the best inputs for local/regional level assessment, most of the studies used global climate change scenarios applied at the local/regional scale (Mousavi et al., 2011; Condon and Peter Sheng, 2012). Although climate change scenarios, in particular, sea level rise scenarios need to be considered in assessing future realistic tropical cyclone risk, more than half of the studies (63%) in the current literature avoided these scenarios in their assessment. 3.4. Challenges in cyclone disaster management studies using remote sensing and spatial analysis The use of remote sensing and spatial analysis for generating required information for cyclone disaster management is increasing rapidly (Klemas, 2009). However, many challenges are associated with the provision of information in every phase of cyclone disaster management using these tools. The review identified specific challenges in the analyses employed in the reviewed studies, and these are presented quantitatively in Table 5 and elaborated in the following paragraph. The most commonly reported challenges were data availability and their spatial resolution, processing techniques and the validation of results (Table 5). For tropical cyclone impact assessment and recovery, it is essential to collect moderate to very high spatial resolution datasets of before, after and a few years after the cyclone event (Hoque et al., 2016; Bhowmik and Cabral, 2013). However, many of researchers found it was challenging to manage the required dataset for their area of interest within the required date and time (Wang and Xu, 2009). Additionally, most of the moderate Table 4 Status of climate change scenarios inclusion in the 35 papers (see appendix 1) which examined tropical cyclone risk assessment and modelling. Climate change Scenarios

Number of papers

Percentage of papers

Yes Data type Local/regional data Global data No

12

34%

3 9 23

25 75 65

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120 Table 5 Challenges reported in the 23 original research papers (see appendix 1) on cyclone disaster management using remote sensing and spatial analysis out of a total of 53 papers. Challenges

Number of papers

Percentage of papers

Data availability Spatial resolution Processing Validation

16 10 7 8

73% 43% 30% 35%

to high spatial resolution satellite imagery is not freely available and it is costly. These types of satellite imagery are also needed for risk assessment and modelling. Another challenging task was to collect high resolution DEM (Ward et al., 2011). The DEM are essential input for risk assessment and modelling (Fang et al., 2016). The researchers also found difficulties in the selection of processing techniques. For instance, many algorithms are available in the change detection technique, and it was hard to select the best one that could generate the best outputs (Wang and Xu, 2010). Conversely, the selection of component, criteria and their weighting, equation, scale and processing techniques for risk assessment and modelling were also found to be challenging. In addition, the accurate assessment of remote sensing data, validation of spatial models, and results were also deemed to be challenging in the current literature.

4. Conclusions and future directions The review demonstrated that despite various cyclone disaster management studies that have been conducted in across the world using remote sensing and spatial analysis, there are significant knowledge gaps in the current literature which require further research. Many studies have been conducted in different parts of the world on cyclone disaster management using remote sensing and spatial analysis. Though some coastal countries from the African continent are regularly affected by tropical cyclones, no study was identified from this continent in the current literature. SAR imagery can be obtained in cloudy conditions during and immediately after any cyclone event, when the scope of the optical imagery is limited. However, the use of SAR imagery for tropical cyclone impact assessment and recovery was very limited in the current studies. Object-based image analysis has received more attention as an effective image classification approach as the availability of moderate-to high-spatial resolution remote sensing data are growing. However, the object-based classification approach is rarely used in the current literature under post-classification comparison techniques for tropical cyclone impact assessment and recovery and future studies can consider this approach. Furthermore, there was much less research which assessed the

117

overall impact and recovery in the landscape caused by tropical cyclones using remote sensing and spatial analysis. The selection of scale, component and criteria are critical in the tropical cyclone risk assessment procedure. Currently, very few studies have examined tropical cyclone risk assessment incorporating spatial multi-criteria using AHP at the local scale. Mitigation capacity was also not considered as a part of the risk component in most of the studies, though this is considered to be a vital risk component in generating the actual risk information. In addition, most of the storm surge models used in the risk modelling procedure employed advanced modelling techniques at the regional scale. Consequently, a simple modelling approach is required at the local scale for producing detailed risk models. While global climate change scenarios have potential impacts on tropical cyclone activities, cyclone risk mapping and modelling should consider future climate change scenarios, in particular, sea level rises. However, comparatively few studies have included climate change scenarios in their assessment. Of the studies which considered climate change scenarios, most them were limited to global scenarios and applied at the regional scale. Additionally, the use of LiDAR data was limited in the current literature, accordingly, future studies could apply this data to every area of cyclone disaster management. The review also found some challenges associated with the current studies. However, these challenges will be overcome in the coming decades due to substantial advances which are expected in remote sensing and spatial analysis techniques. Cyclone disaster management will also be more effective using these tools. The more advanced satellite sensors are now being launched frequently at the government and private level. These sensors can provide remote sensing data at a much finer scale within the required dates and timeframes. Similarly, the availability of other spatial data are growing with advances in technology and increasing demand. Several national and international organisations are working to provide free access to required data for natural disaster management. The processing and validation procedures are likely to be easier with the advent of new technology and processing software. Acknowledgment We are grateful to Australian Government Research Training Program (RTP) scholarship and the University of Queensland (UQ), Australia for funding this research. We also thank the Remote Sensing Research Centre (RSRC), School of Earth and Environmental Sciences, UQ for providing research facilities. The authors also gratefully acknowledge the constructive comments and suggestions provided by two anonymous reviewers, editor, Dr. Sharif Ahmed Mukul and Shawkat Islam Sohel. Appendix 1

Details of 53 studies used in the systematic review on tropical cyclone disaster management using remote sensing and spatial analysis. Management phases

Author (year)

Journal

Country

Specific studies used in Tables (1e5)

Response (n ¼ 11)

Yuvaraj et al. (2015)  n-Jua rez et al. (2014) Negro Zhang et al. (2013) Lou et al. (2012) Szantoi et al. (2012)

Current Science Remote Sensing International Journal of Remote Sensing Natural Hazards International Journal of Applied Earth Observation and Geo-information Journal of Coastal Research

India Australia China China United States

✓ ✓ ✓ ✓ ✓

1

Klemas (2009)

United States

2

3

4

5

✓ ✓

✓ (continued on next page)

118

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120

(continued ) Management phases

Author (year)

Journal

Country

Specific studies used in Tables (1e5) 1

Wang and Xu (2010) Wang and Xu (2009) Lee et al. (2008) Philpott et al. (2008) Barnes et al. (2007) Recovery (n ¼ 7)

Prevention/ Reduction (n ¼ 17)

Dutta et al. (2015) Bhowmik and Cabral (2013) Rathfon et al. (2013) Yu et al. (2013) Rodgers et al. (2009) Chock (2005) Kushwaha (1997) Xu et al. (2015) Gao et al. (2014) Kunte et al. (2014) Brakenridge et al. (2013) Darsan et al. (2013) Rao et al. (2013) Roy and Blaschke (2015) Yin et al. (2013) Bobby (2012) Shepard et al. (2012) Kumar and Kunte (2012) Mahendra et al. (2011) Poulos (2010) Rafiq et al. (2010)

Preparedness (n ¼ 18)

Khalid and Babb (2008) Taramelli et al. (2008) Poompavai and Ramalingam (2013) Lewis et al. (2014) Maloney and Preston (2014) Lewis et al. (2013) Saxena et al. (2013) Li and Li (2013) Condon and Peter Sheng (2012) Ozcelik et al. (2012) Tebaldi et al. (2012) Mousavi et al. (2011) Frazier et al. (2010) Irish et al. (2010) Lin et al. (2010) Arthur et al. (2008) Karim and Mimura (2008) Puotinen (2007) Rao et al. (2007) Ward et al. (2011) Jakobsen et al. (2006)

Environmental Monitoring and Assessment Environmental Monitoring and Assessment Forest Ecology and Management Agriculture, Ecosystems and Environment IEEE Transactions On Geoscience and Remote Sensing Natural Hazards Earth Science Research Disaster Quaternary International Estuaries and Coasts Journal of Wind Engineering and Industrial Aerodynamics Geocarto International Natural Hazards Journal of Geophysical Research: Atmospheres Ocean & Coastal Management Natural Hazards Journal of Coastal Conservation Natural Hazards Geomatics, Natural Hazards and Risk Natural Hazards International Journal of Engineering Research & Technology Natural Hazards Natural Hazards Ocean & Coastal Management Natural Hazards Earth Resources and Environmental Remote Sensing Journal of Maps Natural Hazards and Earth System Science Journal of Indian Society of Remote Sensing Natural Hazards Climate Risk Management Quarterly Journal of the Royal Meteorological Society Natural Hazards Natural Hazards Natural Hazards International Journal of Climatology Environmental Research Letter Climate Change Applied Geography Ocean and Coastal Management Journal of Geophysical Research The Australian Journal of Emergency Management Global Environmental Change International Journal of Geographical Information Science Natural Hazards Natural Hazards Coastal Engineering Journal

References Appelquist, L.R., Balstrøm, T., 2015. Application of a new methodology for coastal multi-hazard-assessment & management on the state of Karnataka, India. J. Environ. Manag. 152, 1e10. Arthur, C., Schofield, A., Cechet, B., 2008. Assessing the impacts of tropical cyclones. Aust. J. Emerg. Manag. 23 (4), 14. Barnes, C.F., 2007. Image-driven data mining for image content Segmentation, classification, and attribution." Geoscience and Remote Sensing. IEEE Trans. 45 (9), 2964e2978. Barnes, C.F., Fritz, H., Yoo, J., 2007. Hurricane disaster assessments with imagedriven data mining in high-resolution satellite imagery. Geosci. Remote Sens. IEEE Trans. 45 (6), 1631e1640. Bhowmik, A.K., Cabral, P., 2013. Cyclone sidr impacts on the sundarbans floristic diversity. Earth Sci. Res. 2 (2), 62. Blaschke, T., Hay, G.J., Kelly, M., Lang, S., Hofmann, P., Addink, E., Queiroz Feitosa, R.,

United States United States Taiwan Mexico United States

✓ ✓ ✓ ✓ ✓

India & Bangladesh Bangladesh United States China United States United States

✓ ✓ ✓ ✓ ✓ ✓

India China China India India, Bangladesh, USA & Myanmar Trinidad and Tobago India Bangladesh China Bangladesh

✓

2

3

4

5 ✓ ✓ ✓

✓

✓ ✓ ✓ ✓ ✓ ✓ ✓

✓ ✓ ✓ ✓

✓ ✓ ✓ ✓

✓ ✓ ✓ ✓ ✓

✓ ✓ ✓ ✓ ✓

United States India India United States Pakistan

✓ ✓ ✓ ✓ ✓

✓ ✓ ✓ ✓ ✓

Granada West indies & Central America India India and Bangladesh USA India and Bangladesh

✓ ✓ ✓ ✓ ✓ ✓

✓ ✓ ✓ ✓ ✓ ✓

India China United States Myanmar United States United States United States United States United States Australia

✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

Bangladesh Australia

✓ ✓

✓ ✓

✓

India Indonesia Bangladesh

✓ ✓ ✓

✓ ✓ ✓

✓ ✓

✓ ✓ ✓

✓ ✓

✓

✓

et al., 2014. Geographic object-based image analysis e towards a new paradigm. ISPRS J. Photogramm. Remote Sens. 87 (0), 180e191. http://dx.doi.org/10.1016/ j.isprsjprs.2013.09.014. Bobby, S.A., 2012. Disaster Risk Index (DRI) for Tropical Cyclone of Bangladesh. Paper presented at the International Journal of Engineering Research and Technology. Brakenridge, G., Syvitski, J., Overeem, I., Higgins, S., Kettner, A., Stewart-Moore, J., Westerhoff, R., 2013. Global mapping of storm surges and the assessment of coastal vulnerability. Nat. Hazards 1e18. Burroughs, W.J., 2007. Climate Change: a Multidisciplinary Approach. Cambridge University Press, New York. Chock, G.Y., 2005. Modeling of hurricane damage for Hawaii residential construction. J. Wind Eng. Indus. Aerodyn. 93 (8), 603e622. Condon, A., Peter Sheng, Y., 2012. Evaluation of coastal inundation hazard for present and future climates. Nat. Hazards 62 (2), 345e373. http://dx.doi.org/ 10.1007/s11069-011-9996-0.

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120 Coppola, D.P., 2006. Introduction to International Disaster Management. Butterworth-Heinemann, Burlington,USA. Darsan, J., Asmath, H., Jehu, A., 2013. Flood-risk mapping for storm surge and tsunami at cocos bay (manzanilla), trinidad. J. Coast. Conserv. 17 (3), 679e689. Deo, A., Ganer, D., 2014. Tropical cyclone activity over the Indian ocean in the warmer climate. In: Mohanty, U.C. (Ed.), Monitoring and Prediction of Tropical Cyclones in the Indian Ocean and Climate Change. Springer, New Delhi, pp. 72e80. Dewan, A.M., 2013a. Hazards, risk, and vulnerability. In: Floods in a Megacity. Springer, pp. 35e74. Dewan, A.M., 2013b. Vulnerability and risk assessment. In: Floods in a Megacity. Springer, Dordrecht, pp. 139e177. Dutta, D., Das, P.K., Paul, S., Sharma, J.R., Dadhwal, V.K., 2015. Assessment of ecological disturbance in the mangrove forest of Sundarbans caused by cyclones using MODIS time-series data (2001e2011). Nat. Hazards 79 (2), 775e790. Elsberry, R., 2014. Advances in research and forecasting of tropical cyclones from 1963e2013. Asia-Pacific J. Atmos. Sci. 50 (1), 3e16. http://dx.doi.org/10.1007/ s13143-014-0001-1. Fang, Y., Yin, J., Wu, B., 2016. Flooding risk assessment of coastal tourist attractions affected by sea level rise and storm surge: a case study in Zhejiang Province, China. Nat. Hazards 84 (1), 611e624. http://dx.doi.org/10.1007/s11069-0162444-4. Frazier, T.G., Wood, N., Yarnal, B., Bauer, D.H., 2010. Influence of potential sea level rise on societal vulnerability to hurricane storm-surge hazards, Sarasota County, Florida. Appl. Geogr. 30 (4), 490e505. http://dx.doi.org/10.1016/ j.apgeog.2010.05.005. Gallina, V., Torresan, S., Critto, A., Sperotto, A., Glade, T., Marcomini, A., 2016. A review of multi-risk methodologies for natural hazards: consequences and challenges for a climate change impact assessment. J. Environ. Manag. 168, 123e132. Gao, Y., Wang, H., Liu, G., Sun, X., Fei, X., Wang, P., Lv, T., Xue, Z., He, Y., 2014. Risk assessment of tropical storm surges for coastal regions of China. J. Geophys. Res. Atmos. 119 (9), 5364e5374. Hoque, M.A.-A., Phinn, S., Roelfsema, C., Childs, I., 2016. Assessing tropical cyclone impacts using object-based moderate spatial resolution image analysis: a case study in Bangladesh. Int. J. Remote Sens. 37 (22), 5320e5343. http://dx.doi.org/ 10.1080/01431161.2016.1239286. Hoque, Muhammad Al-Amin, Phinn, Stuart, Roelfsema, Chris, Childs, Iraphne, 2017a. "Modelling tropical cyclone risks for present and future climate change scenarios using geospatial techniques.". Int. J. Digital Earth. http://dx.doi.org/ 10.1080/17538947.2017.1320595. Hoque, Muhammad Al-Amin, Phinn, Stuart, Roelfsema, Chris, Childs, Iraphne, 2017b. "Tropical cyclone disaster management using remote sensing and spatial analysis: a review.". Int. J. Disaster Risk Reduct. 22, 345e354. http://dx.doi.org/ 10.1016/j.ijdrr.2017.02.008. Hussain, M., Arsalan, M.H., Siddiqi, K., Naseem, B., Rabab, U., 2005. Emerging Geoinformation Technologies (GIT) for natural disaster management in Pakistan: an overview. In: Kurnaz, S., Ince, F., Inbasioglu, S., Basturk, S. (Eds.), RAST 2005: Proceedings of the 2nd International Conference on Recent Advances in Space Technologies. Hussain, M., Chen, D., Cheng, A., Wei, H., Stanley, D., 2013. Change detection from remotely sensed images: from pixel-based to object-based approaches. ISPRS J. Photogramm. Remote Sens. 80, 91e106. Irish, J.L., Frey, A.E., Rosati, J.D., Olivera, F., Dunkin, L.M., Kaihatu, J.M., Ferreira, C.M., Edge, B.L., 2010. Potential implications of global warming and barrier island degradation on future hurricane inundation, property damages, and population impacted. Ocean Coast. Manag. 53 (10), 645e657. Islam, S.M.T., Chik, Z., 2011. Disaster in Bangladesh and management with advanced information system. Disaster Prev. Manag. 20 (5), 521e530. http://dx.doi.org/ 10.1108/09653561111178952. Jakobsen, F., Azam, M.H., Zahid Ahmed, M.M., Mahboob-ul-Kabir, M., 2006. Cyclone storm surge levels along the Bangladeshi coastline in 1876 and 1960e2000. Coast. Eng. J. 48 (03), 295e307. Joyce, K.E., Belliss, S.E., Samsonov, S.V., McNeill, S.J., Glassey, P.J., 2009a. A review of the status of satellite remote sensing and image processing techniques for mapping natural hazards and disasters. Prog. Phys. Geogr. 33 (2), 183e207. Joyce, K.E., Wright, K.C., Samsonov, S.V., Ambrosia, V.G., 2009b. Remote Sensing and the Disaster Management Cycle. INTECH Open Access Publisher, Rijeka, Croatia. Karim, M.F., Mimura, N., 2008. Impacts of climate change and sea-level rise on cyclonic storm surge floods in Bangladesh. Glob. Environ. Change 18 (3), 490e500. http://dx.doi.org/10.1016/j.gloenvcha.2008.05.002. Khalid, F., Babb, R., 2008. Hazard and risk assessment from hurricane ivan (2004) in Grenada using geographical information systems and remote sensing. J. Maps 4 (Suppl. 1), 4e10. Khan, M.S.A., 2008. Disaster preparedness for sustainable development in Bangladesh. Disaster Prev. Manag. 17 (5), 662e671. Klemas, V.V., 2009. The role of remote sensing in predicting and determining coastal storm impacts. J. Coast. Res. 1264e1275. Krishnamohan, K., Mohanakumar, K., Joseph, P., 2014. Climate change in tropical cyclones and monsoon depressions of North Indian Ocean. In: Mohanty, U., Mohapatra, M., Singh, O., Bandyopadhyay, B., Rathore, L. (Eds.), Monitoring and Prediction of Tropical Cyclones in the Indian Ocean and Climate Change. Springer, pp. 33e39. Kumar, A.A., Kunte, P.D., 2012. Coastal vulnerability assessment for Chennai, east

119

coast of India using geospatial techniques. Nat. Hazards 64 (1), 853e872. http:// dx.doi.org/10.1007/s11069-012-0276-4. Kunte, P.D., Jauhari, N., Mehrotra, U., Kotha, M., Hursthouse, A.S., Gagnon, A.S., 2014. Multi-hazards coastal vulnerability assessment of Goa, India, using geospatial techniques. Ocean Coast. Manag. 95, 264e281. Kushwaha, S., 1997. Environmental monitoring and cyclone imapct assessment on Sriharikota Island, India. Geocarto Int. 12 (2), 55e62. Lee, M.-F., Lin, T.-C., Vadeboncoeur, M.A., Hwong, J.-L., 2008. Remote sensing assessment of forest damage in relation to the 1996 strong typhoon Herb at Lienhuachi Experimental Forest, Taiwan. For. Ecol. Manag. 255 (8), 3297e3306. Lewis, M., Bates, P., Horsburgh, K., Neal, J., Schumann, G., 2013. A storm surge inundation model of the northern Bay of Bengal using publicly available data. Q. J. R. Meteorol. Soc. 139 (671), 358e369. http://dx.doi.org/10.1002/qj.2040. Lewis, M., Horsburgh, K., Bates, P., 2014. Bay of Bengal cyclone extreme water level estimate uncertainty. Nat. Hazards 72 (2), 983e996. http://dx.doi.org/10.1007/ s11069-014-1046-2. Li, K., Li, G.S., 2013. Risk assessment on storm surges in the coastal area of Guangdong Province. Nat. Hazards 68, 1129e1139. Lin, N., Emanuel, K., Smith, J., Vanmarcke, E., 2010. Risk assessment of hurricane storm surge for New York City. J. Geophys. Res. Atmos. (1984e2012) 115 (D18), 1e11. Lou, W.P., Chen, H.Y., Shen, X.L., Sun, K., Deng, S.R., 2012. Fine assessment of tropical cyclone disasters based on GIS and SVM in Zhejiang Province, China. Nat. Hazards 64 (1), 511e529. http://dx.doi.org/10.1007/s11069-012-0254-x. Mahendra, R.S., Mohanty, P.C., Bisoyi, H., Kumar, T.S., Nayak, S., 2011. Assessment and management of coastal multi-hazard vulnerability along the CuddaloreVillupuram, east coast of India using geospatial techniques. Ocean Coast. Manag. 54 (4), 302e311. http://dx.doi.org/10.1016/j.ocecoaman.2010.12.008. Maloney, M.C., Preston, B.L., 2014. A geospatial dataset for US hurricane storm surge and sea-level rise vulnerability: development and case study applications. Clim. Risk Manag. 2, 26e41. Martino, L., Ulivieri, C., Jahjah, M., Loret, E., 2009. Remote sensing and GIS techniques for natural disaster monitoring. In: Olla, P. (Ed.), Space Technologies for the Benefit of Human Society and Earth. Springer, pp. 331e382. Mendelsohn, R., Emanuel, K., Chonabayashi, S., Bakkensen, L., 2012. The impact of climate change on global tropical cyclone damage. Nat. Clim. Change 2 (3), 205e209. Moe, T.L., Pathranarakul, P., 2006. An integrated approach to natural disaster management: public project management and its critical success factors. Disaster Prev. Manag. 15 (3), 396e413. Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., P. G, 2009. Preferred reporting items for systematic reviews and meta-analyses: the prisma statement. Ann. Intern. Med. 151 (4), 264e269. http://dx.doi.org/10.7326/0003-4819-151-4200908180-00135. Mousavi, M.E., Irish, J.L., Frey, A.E., Olivera, F., Edge, B.L., 2011. Global warming and hurricanes: the potential impact of hurricane intensification and sea level rise on coastal flooding. Clim. Change 104 (3e4), 575e597. n-Jua rez, R.I., Chambers, J.Q., Hurtt, G.C., Annane, B., Cocke, S., Powell, M., Negro Stott, M., Goosem, S., Metcalfe, D.J., Saatchi, S.S., 2014. Remote sensing assessment of forest disturbance across complex mountainous terrain: the pattern and severity of impacts of tropical cyclone yasi on Australian rainforests. Remote Sens. 6 (6), 5633e5649. Ozcelik, C., Gorokhovich, Y., Doocy, S., 2012. Storm surge modelling with geographic information systems: estimating areas and population affected by cyclone Nargis. Int. J. Climatol. 32 (1), 95e107. http://dx.doi.org/10.1002/joc.2252. Peduzzi, P., Chatenoux, B., Dao, H., De Bono, A., Herold, C., Kossin, J., Mouton, F., Nordbeck, O., 2012. Global trends in tropical cyclone risk. Nat. Clim. Change 2 (4), 289e294. Philpott, S.M., Lin, B.B., Jha, S., Brines, S.J., 2008. A multi-scale assessment of hurricane impacts on agricultural landscapes based on land use and topographic features. Agric. Ecosyst. Environ. 128 (1), 12e20. Pickering, C., Byrne, J., 2014. The benefits of publishing systematic quantitative literature reviews for PhD candidates and other early-career researchers. High. Educ. Res. Dev. 33 (3), 534e548. Poompavai, V., Ramalingam, M., 2013. Geospatial analysis for coastal risk assessment to cyclones. J. Indian Soc. Remote Sens. 1e20. Poulos, H., 2010. Spatially explicit mapping of hurricane risk in New England, USA using ArcGIS. Nat. Hazards 54 (3), 1015e1023. Puotinen, M., 2007. Modelling the risk of cyclone wave damage to coral reefs using GIS: a case study of the Great Barrier Reef, 1969e2003. Int. J. Geogr. Inf. Sci. 21 (1), 97e120. Rafiq, L., Blaschke, T., Zeil, P., 2010. Application of satellite derived information for disaster risk reduction; vulnerability assessment for SW-coast of Pakistan. In: Michel, U., Civco, D.L. (Eds.), Earth Resources and Environmental Remote Sensing-gis Applications. Rao, A., Chittibabu, P., Murty, T., Dube, S., Mohanty, U., 2007. Vulnerability from storm surges and cyclone wind fields on the coast of Andhra Pradesh, India. Nat. Hazards 41 (3), 515e529. Rao, V.R., Subramanian, B.R., Mohan, R., Kannan, R., Mageswaran, T., Arumugam, T., Rajan, B., 2013. Storm surge vulnerability along ChennaieCuddalore coast due to a severe cyclone THANE. Nat. Hazards 68 (2), 453e465. http://dx.doi.org/ 10.1007/s11069-013-0630-1. Rathfon, D., Davidson, R., Bevington, J., Vicini, A., Hill, A., 2012. Quantitative assessment of post-disaster housing recovery: a case study of Punta Gorda, Florida, after Hurricane Charley. Disasters 37 (2), 333e355.

120

M.A.-A. Hoque et al. / Ocean & Coastal Management 146 (2017) 109e120

Rathfon, D., Davidson, R., Bevington, J., Vicini, A., Hill, A., 2013. Quantitative assessment of post-disaster housing recovery: a case study of Punta Gorda, Florida, after Hurricane Charley. Disasters 37 (2), 333e355. Rodgers III, J.C., Murrah, A.W., Cooke, W.H., 2009. The impact of hurricane katrina on the coastal vegetation of the weeks bay reserve, Alabama from NDVI data. Estuaries Coasts 32 (3), 496e507. nyi, R., 2012. Tropical cyclone track forecasting techniques d A Roy, C., Kovorda review. Atmos. Res. 104e105 (0), 40e69. http://dx.doi.org/10.1016/ j.atmosres.2011.09.012. Roy, D.C., Blaschke, T., 2015. Spatial vulnerability assessment of floods in the coastal regions of Bangladesh. Geomat., Nat. Hazards Risk 6 (1), 21e44. Saxena, S., Purvaja, R., Suganya, G.M.D., Ramesh, R., 2013. Coastal hazard mapping in the Cuddalore region, South India. Nat. Hazards 66 (3), 1519e1536. http:// dx.doi.org/10.1007/s11069-012-0362-7. Shepard, C.C., Agostini, V.N., Gilmer, B., Allen, T., Stone, J., Brooks, W., Beck, M.W., 2012. Assessing future risk: quantifying the effects of sea level rise on storm surge risk for the southern shores of Long Island, New York. Nat. Hazards 60 (2), 727e745. http://dx.doi.org/10.1007/s11069-011-0046-8. Shultz, J.M., Russell, J., Espinel, Z., 2005. Epidemiology of tropical cyclones: the dynamics of disaster, disease, and development. Epidemiol. Rev. 27 (1), 21e35. Szantoi, Z., Malone, S., Escobedo, F., Misas, O., Smith, S., Dewitt, B., 2012. A tool for rapid post-hurricane urban tree debris estimates using high resolution aerial imagery. Int. J. Appl. Earth Observ. Geoinform. 18, 548e556. Taramelli, A., Melelli, L., Pasqui, M., Sorichetta, A., 2008. Estimating hurricane hazards using a GIS system. Nat. Hazards Earth Syst. Sci. 8, 839e854. € ck, H., Post, J., Roth, A., Zosseder, K., Strunz, G., Dech, S., 2008. A conceptual Taubenbo vulnerability and risk framework as outline to identify capabilities of remote sensing. Nat. Hazards Earth Syst. Sci. 8 (3), 409e420. Tebaldi, C., Strauss, B.H., Zervas, C.E., 2012. Modelling sea level rise impacts on storm surges along US coasts. Environ. Res. Lett. 7 (1), 014032. Van Westen, C.J., 2013. Remote sensing and GIS for natural hazards assessment and disaster risk management. In: Schroder, J.F., Bishop, M.P. (Eds.), Treatise on Geomorphology. Academic Press, Elsevier, San Diego, pp. 259e298. Wang, F., Wen, R., Zhong, S., 2010. Key issues in mapping technologies for disaster management. In: Paper Presented at the Information Engineering and

Computer Science (ICIECS), 2010 2nd International Conference on. Wang, F., Xu, Y.J., 2010. Comparison of remote sensing change detection techniques for assessing hurricane damage to forests. Environ. Monit. Assess. 162 (1e4), 311e326. http://dx.doi.org/10.1007/s10661-009-0798-8. Wang, F.G., Xu, Y.J., 2009. Hurricane Katrina-induced forest damage in relation to ecological factors at landscape scale. Environ. Monit. Assess. 156 (1e4), 491e507. http://dx.doi.org/10.1007/s10661-008-0500-6. Ward, P., Marfai, M., Yulianto, F., Hizbaron, D., Aerts, J., 2011. Coastal inundation and damage exposure estimation: a case study for Jakarta. Nat. Hazards 56 (3), 899e916. Weinkle, J., Maue, R., Pielke Jr., R., 2012. Historical global tropical cyclone landfalls. J. Clim. 25 (13), 4729e4735. Xu, X., Sun, D., Guo, T., 2015. A systemic analysis of typhoon risk across China. Nat. Hazards 77 (1), 461e477. Yamazaki, F., Matsuoka, M., 2007. Remote sensing technologies in post-disaster damage assessment. J. Earthq. Tsunami 1 (03), 193e210. Yin, J., Xu, S., Wang, J., Zhong, H., Hu, Y., Yin, Z., Wang, K., Zhang, X., 2010. Vulnerability assessment of combined impacts of sea level rise and coastal flooding for China's coastal region using remote sensing and GIS. In: Paper Presented at the Geoinformatics, 2010 18th International Conference on. Yin, J., Yin, Z., Xu, S., 2013. Composite risk assessment of typhoon-induced disaster for China's coastal area. Nat. Hazards 69 (3), 1423e1434. http://dx.doi.org/ 10.1007/s11069-013-0755-2. Yu, F., Switzer, A.D., Lau, A.Y.A., Yeung, H.Y.E., Chik, S.W., Chiu, H.C., Huang, Z., Pile, J., 2013. A comparison of the post-storm recovery of two sandy beaches on Hong Kong Island, southern China. Quat. Int. 304 (0), 163e175. http://dx.doi.org/ 10.1016/j.quaint.2013.04.002. Yuvaraj, E., Dharanirajan, K., Saravanan, Narshimulu, G., 2015. Post-disaster assessment of impact of cyclone lehar in south Andaman island. Curr. Sci. 108 (1), 85e90. Zhang, X.Y., Wang, Y., Jiang, H., Wang, X.M., 2013. Remote-sensing assessment of forest damage by Typhoon Saomai and its related factors at landscape scale. Int. J. Remote Sens. 34 (21), 7874e7886. http://dx.doi.org/10.1080/ 01431161.2013.827344.