Empirical Predictability of the Geo-Effectiveness of CMEs: A Solar Wind Perspective

A major raison d’être of space weather research is the prediction of space weather events in general, and specifically the prediction of the geo-effectiveness of transients (primarily CMEs). One key component of those efforts is the prediction of the solar wind IMF for as long in advance and as well as possible, recognizing the fact that the IMF is one of the major contributors to energy input into Earth’s magnetosphere. A comprehensive prediction capability is undoubtedly still quite a number of years away. Any sufficiently successful approach will by nature have to be multi-disciplinary: observations of the Sun to capture emerging CMEs, modeling to predict the propagation and large-scale development of CMEs, and plasma parameter predictions to derive a sufficiently long-term prediction (i.e., several hours to a day) for the L1 point. At L1, prediction and data will come together for direct comparison in real-time. Rudiments of such a prediction chain may exist. To aid these efforts, we have been focusing on increasing our knowledge of the structure of the solar wind both inside and outside of CMEs. We have developed a framework that is able to statistically investigate the occurrence and characteristic of time series patterns of L1 solar wind observations. In this presentation we will contrast “typical”, representative time series behavior of relevant solar wind parameters (e.g., IMF, speed, dynamic pressure, etc.) during and outside of CMEs. We focus on intermediate (3 hours) to long (24 hours) time series, exploring in particular the range of parameters and time scales over which solar wind data can possibly be forecast in the future. This study of time series patterns goes beyond the study of solar wind threshold values (that are used to automatically identify CME intervals) as well as beyond purely statistical and superposed epoch analysis studies that are used to determine heavily averaged values and time series profiles during CME encounters. A “production type” computational framework capable of processing historic data as well as real-time L1 data (for rudimentary empirical prediction) will be presented.