Monitoring of Inland Excess Water Inundations Using Machine Learning Algorithms

Nowadays, climate change not only leads to riverine floods and flash but also inland excess water (IEW) inundations drought due extreme hydrological processes. The Carpathian Basin is extremely affected by fast-changing weather conditions during the year. IEW (sometimes referred as logging) formed when, limited runoff, infiltration, evaporation, surplus remains on surface or in places where groundwater flowing lower areas appears leaking through porous soil. In this study, eight different machine learning approaches were applied derive three dates 2021 (23 February, 7 March, 20 March). Index-based are simple provide relatively good results, they need be adapted specific circumstances for each area date. With an overall accuracy of 0.98, a Kappa 0.65, QADI score 0.020, deep method Convolutional Neural Network (CNN) gave best compared more traditional Maximum Likelihood (ML), Random Forest (RF), Support Vector Machine (SVM) artificial neural network (ANN) that evaluated. CNN-based maps can used operational control management authorities.