1. Hydrological stochastic simulation is a technology that uses the method of hydrological time series analysis to establish a model for a given hydrological time series, and then generates an artificial series by Monte Carlo method according to the selected model. The generated artificial sequence needs to be statistically tested, and the model and generated sequence should be re-established if it does not meet the requirements.
2. Water conservancy and hydropower, flood control and water resources dispatching play a fundamental guiding role in hydrological forecasting. At present, there are two commonly used hydrological forecasting methods in China: data-driven model and process-driven model. The data-driven model is not based on hydrological process, but on the mathematical relationship between data to establish an effective hydrological model. Based on the concept of hydrology, the process-driven model completes the evolution model of rivers and basins. In the medium and long-term hydrological forecast, the process-driven model is more suitable, mainly to improve the structure of the rainfall-runoff model in the basin, and is suitable for hydrological forecast with relatively long time span.