Correlation evaluation method for forecasting factor and solar flare

Abstract

The invention relates to the research technology of solar activities, in particular to a correlation evaluation method for forecasting factors and solar flare, which comprises the following specific steps: step 1, extracting all the forecasting factors and data corresponding to whether a flare event occurs or not, and defining a correlation index of each forecasting factor; step 2, drawing a correlation graph of any one forecast factor and data whether flare occurs or not; step 3, performing statistics on the flare occurrence probability of the forecasting factor in each data segment to obtain a correlation index of the forecasting factor; step 4, giving a correlation conclusion between the forecasting factor and whether a flare event occurs or not according to the numerical value of thecorrelation index; and step 5, executing the step 2 to the step 4 on other forecasting factors to obtain a correlation conclusion of each forecasting factor. According to the method, a multi-data-segment evaluation method is adopted, and a correlation conclusion can be obtained completely and accurately.

Description

technical field [0001] The invention relates to a research technology of solar activity, in particular to a method for evaluating the correlation between a predictor based on a density statistics method and the occurrence of solar flares. [0002] Background of the invention [0003] A solar flare is a sudden and violent energy release process, and when a flare occurs, it will also endanger human survival. Accurately forecasting the situation of solar eruption activities in the future can prevent and deal with disasters in a timely manner. [0004] Prediction of solar flares requires accurate predictors, and there are two main methods for extracting predictors. One is to observe sunspots and then perform sunspot group classification: the most commonly used is the McIntosh classification method, which reduces the 9 types in the traditional Zurich classification to 7 types, and uses them as the characteristic parameters of the solar flare prediction process The other is to di...

AI Technical Summary

Problems solved by technology

[0005] At present, under the condition of obtaining a large amount of open source data, a set of linear or nonlinear mappings can be obtained by reducing the dimensionality of the predictors; however, although this method can eliminate redundant information in the predictors, it cannot directly Determining whether the predictors used are complete

As for how to build a relationship model between predictors and solar flares, and effectively evaluate the nonlinear relationship between predictors and flares, there is no authoritative public literature that elaborates on it

Images