摘要
重现性地磁扰动作为地磁活动的重要部分,其特性在磁情预报中起着不可缺少的重要作用.本文利用地磁扰动在一定条件下具有27天重现的特性和现代时序分析法中的自适应滤波方法,对北京地磁台K指数的预报方法进行了研究,并分别在太阳黑子活动低年和高年的两个时间段,进行了预报实验:低年时段预报结果精度为0.66;在高年时段,预报精度为0.79.进一步分析还表明,理论计算方法应当与多方面的物理分析相结合,这样才能为磁扰的定量分析和预报提供更充分与合理的计算条件和物理根据.
The recurrence of geomagnetic activities has provided the. possibility for the prediction. A quantitative, computable. method--self--adaptive filtering, is introducedto forecast K index; its objective is, going with the filtering, to estimate and rectifythe model parameters (on--line) continuously by means of the new observations. Twoperiods July 14, 1984-- Jan. 18, 1985 and Sept. 15, 1991--Mar.21, 1992 are selected for testing the method. The prediction precision is estimated by covariance(Cog) and standard variance (St). For the first period, CoP ~ 0.66, St = 4.74; thesecond is Coy = 0.79, St = 6.43. The results also indicate the significance of combining the various observations, forecaster's experience with the theoretical computations.
出处
《空间科学学报》
CAS
CSCD
北大核心
1994年第3期237-243,共7页
Chinese Journal of Space Science
关键词
地磁活动
数值预报
自适应滤波
Recurrency, Geomagnetic disturbances, Numericl prediction, Selfadaptive filtering