Swarm‐based optimizer for convolutional neural network: An application for flood susceptibility mapping

This paper proposes a novel hybrid method for flood susceptibility mapping using a geographic information system (ArcGIS) and satellite images based on the analytical hierarchy process (AHP). Here, the following nine multisource environmental controlling factors influencing flood susceptibility were considered for relative weight estimation in AHP: elevation, land use, slope, topographic wetness index, curvature, river distance, flow accumulation, drainage density, and rainfall. The weight for each factor was determined from AHP and analyzed to investigate critical regions that are more vulnerable to floods using the overlay weighted sum technique to integrate the nine layers. As a case study, the ArcGIS-based framework was applied in Seoul to obtain a flood susceptibility map, which was categorized into six regions (very high risk, high risk, medium risk, low risk, very low risk, and out of risk). Finally, the flood map was verified using real flood maps from the previous five years to test the model’s effectiveness. The flood map indicated that 40% of the area shows high flood risk and thus requires urgent attention, which was confirmed by the validation results. Planners and regulatory bodies can use flood maps to control and mitigate flood incidents along rivers. Even though the methodology used in this study is simple, it has a high level of accuracy and can be applied for flood mapping in most regions where the required datasets are available. This is the first study to apply high-resolution basic maps (12.5 m) to extract the nine controlling factors using only satellite images and ArcGIS to produce a suitable flood map in Seoul for better management in the near future.

Download Full-text

Landslide susceptibility mapping based on convolutional neural network and conventional machine learning methods

10.21203/rs.3.rs-190195/v1 ◽

2021 ◽

Author(s):

Rui Liu ◽

Xin Yang ◽

Chong Xu ◽

Luyao Li ◽

Xiangqiang Zeng

Keyword(s):

Neural Network ◽

Machine Learning ◽

Convolutional Neural Network ◽

Landslide Susceptibility ◽

Susceptibility Mapping ◽

Landslide Susceptibility Mapping ◽

Support Vector ◽

Learning Methods ◽

Machine Learning Methods ◽

Conventional Machine

Abstract Landslide susceptibility mapping (LSM) is a useful tool to estimate the probability of landslide occurrence, providing a scientific basis for natural hazards prevention, land use planning, and economic development in landslide-prone areas. To date, a large number of machine learning methods have been applied to LSM, and recently the advanced Convolutional Neural Network (CNN) has been gradually adopted to enhance the prediction accuracy of LSM. The objective of this study is to introduce a CNN based model in LSM and systematically compare its overall performance with the conventional machine learning models of random forest, logistic regression, and support vector machine. Herein, we selected the Jiuzhaigou region in Sichuan Province, China as the study area. A total number of 710 landslides and 12 predisposing factors were stacked to form spatial datasets for LSM. The ROC analysis and several statistical metrics, such as accuracy, root mean square error (RMSE), Kappa coefficient, sensitivity, and specificity were used to evaluate the performance of the models in the training and validation datasets. Finally, the trained models were calculated and the landslide susceptibility zones were mapped. Results suggest that both CNN and conventional machine-learning based models have a satisfactory performance (AUC: 85.72% − 90.17%). The CNN based model exhibits excellent good-of-fit and prediction capability, and achieves the highest performance (AUC: 90.17%) but also significantly reduces the salt-of-pepper effect, which indicates its great potential of application to LSM.

Download Full-text