Precipitation intensity intelligent correction method based on mode set information
By acquiring ensemble forecast information from the current and historical periods, and using a precipitation intensity correction model trained by machine learning, the problem of inaccurate intensity in precipitation ensemble forecasts has been solved, achieving higher precision precipitation forecasts.
CN121634347APending Publication Date: 2026-03-10EARTH SYST NUMERICAL PREDICTION CENT OF CHINA METEOROLOGICAL ADMINISTRATION
View PDF 0 Cites 0 Cited by
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-19
- Publication Date
- 2026-03-10
AI Technical Summary
Technical Problem
Existing precipitation ensemble forecasting systems suffer from false alarms and missed predictions in heavy precipitation forecasts, resulting in insufficient accuracy and failing to meet application requirements.
Method used
By acquiring ensemble forecast information from the current model and historical data for the same period, the precipitation percentile sequence is determined, and a precipitation intensity correction model trained by machine learning is used to correct the forecast precipitation intensity.
Benefits of technology
It improves the accuracy of precipitation forecast results and effectively enhances the post-processing correction accuracy of precipitation ensemble numerical forecasts.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN121634347A_ABST
Abstract
The invention relates to the technical field of meteorological prediction, and provides a rainfall intensity intelligent correction method based on mode set information. The method comprises the following steps: determining a rainfall percentile sequence corresponding to each forecast time efficiency according to ensemble forecast information of each forecast time efficiency of a selected area at a current mode report starting time and ensemble forecast information of each forecast time efficiency in a historical same time period; taking the rainfall percentile sequence as a rainfall forecast sequence feature, inputting the rainfall percentile sequence feature into a rainfall intensity correction model, and outputting corrected rainfall intensity by the rainfall intensity correction model; wherein the rainfall intensity correction model is a model which is obtained by taking the rainfall forecast sequence characteristics of the sample data and the actual rainfall sequence characteristics as input and through machine learning and training and is used for correcting and forecasting the rainfall intensity. According to the method, the post-processing correction precision of rainfall set numerical forecasting can be effectively improved.
Need to check novelty before this filing date? Find Prior Art