Real-time Flood Classification Forecasting Based on k-means++ Clustering and Neural Network

Hu Caihong,Zhang Xueli,Liu Chengshuai,Li Changqing,Jian Shengqi,Wang Jinxing

doi:10.1007/s11269-021-03014-y

Hu Caihong, Zhang Xueli + Show 4 more

Open Access

PDF Available

https://doi.org/10.1007/s11269-021-03014-y

Copy DOI

Export

Save

Cite

Abstract
Highlights/Summary
Full-Text PDF
Similar Papers

Abstract

Listen

Floods are among the most dangerous disasters that affect human beings. Timely and accurate flood forecasting can effectively reduce losses to human life and property and improve the utilization of flood resources. In this study, a real-time flood classification and prediction method (RFC-P) was constructed based on factor analysis, the k-means++ clustering algorithm, SSE, a backpropagation neural network (BPNN) and the M-EIES model. Model parameters of different flood types were obtained to forecast floods. The RFC-P method was applied to the Jingle sub-basin in Shanxi Province. The results showed that the RFC-P method can be used for the real-time classification and prediction of floods. The parameters of the flood classification and prediction model were consistent with the characteristics of the flood events. Compared with the results of unclassified predictions, the Nash coefficient increased by 5%–11.62%, the relative error of the average flood peak was reduced by 6.08%–12.7%, the relative error of the average flood volume was reduced by 5.74%–8.07%, and the time difference of the average peak was reduced by 43%–66% based on the proposed approach. The methodology proposed in this study can be used to identify extreme flood events and provide scientific support for flood classification and prediction, flood control and disaster reduction in river basins, and the efficient utilization of water resources.

Highlights

Floods have always been one of the most dangerous disasters affecting human health
The products of this study are as follows: (1) a real-time flood classification framework based on k-means++, factor analysis, and a BP neural network was established to classify long time series flood events, and hydrological model parameters of corresponding flood categories were obtained through an intelligent optimization algorithm
All the flood events occurred from May to October (Fig. 3a), and 78.41% of the flood events occurred in July and August in summer

Summary

Introduction

Floods have always been one of the most dangerous disasters affecting human health. In addition, with the increase in global temperatures and the intensification of human activities, the frequency of flood disasters is gradually increasing, which poses a considerable threat to human life, property, and ecosystems (Moore et al, 2005). Flood forecasting has always been a challenging topic in hydrology (UNPD, 2005). The hydrological model is a generalization of the law of the water cycle (Gupta et al, 1998), and it cannot be completely consistent with the actual water cycle. There are many factors that affect forecast results (Arkesteijn and Pande, 2013; Gourley and Vieux, 2006; Wagener and Gupta, 2005), among which the model parameters are an important part of the hydrological model of a watershed and have certain physical significance. The rationality of the parameter settings directly affects the final forecast results (Benke et al, 2008). If a set of hydrological forecast model parameters is used to forecast a flood in an entire basin, certain errors will be generated

Methods

Results

Discussion

Conclusion