Abstract:In recent decades, the frequency of torrential rain coupled with various complex flood patterns increased. Since floods are changeable and unpredictable, deeper understanding of flood incidents is necessary for better watershed management. The primary objective of this research is to investigate and characterize annual flood patterns embedded in a river catchment located in west of Japan. To fulfill this aim, we proposed a method based on information and complexity principles to examine the annual variation of flood patterns embedded in a riverine system. The key strength of the proposed approach is being established on two fundamental pillars: (1) word pattern; gives an image about the detected flood patterns, i.e., simple flood occurs within one day or severe flood persists for two days or more, and (2) information-complexity indices that report the frequency and randomness of the detected patterns. Information content was quantified using Mean Information Gain (MIG), whereas Effective Measure Complexity (EMC) and Fluctuation Complexity (FC) were indices used to define the complexity in the studied records. The results show that the proposed method is very powerful in detecting hidden patterns. Furthermore, we succeed in capturing the stations that exhibited the same flood patterns. The main finding of the first pillar of this approach revealed that flood events were fundamentally triggered by precipitation occurred during East Asian monsoon and tropical cyclones. Alternatively, the information-complexity pillar was a powerful tool in capturing different internal structures of flood patterns. Hence, higher MIG values indicated higher degree of randomness. On the other hand, higher EMC values reflected the length of flood events, while higher FC values showed higher number of separated flood events. Overall, this study discusses the competence of a new approach capable to capture hidden patterns in dataset and can be extended to numerous applications that investigate system behavior.

Flood Clustering Analysis Based on Feature Selection

Data Fusion Using Feature Selection Based Causal Network Algorithm

Study on applying hierarchical clustering method to flood forecasting

A novel framework for feature simplification and selection in flood susceptibility assessment based on machine learning

Modeling flood susceptibility using data-driven approaches of naive Bayes tree, alternating decision tree, and random forest methods

Preliminary Discussion on Flood Forecast Model Based on Support Vector Machine

Is Clustering Time-Series Water Depth Useful? An Exploratory Study for Flooding Detection in Urban Drainage Systems

Flooded Area Detection Method Based on Fusion of Optical and SAR Remote Sensing Images

Risk Assessment of Coal Mine Flood Disasters Based on Projection Pursuit Clustering Model

Identification of the Key Driving Factors of Flash Flood Based on Different Feature Selection Techniques Coupled with Random Forest Method

Mid and Long Term Hydrology Forecast Model and Its Application

A Feature Selection Method Based on Feature Grouping and Genetic Algorithm

A comparative analysis of feature selection models for spatial analysis of floods using hybrid metaheuristic and machine learning models

A novel integrated Urban flood risk assessment approach coupling GeoDetector-Dematel and clustering method

Characterizing annual flood patterns variation using information and complexity indices

Regional Flood Risk Assessment Via Coupled Fuzzy C -Means Clustering Methods: an Empirical Analysis from China’s Huaihe River Basin

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

Feature Selection Based on Data Clustering

Integrated Catchment Classification Across China Based on Hydroclimatological and Geomorphological Similarities Using Self-Organizing Maps and Fuzzy C-Means Clustering for Hydrological Modeling

Regional flood risk assessment based on fuzzy cluster factor analysis

Flood prediction based on feature selection and a hybrid deep learning network