Exposure Estimation from Multi-Resolution Optical Satellite Imagery for Seismic Risk Assessment

Abstract

Given high urbanization rates and increasing spatio-temporal variability in many present-day cities, exposure information is often out-of-date, highly aggregated or spatially fragmented, increasing the uncertainties associated with seismic risk assessments. This work therefore aims at using space-based technologies to estimate, complement and extend exposure data at multiple scales, over large areas and at a comparatively low cost for the case of the city of Bishkek, Kyrgyzstan. At a neighborhood scale, an analysis of urban structures using medium-resolution optical satellite images is performed. Applying image classification and change-detection analysis to a time-series of Landsat images, the urban environment can be delineated into areas of relatively homogeneous urban structure types, which can provide a first estimate of an exposed building stock (e.g., approximate age of structures, composition and distribution of predominant building types). At a building-by-building scale, a more detailed analysis of the exposed building stock is carried out using a high-resolution Quickbird image. Furthermore, the multi-resolution datasets are combined with census data to disaggregate population statistics. The tools used within this study are being developed on a free-and open-source basis and aim at being transparent, usable and transferable.

A Multiscale Exposure Model for Seismic Risk Assessment in Central Asia

“…In a tier 2 analysis, freeof-cost medium-resolution satellite images are analyzed to outline the extent of built-up areas and to delineate them into areas of relatively homogeneous urban structure at an aggregated neighborhood scale. The tier 2 analysis provides a detailed processing mask for exposure (Wieland et al, 2012a), and the resulting zonation concurs to define the spatial base layer for a stratified sampling to optimize in situ data capturing at the most detailed per-building scale. In a tier 3 analysis, per-building data is acquired and integrated using standard rapid visual screening (RVS; Federal Emergency Management Agency [FEMA] 154, 2002), novel remote rapid visual screening (RRVS) from omnidirectional camera images (Wieland, Pittore, Parolai, Zschau, Moldobekov, et al, 2012) and high-resolution satellite image analysis (Wieland et al, 2012b).…”

“…In successive stages of zonation, the image pixels are clustered and labeled depending on their approximate construction date and predominant building types using change-detection analysis and machine learning assisted image analysis. The zonation steps are explained in detail in Wieland et al (2012a). Complementary information about parameter tuning and training of image analysis algorithms along with a performance evaluation of different machine learning classifiers that were used for the urban pattern recognition is given in Wieland and Pittore (2014).…”

“…Omnidirectional images have been acquired for the cities of Bishkek (Kyrgyzstan; Wieland et al, 2012a), Osh (Kyrgyz-stan), Jalalabad (Kyrgyzstan), Dushanbe (Tajikistan; Pilz et al, 2013), and Khorog (Tajikistan). Field surveys are easy to perform, and operators do not necessarily have to be skilled engineers because the screening of building is carried out at a later stage by analyzing the captured images.…”

“Internet+” approach to mapping exposure and seismic vulnerability of buildings in a context of rapid socioeconomic growth: a case study in Tangshan, China

“…To improve accuracy, a great deal of post-processing work must be performed. For a large number of buildings, applying these automatic or semiautomatic procedures is more difficult, and sometimes, even impossible (Aytekin et al 2009;Durieux et al 2008;Pittore and Wieland 2013;Su et al 2015;Wieland et al 2012). The GE images we used have only red, green, and blue bands, i.e., their spectral information is much poorer than that of commercial ones, thereby aggravating such difficulties.…”

“…There are generally three major types of methods for analyzing seismic exposures, vulnerabilities, and risks: (1) the conventional and mainstream loss ratio curve (LRC) and/ or damaged probability matrix (DPM)-based methods (e.g., ATC 1985;FEMA 1999;Grünthal 1998;Inel et al 2008;Ploeger et al 2010), (2) exposure and vulnerability (E and V) or vulnerability (V) composite index-focused methods (e.g., Carreño et al 2007;Cutter et al 2003;Cutter and Finch 2008;Su et al 2007), and (3) remote sensing (RS)-based E and/or V methods (e.g., Ehrlich et al 2010Geiß et al 2014Geiß et al , 2015Geiß et al , 2016Pittore and Wieland 2013;Taubenböck et al 2009;Wieland et al 2012) (for more, please see Table 1). The conventional LRC and/or DPM-based methods can yield definite and accurate exposure, vulnerability, and risk results.…”