Abstract:
The Ministry of Water Resources of the People's Republic of China proposed to promote the construction of digital twin in the river basin nationwide recently, aiming at improving the ability of flood and drought disaster prevention and decision support. It is the core of intelligent water conservancy system which has the function of 'four pre' (forecast, early warning, simulation preview, prevention plan). Ningbo which locates in the east of China, always suffered a lot every year from both flood and tide which cause by typhoon in the month of July to September. In order to support the construction of digital twin watershed applications, it is necessary to enhance storm surge forecasting model. The improvement of the accuracy of storm surge forecasting, calculation granularity and calculation domain are proposed to adapt to the early warning and forecasting of marine disasters in the basin. Therefore, the new model which based on the ADCIRC mode is extended its border. The coastline and topography are also updated by using the latest collected data, and the key bank sections are encrypted. The computational domain of the model is discretized by unstructured triangular meshes, and the minimum unit length deceased from 200m to 50m. Furthermore, hybrid model for short-term prediction of tide level in estuary based on LSTM and nonstationary harmonic analysis is also adapted to real-time correction of forecasting. And super-computing in cloud is applicated. The result shows that the enhanced storm surge forecasting model could be used in the applications of digital twin watershed. Β© 2023 IEEE.
