A GIS-Based Artificial Neural Network Model for Flood Susceptibility Assessment

Khoirunisa, Nanda; Ku, Cheng-Yu; Liu, Chih-Yu

doi:10.3390/ijerph18031072

Cited by 66 publications

(36 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The deep learning model then computes, for each point in the area, a score from 0 (non-flooded) to 1 (flooded). These scores are finally divided into several classes, generally using the natural (Jenks) breaks method (e.g., Fang et al, 2020a;Wang et al, 2020;Khoirunisa et al, 2021), to obtain a susceptibility map. An exception is given by Jahangir et al…”

Section: Deep Learning For Flood Susceptibilitymentioning

confidence: 99%

“…Most papers use MLP and CNN. Models based on MLPs consider single points or pixels as inputs (Tien et al, 2020;Ahmadlou et al, 2021;Khoirunisa et al, 2021), while CNNs consider the whole raster files (Zhao et al, 2020b;Khosravi et al, 2020;Wang et al, 2020). Since MLPs lack inductive bias they provide less coherent results, meaning that the variation among neighboring cells can be high.…”

Section: Architecturesmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning Methods for Flood Mapping: A Review of Existing Applications and Future Research Directions

Bentivoglio

Isufi

Jonkman

et al. 2021

Preprint

View full text Add to dashboard Cite

Abstract. Deep Learning techniques have been increasingly used in flood risk management to overcome the limitations of accurate, yet slow, numerical models, and to improve the results of traditional methods for flood mapping. In this paper, we review 45 recent publications to outline the state-of-the-art of the field, identify knowledge gaps, and propose future research directions. The review focuses on the type of deep learning models used for various flood mapping applications, the flood types considered, the spatial scale of the studied events, and the data used for model development. The results show that models based on convolutional layers are usually more accurate as they leverage inductive biases to better process the spatial characteristics of the flooding events. Traditional models based on fully-connected layers, instead, provide accurate results when coupled with other statistical models. Deep learning models showed increased accuracy when compared to traditional approaches and increased speed when compared to numerical methods. While there exist several applications in flood susceptibility, inundation, and hazard mapping, more work is needed to understand how deep learning can assist real-time flood warning during an emergency, and how it can be employed to estimate flood risk. A major challenge lies in developing deep learning models that can generalize to unseen case studies and sites. Furthermore, all reviewed models and their outputs, are deterministic, with limited considerations for uncertainties in outcomes and probabilistic predictions. The authors argue that these identified gaps can be addressed by exploiting recent fundamental advancements in deep learning or by taking inspiration from developments in other applied areas. Models based on graph neural networks and neural operators can work with arbitrarily structured data and thus should be capable of generalizing across different case studies and could account for complex interactions with the natural and built environment. Neural operators can also speed up numerical models while preserving the underlying physical equations and could thus be used for reliable real-time warning. Similarly, probabilistic models can be built by resorting to Deep Gaussian Processes.

show abstract

Section: Deep Learning For Flood Susceptibilitymentioning

confidence: 99%

Section: Architecturesmentioning

confidence: 99%

Deep Learning Methods for Flood Mapping: A Review of Existing Applications and Future Research Directions

Bentivoglio

Isufi

Jonkman

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Most papers use MLP and CNN. Models based on MLPs consider single points or pixels as inputs (Tien et al, 2020;Ahmadlou et al, 2021;Khoirunisa et al, 2021), while CNNs consider the whole raster files (Zhao et al, 2020b;Khosravi et al, 2020;Wang et al, 2020b). Since MLPs lack inductive bias they provide less coherent results, meaning that the variation among neighboring cells can be high.…”

Section: Architecturesmentioning

confidence: 99%

Deep Learning Methods for Flood Mapping: A Review of Existing Applications and Future Research Directions

Bentivoglio

Isufi

Jonkman

et al. 2022

Preprint

View full text Add to dashboard Cite

Abstract. Deep Learning techniques have been increasingly used in flood management to overcome the limitations of accurate, yet slow, numerical models, and to improve the results of traditional methods for flood mapping. In this paper, we review 58 recent publications to outline the state-of-the-art of the field, identify knowledge gaps, and propose future research directions. The review focuses on the type of deep learning models used for various flood mapping applications, the flood types considered, the spatial scale of the studied events, and the data used for model development. The results show that models based on convolutional layers are usually more accurate as they leverage inductive biases to better process the spatial characteristics of the flooding events. Models based on fully-connected layers, instead, provide accurate results when coupled with other statistical models. Deep learning models showed increased accuracy when compared to traditional approaches and increased speed when compared to numerical methods. While there exist several applications in flood susceptibility, inundation, and hazard mapping, more work is needed to understand how deep learning can assist real-time flood warning during an emergency, and how it can be employed to estimate flood risk. A major challenge lies in developing deep learning models that can generalize to unseen case studies. Furthermore, all reviewed models and their outputs, are deterministic, with limited considerations for uncertainties in outcomes and probabilistic predictions. The authors argue that these identified gaps can be addressed by exploiting recent fundamental advancements in deep learning or by taking inspiration from developments in other applied areas. Models based on graph neural networks and neural operators can work with arbitrarily structured data and thus should be capable of generalizing across different case studies and could account for complex interactions with the natural and built environment. Physics-based deep learning can be used to preserve the underlying physical equations resulting in more reliable speed-up alternatives for numerical models. Similarly, probabilistic models can be built by resorting to Deep Gaussian Processes or Bayesian neural networks.

show abstract

“…(2021) investigated 84 ood hazard spots in Guang Nam province, Vietnam. Additionally, Khoirunisa et al (2021) examined data from 307 oods that occurred in Keelung, Taiwan, between 2015 and 2019.…”

Section: Introductionmentioning

confidence: 99%

Applications of Spatial Statistics and Spatial Analysis in Flooding Assessment Using GIS: A Case Study in Darab Watershed, Iran

Shariati

Kazemi

Samani

et al. 2022

Preprint

View full text Add to dashboard Cite

Floods are amongst the most catastrophic natural disasters, having enormous impacts on both infrastructure and humans. Between 1998 and 2017, this detrimental phenomenon afflicted more than two billion individuals globally. The present investigation utilized two phases of spatial analysis and spatial statistics from ArcGIS in order to precisely assess the hazard of flooding in the Darab watershed, Iran. Following a survey of literature and case studies, seven criteria for flood risk were recognized to be effective: dem, slope, drainage density, NDVI, land cover, geology, and rainfall. The initial phase was operated using fuzzy logic and AHP in order to overlap layers. In order to prepare of flood hazard mapping, kernel density and zonal statistics were used to compute flood hazard density and flood hazard zoning. In flood hazard zones, spatial statistics in ArcGIS Pro were applied to detect high-and low-risk clusters. An innovation in this study is the use of local data to detect high-risk and low-risk flood clusters. According to the findings, the Gamma operator (0.9) was recognized as the optimal operator for flooding zoning, and only 7344/63432 hectares of the study area had a high risk of flooding, with only two high-risk flooding clusters in the Darab watershed. The boundary was near the heights in the north and northeast of the study, which corresponded to townships with a population of less than 5,000 people. Overall, the findings of this study demonstrated that clusters and outliers' analysis, as well as hotspot analysis, are effective complementary techniques to recognize high- and low-risk flood clusters. Combining spatial analysis with spatial statistics may be a reliable and efficient approach for natural scientific investigations.

show abstract

A GIS-Based Artificial Neural Network Model for Flood Susceptibility Assessment

Cited by 66 publications

References 45 publications

Deep Learning Methods for Flood Mapping: A Review of Existing Applications and Future Research Directions

Deep Learning Methods for Flood Mapping: A Review of Existing Applications and Future Research Directions

Deep Learning Methods for Flood Mapping: A Review of Existing Applications and Future Research Directions

Applications of Spatial Statistics and Spatial Analysis in Flooding Assessment Using GIS: A Case Study in Darab Watershed, Iran

Contact Info

Product

Resources

About