The Coronal Hole Influence Parameter (CHIP) as an Important Parameter for Developing the Geomagnetic Storm Prediction Models

It is well demonstrated that fast and wide halo coronal mass ejections (CMEs) can produce geomagnetic storms. Several attempts to predict well in advance the geomagnetic storms have been carried out. Moon et al. [2005], and Kim et al. [2008, 2010] reported on an important geo-effective parameter named direction parameter (D); defined as the ratio of the shortest to the longest distance measured from the solar disk center to the CME front. Such parameter quantifies the asymmetry degree of the CME shape to indicate how much CME propagation can be directed to Earth. They also developed an empirical model for predicting geomagnetic storms; their occurrence and strength as well. Their model is based on the CME parameters such as CME speed (V), CME longitude (L), magnetic field orientation (M), and the direction parameter (D). They showed that Dst index is best correlated with D. Here, we define another empirical model relying on the coronal hole influence parameter (CHIP) which was introduced by Gopalswamy et al. [2009] as another geo-effective CME parameter that can decide the CME propagation towards to or away from the Sun-Earth line. This is important for developing the empirical prediction models. The results show that D can be successfully replaced by the CHIP in the storm prediction model and pertain similar predicted Dst values with an accuracy 56 % for both parameters. We also investigate the correlation between the Dst index versus CME speed, sheath speed, and ICME speed for a sample of 19 disc-center MCs and 13 disc-center non-MCs. The correlation is found to be the strongest for MCs (CC=-0.69, -0.82, and-0.84, respectively) and the smallest for non-MCs (CC=-0.20, -0.33, -0.49, respectively). We also investigate the best set of parameters which depend on CME-ICME properties and can decide the most important parameter for Dst prediction. The results show that ICME speed is better correlated with Dst than CME speed (CC=-0.77, and -0.61, respectively) but due to its short forewarning time (30 minutes), we use the CME speed in our empirical model.