Prior investigation for flash floods and hurricanes, concise capsulization of hydrological technologies and instrumentation: A survey

Intense and severe torrents, tornadoes and thunderstorm causes numerous casualties in fraction of second and extreme devastation of infrastructure in many countries. Flash floods are considered one of the most hazardous natural disasters. Several approaches have been made for an authentic and robust early warning system to forecast the flash floods vigorously. An intelligent and vibrant model for the recognition of floods includes the estimation of water level, Global Positioning System-Precipitable Water Vapor (GPS PWV), precipitation velocity, wind speed, direction, upstream levels of river, soil moisture, oceanic bottom pressure and color of the water with accurate and reliable cognition algorithms. UGS (unattended ground sensors) and langrangian micro transducers are deployed on the ground and spread on the sea surface respectively to investigate the hydrological and meteorological differences on real time basis. By the utilization of fuzzy logic, Kalman filtering, Adaptive neuro fuzzy interference system (ANFIS), Particle Swarm Optimization (PSO) and Neural network autoregressive model with exogenous input (NNARX) based structure. Reduction of complexities in data collection, high false alarm rates, communication issues, low WSN battery backup and all related hindrances have been the focal point of this research paper.