Prediction technique for extended geomagnetic storm forecast
The prediction of the geomagnetic storm maximum amplitude (the negative peak Dst value) several hours before the storm maximum (red line) and the prediction of the Dst index 2-4 hours ahead (blue points) is performed in real time using the hourly averaged solar wind observations. The prediction algorithm of the Dst maximal amplitude is based on the relation of the solar wind parameters in the beginning of the storm development with the ultimate storm strength.
If the sharp jump of solar wind parameters was detected than a suddenly developing major storm (Dst<-100) is predicted, for which fast saturation is typical. In this case the prediction of lower and upper limits of the negative peak value of the Dst index (red line) is made in the beginning of the storm development.
A lower limit of the storm strength is estimated on basis of steady-state solution of the differential equation of the Dst index evolution (introduced by Burton et al, 1975). An upper limit estimate is made on basis of choice of the intermediate point between the current state and saturation point on the storm saturation trajectory. Gradually developing geomagnetic storms are characterized by the longer saturation interval and usually do not achieve the steady-state solution of the differential equation (the saturation point). For this kind of storms the Dst index achieves its maximum at the intermediate point on the storm saturation trajectory that is considered as the storm strength prediction (red line).
Advance warning time about the storm strength on average is 6 hours and varies from 1 to 22 hours.
Using the differential equation of the Dst index evolution the prediction of the Dst index is made in the storm development phase on average 2-3 hours ahead depending on storm developing behavior and on average 3-4 hours in the recovery storm phase, for which the relatively monotonous dynamics is typical.
The prediction level of storm is kept active for some time also after the real storm maximum is achieved, in order to account for possible second maximum.
The values of the Dst from Kyoto during the last 24 hours are shown on the figure by blue line. The recovered Dst index during the last 24 hours according to the solar wind data is presented by cyan line. Every six hours the Dst recovery starts from the Dst index, Kyoto if it is available.
In case of the data gaps in the solar wind the observations of the the previous hour are used instead of non-observed data. In case of data gaps in the Dst, Kyoto either the recovered Dst value or the Dst index from the previous hour are substituted for the missed Dst value. In this case the prediction quality is lower. So in the case of data gaps the vertical red line appears on the figure.
Hourly averaged observations of Bz (nT, GSM) during the last 24 hours are shown on the figure by the green line. Black line indicates the hourly averaged observations of the solar wind plasma speed (km/s) during the last 24 hours. These data are shifted in time forward to account for propagation time from L1 to Earth.
Ref: Podladchikova,T. V., and A. A. Petrukovich (2012), Extended geomagnetic storm forecast ahead of available solar wind measurements, Space Weather, 10, S07001, doi:10.1029/2012SW000786.