A small watershed mountain flood forecasting method

Abstract

The invention relates to a small watershed mountain flood forecasting method, and provides a technical scheme for solving the problem of low flood forecasting precision of a small watershed in the prior art, which comprises the following steps of: setting model parameters; inputting real-time data; carrying out runoff production calculation, water source division calculation and confluence calculation; Outputting the flow data of the on-site runoff and the total flow data at the outlet of the whole basin; judging whether the output total flow at the outlet of the whole basin is greater than aflood early warning threshold or not; if the flood early warning threshold value is not greater than the flood early warning threshold value, not operating; And starting a flood early warning scheme if the flood early warning threshold value is greater than the flood early warning threshold value. Source of model parameters in the method is 5-. In 10 years, the rainfall data and the real-time dataof the basin are from the monitoring data of the telemetry system, and in addition, before the real-time data are input, the detection data of the telemetry system are processed by using a robust algorithm. The method is suitable for simulated runoff calculation of a small watershed, is suitable for short-term runoff forecasting, and is high in forecasting precision.

Description

technical field [0001] The invention relates to a mountain torrent forecasting method, in particular to a small watershed torrent forecasting method. Background technique [0002] A small watershed usually refers to a catchment area of ​​100 km below the secondary and tertiary tributaries bounded by the watershed and the outlet section of the downstream channel. 2 In the following relatively independent and closed natural catchment areas, because there are few or no monitoring stations in mountainous areas in small watersheds, hydrological data are scarce, and artificial forecasting is mostly used. At present, the artificial forecast of small watershed floods is mainly calculated by the indirect method. This method first assumes that the rain and flood are of the same frequency, and the design flood is calculated from the design storm. It is also relatively low and cannot meet the requirements for flood forecasting. [0003] In addition, in the process of real-time flood f...

AI Technical Summary

Problems solved by technology

At present, the artificial forecast of small watershed floods is mainly calculated by the indirect method. This method first assumes that the rain and flood are of the same frequency, and the design flood is calculated from the design storm. It is also relatively low and cannot meet the requirements for flood forecasting

[0003] In addition, in the process of real-time flood forecasting, the telemetry system is generally used to monitor the water regime data in real time, but due to the reasons of the telemetry system itself and the specific requirements of the measurement of hydrological elements, the data often carry abnormal errors

In order to ensure the normal operation of the forecasting system and the accuracy of flood forecasting in case of abnormal factors, the traditional processing method uses the difference between the measured flow rate and the modeled flow rate as the information basis for real-time correction to correct the estimated results, model parameters or models in real time. Input, due to the existence of data errors, the actual residual error often cannot meet the above assumptions, which pollutes the correction method and seriously affects the effect of real-time correction

For example, the publication date is July 6, 2016, and the Chinese patent with the publication number CN105740969A discloses a data-driven real-time flood forecasting method for small watersheds. This invention does not perform robustness analysis on real-time data, and the forecasting accuracy is low.

Images