Prediction of magnetic storms by nonlinear models

摘要: The strong correlation between magnetic storms and southward interplanetary field (IMF) is well known from linear prediction filter studies using the Dst IMF data. However, filters change significantly one storm to another thus are limited in their predicting ability. Previous have indicated nonlinearity magnetospheric re- sponse as ring current decay rate varies with value during storms. We present this letter non- models for evolution of based on OMNI database 1964-1990. When solar wind data available advance, can be predicted Solar data, however, not most time or avail- able typically an hour less advance. Therefore, we developed nonlinear predictive alone. In absence these cannot predict onset, but evolution, may identify intense moderate ones. input-output model autonomous alone, a combination two used forecasting tools space weather.