A hybrid novel framework for flood disaster risk control in developing countries based on smart prediction systems and prioritized scenarios
Hadi Akbarian,Mohammad Gheibi,Mostafa Hajiaghaei-Keshteli,Mojtaba Rahmani
DOI: https://doi.org/10.1016/j.jenvman.2022.114939
2022-06-15
Abstract:A Decision Support System (DSS) is a highly efficient concept for managing complex objects in nature or human-made phenomena. The main purpose of the present study is related to designing and implementation of real-time monitoring, prediction, and control system for flood disaster management as a DSS. Likewise, the problem of statement in the research is correlated to implementation of a system for different climates of Iran as a unique flood control system. For the first time, this study coupled hydrological data mining, Machine Learning (ML), and Multi-Criteria Decision Making (MCDM) as smart alarm and prevention systems. Likewise, it created the platform for conditional management of floods in Iran's different clusters of climates. According to the KMeans clustering system, which determines homogeneity of the hydrology of a specific region, Iran's rainfall is heterogeneous with 0.61 score, which is approved high efficiency of clustering in a vast country such as Iran with four seasons and different climates. In contrast, the relation of rainfall and flood disaster is evaluated by Nearest Neighbors Classification (NNC), Stochastic Gradient Descent (SGD), Gaussian Process Classifier (GPC), and Neural Network (NN) algorithms which have an acceptable correlation coefficient with a mean of 0.7. The machine learning outputs demonstrated that based on valid data existence problems in developing countries, just with verified precipitation records, the flood disaster can be estimated with high efficiency. In the following, Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) method as a Game Theory (GT) technique ranked the preventive flood damages strategies through three social (Se 1), environmental (Se 2), and economic (Se 3) crises scenarios. The solutions of flood disaster management are collected from literature review, and the opinion approves them of 9 senior experts who are retired from a high level of water resource management positions of Iran. The outcomes of the TOPSIS method proved that National announcement for public-institutional participation for rapid response and funding (G1-2), Establishment of delay structures to increase flood focus time to give the animals in the ecosystem the opportunity to escape to the upstream points and to preserve the habitat (G 2-8), and Granting free national financial resources by government agencies in order to rebuild sensitive infrastructure such as railways, hospitals, schools, etc. to the provincial treasury (G3-10) are selected as the best solution of flood management in Social, Environmental, and Economic crises, respectively. Finally, the collected data are categorized in Social, Environmental, and Economic aspects as three dimensions of Sustainable Development Goals (SDGs) and ranked based on the opinion of 32 experts in the five provinces of present case studies.