结合活动区光球磁场参量和黑子参量的太阳耀斑预报模型

Solar flare forecasting model with active region photospheric magnetic field properties and sunspot factors

尝试将太阳活动区磁场特征物理量与黑子参量结合起来研究太阳耀斑短期预报,探讨综合预报因子在耀斑预报中有效性.太阳黑子参量选取黑子面积、磁分型、Macintosh分型;活动区光球磁场特征物理量选择纵向磁场最大水平梯度、强梯度中性线长度、孤立奇点数目.首先对上述参量的原始数据集通过产率拟合得到规范化的建模数据集,应用多层感知机神经网络方法建立耀斑预报模型.将综合预报模型的预报结果和单独采用黑子数据和磁场参量的两个模型进行了比对,结果显示二者相结合的预报模型的预报准确率有所提高,同时虚报率有所下降.

Solar active region photospheric magnetic physical properties and sunspots factors are connected to research on solar flare short term forecasting. The significance of integrated parameters is discussed. Sunspots parameters are area, magnetic classification, Macintosh classification; and the magnetic physical properties are parameterized with the maximum horizontal gradient, the length of neutral line with the strong gradients, the number of singular points. Using flare productivity fitting function, the initial data set of predictors is normalized to form the modeling data set. Based on this data set, multi-layer perceptron is applied to building flare forecasting model. In experiment, the integrated forecasting model is compared with other two models which take sunspots and magnetic parameters separately. The results indicate that the integrated model has a higher accuracy rate and a lower false alarm rate.