Abnormal relative tendency generation method and system for climate prediction

Abstract

The invention discloses an abnormal relative tendency generation method and system for climate prediction. Performing time smoothing on the original data of the climate variables to remove the high-frequency variation rate; subtracting the climate state of the smooth data to obtain a distance plane; on the basis, according to the target frequency band, selecting the length of the early-stage average time period of each moment, and defining a corresponding recent abnormal background; the recent abnormal background is subtracted, the low frequency variation rate is removed, and the abnormal relative tendency is obtained; and modeling and predicting the abnormal relative tendency, and substituting and adding the low-frequency variation rate as a known recent abnormal background to realize prediction of a distance and an original field. According to the method, the problems of time boundary and multi-scale of climate prediction are solved, target frequency band information is extracted at the tail end of a time sequence only by using historical and current data, the relative tendency of high-frequency anomaly on a known recent anomaly background can be highlighted, only the relative tendency of anomaly needs to be predicted, errors caused by introduction of prediction low-frequency variation rate are avoided, and the prediction accuracy is improved. And the accuracy and the stability of climate prediction are effectively improved.

Description

Technical field [0001] The present invention belongs to the field of digital signal processing and climate prediction, and more particularly to the target band information extraction method of history and current data and the climate prediction method based on the method, can be used for a variety of time scale of the season, the secondary season. predict. Background technique [0002] Existing common filtering methods for climate research. Climate research often pays attention to changes in the climate system in certain specific frequency bands, to extract information on the frequency band, weaken other frequency bands to the study, need to apply filters for raw data applications. Existing Filters commonly used in climate research mainly include FFT filters (Steinacker, 2021), LANCZOS filter (LANCZOS, 1956; Duchon, 1979) and Butterworth, 1930; Murakami, 1979). The FFT filter belongs to the frequency domain filter, and the discrete Fourier transform is the theoretical basis; the ...

AI Technical Summary

Problems solved by technology

[0007] Purpose of the invention: Aiming at the problem that existing commonly used filtering methods in climate research cannot be applied at the end of the time series, and the existing solutions to the end problem may introduce initial value errors into climate predictions, may cause inconsistencies in magnitude and phase, lack of universal properties and general To solve the problem of adaptability, the present invention provides a method and system for generating relative tendency of anomalies for climate prediction. As a causal system, it only needs to use historical and current data for calculation, and does not involve future data extension. It can include All the data including the end points are applied consistently, and have a universal nature completely determined by their calculation coefficients, which can achieve the purpose of highlighting the information of the target frequency band, achieve non-traditional filtering effects, and be applied to climate prediction on multiple time scales

Images