Abstract. Increasing amounts of data, together with more computing power and better machine learning algorithms to analyse the data, are causing changes in almost every aspect of our lives. This trend is expected to continue as more data keep becoming available, computing power keeps improving and machine learning algorithms keep improving as well. Flood risk and impact assessments are also being influenced by this trend, particularly in areas such as the development of mitigation measures, emergency response preparation and flood recovery planning. Machine learning methods have the potential to improve accuracy as well as reduce calculating time and model development cost. It is expected that in the future more applications will become feasible and many process models and traditional observation methods will be replaced by machine learning. Examples of this include the use of machine learning on remote sensing data to estimate exposure and on social media data to improve flood response. Some improvements may require new data collection efforts, such as for the modelling of flood damages or defence failures. In other components, machine learning may not always be suitable or should be applied complementary to process models, for example in hydrodynamic applications. Overall, machine learning is likely to drastically improve future flood risk and impact assessments, but issues such as applicability, bias and ethics must be considered carefully to avoid misuse. This paper presents some of the current developments on the application of machine learning in this field and highlights some key needs and challenges.
The estimation of urban growth in megacities is a critical and intricate task for researchers and decision-makers owing to the complexity of these urban systems. Currently, the majority of megacities are located in Asia which is one of the most disaster-prone regions in the world. The high concentrations of people, infrastructure and assets in megacities create high loss potentials for natural hazards; therefore, the forecasting of exposure metrics such as built-up area is crucial for disaster risk assessment. This study aims to identify and project the dynamics of built-up area at risk using a spatio-temporal approach considering seismic hazard in three Asian megacities, namely Jakarta, Metro Manila and Istanbul. First, Landsat Thematic Mapper images were processed to obtain the built-up areas of 1995 and 2016 for Metro Manila, and of 1995 and 2018 for Jakarta and Istanbul. The SLEUTH urban growth model, a cellular automaton (CA)-based spatial model that simulates urban growth using historical geospatial data, was then employed to predict the urban growth of these megacities by 2030. Finally, seismic hazard maps obtained for 10% and 2% probabilities of exceedance were overlaid with built-up area maps. For a seismic hazard of 10% probability of exceedance in 50 years, the total urban area subjected to Modified Mercalli intensities (MMI) VIII and IX has increased nearly 65% over 35 years in Metro Manila. For Jakarta and Istanbul, the total urban area at the MMI VIII level has increased nearly 79% and 54% over 35 years, respectively. For a seismic hazard of 2% probability of exceedance in 50 years, the total urban area subjected to MMI IX has increased nearly 75%, 65% and 49% over 35 years in Jakarta, Metro Manila and Istanbul, respectively. The results show that urban growth modelling can be utilized to assess the built-up area exposed to high risk as well as to plan urban growth considering natural hazards in megacities.
The Pearl River Delta metropolitan region is one of the most densely urbanized megapolises worldwide with high exposure to weather-related disasters such as storms, storm surges and river floods. Shenzhen megacity has been the fastest growing city in the Pearl River Delta region with a significant increase of resident population from 0.32 million in 1980 to 13.03 million in 2018. Being a flood-prone city, Shenzhen’s rapid urbanization has further exacerbated potential flood losses and forthcoming risk. Thus, evaluating the changes in its exposure from present to future is essential for flood risk assessment, mitigation and management purposes. The main objective of this study is to present a methodology to assess the spatio-temporal dynamics of flood exposure from present to future using high-resolution and open-source data with a particular focus on the built-up area. To achieve this, the SLEUTH model, a cellular automata-based urban growth model, was employed for predicting the built-up area in Shenzhen in 2030. An almost threefold increase was observed in total built-up area from 421 km2 in 1995 to 1166 km2 in 2030, with the 2016 built-up area being 858 km2. Built-up areas, both present (2016) and projected (2030), were then used as the land cover input for flood hazard assessment based on a fuzzy comprehensive evaluation model, which classified the flood hazard into five levels. The analysis indicates that the built-up area subjected to the two highest flood hazard levels will increase by almost 88% (212 km2) from present to future. The approach presented here can be leveraged by policymakers to identify critical areas that should be prioritized for flood mitigation and protection actions to minimize potential losses.
Asia has the fastest growing population and economy, but it is also the most disaster‐prone region in the world. Resilience to disaster impacts from natural hazards will be key to the long‐term sustainability of this rapidly growing region. The first step to building resilience is to identify the key threats that this region faces. We describe these key threats as Black Elephants: a cross between a “black swan” and the proverbial "elephant in the room" — they are extreme events that are known but difficult to address and often ignored. We examine the primary drivers of these looming risks and find that the drivers include underestimated or intensifying hazards, growing exposure, high vulnerability, and unaccounted complexities from multi‐hazard events. In mitigating these key risks, we discuss psychological barriers to action and highlight the importance of information, language, and hope. The known but complex impacts from natural hazards in Asia must be further acknowledged and managed in order to build a more sustainable, resilient future in an increasingly globally connected world.
Abstract. Increasing amounts of data, together with more computing power and better machine learning algorithms to analyse the data are causing changes in almost every aspect of our lives. This trend is expected to continue as more data becomes available, computing power increases and machine learning algorithms improve. Flood risk and impact assessments are also being influenced by this trend, particularly in areas such as the development of mitigation measures, emergency response preparation, and flood recovery planning. Machine learning methods have the potential to improve accuracy as well as reduce calculating time and model development cost. It is expected that in the future more applications become feasible and many process models and traditional observation methods will be replaced by machine learning. Examples of this include the use of machine learning on remote sensing data to estimate exposure or on social media data to improve flood response. Some improvements may require new data collection efforts, such as for the modelling of flood damages or defence failures. In other fields, machine learning may not be suitable or should be applied complementary to process models, for example in hydrodynamic applications. Overall, machine learning is likely to drastically improve future flood risk and impact assessments, but issues such as applicability, bias and ethics must be considered carefully. This paper presents some of the current developments on the application of machine learning for flood risk and impact assessment, and highlights some key needs and challenges.
The inputs of the urban growth model were prepared by me and Ms. Zhu Tinger. • The urban growth modelling and probabilistic seismic hazard analysis were carried out by me. • I integrated the built-up area and seismic hazard maps. I also analyzed the results.
Scrutinizing the evolving exposure and possible consequent forthcoming seismic losses in rapidly urbanizing megacities is essential for decision-makers. Here we present a framework for the evaluation of spatio-temporal seismic loss dynamics where we propose a probabilistic macro-level loss estimation approach that is based on socioeconomic exposure indicators. We follow this framework to model the urban growth, disaggregate population to urban cells, and estimate grid-level wealth in three Asian megacities, namely Jakarta, Metro Manila, and Istanbul. Then, we calculate present and future probabilistic risk metrics based on the combination of evolving exposure, probabilistic seismic hazard analysis and vulnerability curves. The results reveal that our approach can produce present loss estimates that are in the same order of magnitude as the conventional approaches. The predictions suggest that present average annual loss could increase almost twofold in Jakarta and in Metro Manila, and by almost 57% in Istanbul by 2030. Our framework can be used to trigger discussions between scientific community and decision-makers for better long-term risk reduction and risk awareness strategies.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.