Integrated machine learning methods with resampling algorithms for flood susceptibility prediction.

Sci Total Environ

Kalman Kando Faculty of Electrical Engineering, Obuda University, Budapest, Hungary; School of the Built Environment, Oxford Brookes University, Oxford OX30BP, UK.

Published: February 2020

Flood susceptibility projections relying on standalone models, with one-time train-test data splitting for model calibration, yields biased results. This study proposed novel integrative flood susceptibility prediction models based on multi-time resampling approaches, random subsampling (RS) and bootstrapping (BT) algorithms, integrated with machine learning models: generalized additive model (GAM), boosted regression tree (BTR) and multivariate adaptive regression splines (MARS). RS and BT algorithms provided 10 runs of data resampling for learning and validation of the models. Then the mean of 10 runs of predictions is used to produce the flood susceptibility maps (FSM). This methodology was applied to Ardabil Province on coastal margins of the Caspian Sea which faced destructive floods. The area under curve (AUC) of receiver operating characteristic (ROC) and true skill statistic (TSS) and correlation coefficient (COR) were utilized to evaluate the predictive accuracy of the proposed models. Results demonstrated that resampling algorithms improved the performance of Standalone GAM, MARS and BRT models. Results also revealed that Standalone models had better performance with the BT algorithm compared to the RS algorithm. BT-GAM model attained superior performance in terms of statistical measures (AUC = 0.98, TSS = 0.93, COR = 0.91), followed by BT-MARS (AUC = 0.97, TSS = 0.91, COR = 0.91) and BT-BRT model (AUC = 0.95, TSS = 0.79, COR = 0.79). Results demonstrated that the proposed models outperformed the benchmark models such as Standalone GAM, MARS, BRT, multilayer perceptron (MLP) and support vector machine (SVM). Given the admirable performance of the proposed models in a large scale area, the promising results can be expected from these models for other regions.

Download full-text PDF

Source
http://dx.doi.org/10.1016/j.scitotenv.2019.135983DOI Listing

Publication Analysis

Top Keywords

flood susceptibility
16
proposed models
12
models
11
integrated machine
8
machine learning
8
resampling algorithms
8
susceptibility prediction
8
standalone models
8
standalone gam
8
gam mars
8

Similar Publications

An open-source geodatabase and its associate WebGIS platform (CONNECTOSED) were developed to collect and utilize data for the Sediment Flow Connectivity Index (SfCI) for the Apulia region of southern Italy. Maps depicting sediment mobility and connectivity across the hydrographic basins of the Apulia region were generated and stored in the geodatabase. This geodatabase is organized into folders containing data in TIFF, shapefile, Jpeg and Pdf formats, including input variables (digital elevation model, land cover map, rainfall map, and soil units dataset for each hydrographic basin), classification graphs (ranking of variable values), dimensionless index maps (slope, ruggedness, rainfall, land cover, and soil stability) and key products (maps of sediment mobility, SfCI, and applied SfCI).

View Article and Find Full Text PDF

Soil erosion susceptibility maps and raster dataset for the hydrological basins of North Africa.

Sci Data

January 2025

University of Southern California, Viterbi School of Engineering, 3737 Watt Way, Powell Hall of Engineering, Los Angeles, CA, 90089, USA.

Soil erosion in North Africa modulates agricultural and urban developments as well as the impacts of flash floods. Existing investigations and associated datasets are mainly performed in localized urban areas, often representing a limited part of a watershed. The above compromises the implementation of mitigation measures for this vast area under accentuating extremes and continuous hydroclimatic fluctuations.

View Article and Find Full Text PDF

Interpretable flash flood susceptibility mapping in Yarlung Tsangpo River Basin using H2O Auto-ML.

Sci Rep

January 2025

Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research (IGSNRR), Chinese Academy of Sciences (CAS), Beijing, 100101, China.

Flash flood susceptibility mapping is essential for identifying areas prone to flooding events and aiding decision-makers in formulating effective prevention measures. This study aims to evaluate the flash flood susceptibility in the Yarlung Tsangpo River Basin (YTRB) using multiple machine learning (ML) models facilitated by the H2O automated ML platform. The best-performing model was used to generate a flash flood susceptibility map, and its interpretability was analyzed using the Shapley Additive Explanations (SHAP) tree interpretation method.

View Article and Find Full Text PDF

Global climate change has significantly increased the frequency and intensity of extreme precipitation events, thereby heightening flood risks for mountainous settlements. However, due to geographical and socio-economic constraints in these regions, flood-related sample data are generally scarce. This study introduces a Mean Filter (MF) grounded in the point-area duality perspective, combined with a feature selection approach utilizing multi-objective optimization algorithms.

View Article and Find Full Text PDF
Article Synopsis
  • Climate change impacts psychological well-being alongside physical infrastructure, with floods causing significant disruptions to communities and forcing people to relocate.
  • Major disasters, like the 2022 floods in Pakistan, are correlated with increased anxiety among displaced individuals, highlighting the urgent need for understanding this connection.
  • Vulnerable populations, including low-income communities, older adults, and children, face heightened stress during natural disasters, making them more susceptible to the negative effects of climate-related events.
View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!