Evaluation of Solar-Wind Magnetosphere Coupling Functions Using the WINDMI Model

Solar-wind Magnetosphere coupling functions from Newell et. al. 2007 are used as inputs into the WINDMI physics based nonlinear dynamical model of the Earth's nightside magnetosphere. The coupling functions are scaled to the rectified vBs coupling function of Burton et. al. 1975 for this purpose. The outputs of the model are the AL and Dst indices that are compared to the indices measured on the ground. The physical parameters of the model are optimized using a genetic algorithm (GA) computation scheme for each different input using the October 3-7 2000 and April 15-24 2002 Geomagnetic Storm Datasets. The optimization of the model parameters is performed over 3-6 hour intervals to incorporate the changing physical state of the nightside magnetosphere. The predictive capability of each input and the AL and Dst index are compared through using the average relative variance (ARV) and the cross-correlation (COR) measures. The coupling functions are ranked according to performance against the ARV and COR measures. The physical parameters of the WINDMI model obtained from the optimization is also presented. The best performing coupling function will be used as part of a hybrid physics black-box space weather prediction tool in future.