An Object-Based Image Analysis of WorldView-3 Image for Urban Flood Vulnerability Assessment and Dissemination Through ESRI Story Maps

Abstract

The utilization of remote sensing and GIS is increasing in importance across the world for disaster preparedness, assessments, mitigation, and governance. A major focus is on extracting and disseminating useful information from satellite and airborne data sources for effective planning and decision making during and after disasters. This study presents a case study from Ko-Rian, a sub-district in Ayutthaya, Thailand which is frequently hit by monsoon flooding. First, an approach is demonstrated to extract buildings and roads from very-high-resolution satellite images. The results showed that use of selected multiple indices with object-based image analysis produced excellent results in mixed urban environment. Secondly, a procedure and framework are proposed for geodatabase modeling of urban flood vulnerability and disseminating information to a wider community through ESRI story maps. The proposed approach using virtual reality technologies such as Google Street View for assessing physical vulnerability and ESRI story maps to share the information can significantly reduce the time and costs of the traditional survey. The techniques and approach presented in this study can be useful to develop detailed vulnerability profile for effective decision making and dynamic planning for a flood resilient city.

Land cover change analysis between 1990 and 2021 using Landsat images and object-based classification: A case study in Bodrum peninsula, Aegean Region, Turkey

“…It is seen in the literature that pixelbased classification algorithms are frequently used to determine the current state of land cover (Efe, Soykan, & Cürebal, 2012;İkiel, Ustaoğlu, & Kılıç, 2013;Dutucu & İkiel 2016;Alevkayalı & Tağıl 2018;Sekertekin, Marangoz, & Akcin, 2017;İkiel, Ustaoğlu, & Dutucu, 2019;Mishra, Rai, & Rai, 2020). On the other hand, object based classification, is a classification algorithm that is used extensively after the widespread use of very high resolution satellite images, and its use in medium spatial resolution data such as Landsat has also become widespread (Sertel & Algancı, 2016;Algancı, 2018;Aahlaad, Mozumder, & Tripathi, 2021;Sang, Guo, & Wu, 2021). In recent years, there have been many studies in which pixelbased and object-based classifications have been made comparatively.…”

Assessment and appraisal of local governance on urban flood resilience in Bangkok Metropolitan Region: perspectives of SDGs 11 and 13

“…Vulnerability is usually limited to the susceptibility of structures and infrastructure being damaged (i.e. physical vulnerability) (Thanvisitthpon et al , 2020; Aahlaad et al , 2021).…”

Advancing Data for Street-Level Flood Vulnerability: Evaluation of Variables Extracted from Google Street View in Quito, Ecuador

“…There is a growing body of research acknowledging GSV as a component of flood vulnerability assessment. GSV has been used to collect physical building characteristics that contribute to flood vulnerability based on visual inspection of the imagery [7], [15], [42], [43]. Collection methods range from visual inspection by a user, to deep learning methods trained to pick out specific features in the images.…”

“…Building footprints derived from inspection of satellite imagery and cadastral data have been cross-checked against GSV imagery to fill gaps in or validate data [6]. One study found a high reliability of using GSV to assess physical vulnerability when compared to ground control points from a field survey [42]. Validation of GSV data collected in flood vulnerability research is relatively uncommon.…”