Flood susceptibility mapping by integrating frequency ratio and index of entropy with multilayer perceptron and classification and regression tree

2021 ◽  
Vol 289 ◽  
pp. 112449
Author(s):  
Yi Wang ◽  
Zhice Fang ◽  
Haoyuan Hong ◽  
Romulus Costache ◽  
Xianzhe Tang
Author(s):  
Bahareh Ghasemain ◽  
Dawod Talebpoor Asl ◽  
Binh Thai Pham ◽  
Mohammadtghi Avand ◽  
Huu Duy Nguyen ◽  
...  

Shallow landslides through land degrading not only lead to threat the properly and life of human but they also may produce huge ecosystem damages. The aim of this study was to compare the performance of two decision tree machine learning algorithms including classification and regression tree (CART) and reduced error pruning tree (REPTree) for shallow landslide susceptibility mapping in Bijar, Kurdistan province, Iran. We first used 20 conditioning factors and then they were tested by information gain ratio (IGR) technique to select the most important ones. We then constructed a geodatabase based on the selected factors along with a total of 111 landslide locations with a ratio of 80/20 (for calibration/validation). The performance of the models was checked by the true positive rate (TP Rate), false positive rate (FP Rate), precision, recall, F1-Measure, Kappa, mean absolute error, and area under the receiver operatic curve (AUC). Results of IGR specified that the slope angle and TWI had the most contribution to shallow landslide occurrence in the study area. Moreover, results concluded that although these models had a high goodness-of-fit and prediction accuracy, the CART model (AUC=0.856) outperformed the REPTree model (AUC=0.837). Therefore, the CART model can be used as a promising tool and also as a base classifier to hybrid with optimization algorithms and Meta classifiers for spatial prediction of shallow landslide-prone areas.


2021 ◽  
Vol 9 (1) ◽  
pp. 148
Author(s):  
Hugo Leonardo Oliveira Chaves ◽  
Maria Elisa Leite Costa ◽  
Sérgio Koide ◽  
Tati De Almeida ◽  
Rejane Ennes Cicerelli

<p>O mapeamento de suscetibilidade à inundação é importante para o manejo da dinâmica do uso do solo e, consequentemente, da hidrologia urbana local. O presente estudo produziu o mapa de suscetibilidade à inundação na Bacia do Riacho Fundo, Distrito Federal, utilizando o método estatístico bivariado Razão de Frequência (<em>Frequency Ratio</em>), com 30 pontos de inundação observados em 2018 como pontos de treinamento (71%) e outros 12 pontos de inundação (29%) como pontos de validação para desenvolvimento do modelo. O modelo é composto de 12 fatores de influência: declividade, curvatura, aspecto, hipsometria, distância dos rios, índice de potência de escoamento, índice de transporte de sedimento, índice topográfico de umidade, índice de rugosidade do terreno, índice de escoamento superficial, uso e cobertura do solo e geologia. Todas as variáveis com um tamanho de pixel de 12,5 m x 12,5 m. Os fatores de uso e cobertura do solo e geologia local mostraram-se os mais influentes no modelo. A validação do modelo foi realizada utilizando o método da área sob a curva, com uma acurácia de 85,75%. O estudo mostra que o método pode ser usado para auxiliar no estudo de planos de controle e mitigação de inundação em centros urbanos, como a locação preliminar de bacias de detenção.</p><p><strong>Palavras-chave</strong>: suscetibilidade, inundação, mapeamento, razão de frequência, geoprocessamento.</p><p> </p><p align="center">FLOOD SUSCEPTIBILITY MAPPING USING THE FREQUENCY RATIO METHOD APPLIED TO THE RIACHO FUNDO BASIN - FEDERAL DISTRICT</p><p class="Default"><strong>Abstract</strong><strong></strong></p><p>Flood susceptibility mapping is important to the management of the urban hydrological dynamic and to the studies conducted to prevent the flood-based problems. This study has produced a flood susceptibility map using a bivariate statistical analysis named frequency ratio (FR) model applied in the Riacho Fundo catchment, with 30 flooding locations (71%) for statistical analysis as training dataset and 12 remaining points (29%) were applied to validate the developed model. Twelve conditioning factors were considered in this study: slope, curvature, aspect, elevation, distance to river, stream power index (SPI), sediment transport index (STI), topographic wetness index (TWI), terrain roughness index (TRI), superficial runoff index, land use/land cover (LULC) and geology. All these variables were resampled into 12.5×12.5 m pixel size. The model showed LULC and geology as the most influential factors in flooding. The AUC for success rate was 85.75% with the training points. The study shows the method can be used in studies of plans to mitigate and control flooding in urban centers, as preliminary lease of ponds.</p><p><strong>Keywords</strong>: susceptibility, flooding, mapping, frequency ratio, geoprocessing.</p>


Sign in / Sign up

Export Citation Format

Share Document