Balch C C
National Oceanic and Atmospheric Administration, Space Environmental Center, Boulder, CO 80303-3328, USA.
Radiat Meas. 1999 Jun;30(3):231-50. doi: 10.1016/s1350-4487(99)00052-9.
This paper describes a model that has been used at the NOAA Space Environment Center since the early 1970s as a guide for the prediction of solar energetic particle events. The algorithms for proton event probability, peak flux, and rise time are described. The predictions are compared with observations. The current model shows some ability to distinguish between proton event associated flares and flares that are not associated with proton events. The comparisons of predicted and observed peak flux show considerable scatter, with an rms error of almost an order of magnitude. Rise time comparisons also show scatter, with an rms error of approximately 28 h. The model algorithms are analyzed using historical data and improvements are suggested. Implementation of the algorithm modifications reduces the rms error in the log10 of the flux prediction by 21%, and the rise time rms error by 31%. Improvements are also realized in the probability prediction by deriving the conditional climatology for proton event occurrence given flare characteristics.
本文描述了一个自20世纪70年代初以来一直在美国国家海洋和大气管理局空间环境中心使用的模型，作为预测太阳高能粒子事件的指南。文中描述了质子事件概率、峰值通量和上升时间的算法。将预测结果与观测结果进行了比较。当前模型显示出一定能力，可区分与质子事件相关的耀斑和与质子事件无关的耀斑。预测峰值通量与观测峰值通量的比较显示出相当大的离散度，均方根误差几乎达到一个数量级。上升时间的比较也显示出离散度，均方根误差约为28小时。利用历史数据对模型算法进行了分析，并提出了改进建议。算法修改的实施使通量预测对数10中的均方根误差降低了21%，上升时间均方根误差降低了31%。通过推导给定耀斑特征下质子事件发生的条件气候学，概率预测也得到了改进。
