Medium and long term runoff prediction method based on numerical rainfall forecast

Abstract

The invention provides a medium and long term runoff prediction method based on numerical rainfall forecast, which is a product based on numerical rainfall forecast of the European mid-term weather forecast center in the future six months, is fused with a short term runoff forecast model, and greatly increases the forecast period of runoff forecast and the forecast precision of runoff forecast.

Description

technical field [0001] The invention relates to a medium and long-term runoff prediction method based on numerical rainfall forecasting, and belongs to the technical field of hydrology and water resources technology. Background technique [0002] Runoff refers to rainfall, ice and snow melt water, or water flowing along the surface or underground under the action of gravity when watering the ground. Runoff is the basic link of the water cycle and the basic element of water balance. Runoff forecasting is a scientific information work based on mastering objective hydrological laws and predicting future changes in water regimes. It is a prerequisite for realizing the scientific mobilization of water resources. [0003] Based on the basic principles of the hydrological cycle, the basic concepts and classifications of mid- and long-term runoff forecasting are clarified from the perspective of factors affecting the source of water vapor in the basin runoff. Predict the ten-day, m...

AI Technical Summary

Problems solved by technology

[0004] In the existing mid- and long-term runoff forecasts, there are mainly two difficulties. On the one hand, due to the lack of reliable meteorological forecast results and the time scale of mid- and long-term runoff forecasts is mostly on the monthly scale, it is difficult to establish a physical model based on the relationship between rainfall and runoff. On the other hand, affected by many factors such as climate, vegetation, topography and human activities, the river runoff sequence has obvious random and nonlinear characteristics

Aiming at these two problems, current research often uses data-driven models to establish statistical relationships between autocorrelation factors such as runoff and anterior runoff, and teleconnection factors such as climate factors. Key information, which in turn affects the accuracy of the forecast results, and as the forecast period becomes longer, the correlation between the previous runoff and the current runoff rapidly weakens, and the reliability of the forecast decreases rapidly

The patent with the publication number CN113592143A "Medium and long-term runoff ensemble forecast method and system based on rainfall forecast" specifically discloses the construction of runoff ensemble forecast data for the target area based on the forecast period, historical runoff data and the predictors corresponding to the historical runoff data , input the runoff ensemble forecast data into the trained runoff ensemble forecast model, and obtain the medium and long-term runoff ensemble forecast results of the target area. The method of this patent solves the problem that the traditional medium and long-term runoff forecast method can only produce deterministic runoff forecast The problem of the result, but in this method, the coupling model is used for prediction, and most of the methods used are mathematical statistics, without considering the relevant physical meaning, which will affect the prediction accuracy

Images