Geomagnetic activity can affect communications, navigation systems, satellite health, power grids, and space travel. Kp is one of the most common indices used to indicate the severity of the global magnetic disturbances in near-Earth space. Predictions update every 15 minutes. Magnetically active times, e.g., Kp > 5, are notoriously difficult to predict, precisely the times when such predictions are crucial to the space weather users. Taking advantage of the routinely available solar wind measurements at the Lagrangian point (L1) and the nowcast of Kps, Kp forecast models based on neural networks were developed with a focus on improving the forecast for active times.

12-hour and 24-hour Plot Description

These plots show the latest predicted planetary geomagnetic activity index (Kp) from the U. S. Air Force Weather Agency Wing Kp model. The plots are of 1-hour and 4-hour Kp predictions, along with the observed Estimated Kp. 1-hour predictions are plotted with filled-in circles, 4-hour predictions are plotted in plus (+) symbols, and the solid blue line is the observed Estimated Kp index. The predictions are color coded according to the amount of solar wind data that was available for each model run; green (most data) and yellow (little data). The solid horizontal green or yellow bars show the 3-hour interval over which the latest predictions are valid. Vertical error bars are plotted to show the 50% confidence interval.

Prediction points in the top panel are plotted at the model valid time (run time + propagation time), which is usually 30-90 minutes into the future for the 1-hour prediction and 210-270 minutes into the future for the 4-hour prediction. The dashed red line indicates the lowest alert level, G1 (minor), on the NOAA Space Weather Scale for Geomagnetic Storms.

The bottom Lead Time panel shows an estimate of the actual lead time for the 1-hour and 4-hour predictions. The lead time is the time it takes for the solar wind to propagate from the ACE satellite, at L1, to Earth. The actual lead time depends on solar wind speed and can be greater than or less than one hour depending on the solar wind speed. The left plot legend shows the 1-hour lead time and the 4-hour lead time is shown on the right.

Missing Kp values, Lead Time values, or model output indicates the data are not available at SWPC.

7-Day Model Performance Plot Description

The 7-day Model Performance plot shows the latest Wing Kp model output over the past 7 days with statistical characterizations in two panels.

The top panel plots the magnitude of the predicted Kp index. The model provides two predictions:

1. 1-hour prediction, plotted with filled-in circles

2. 4-hour prediction, plotted in plus (+) symbols.

The predictions are color coded according to the amount of solar wind data that were available for each model run, ranging from green (most data) to yellow (little data). The solid blue line is the observed USAF Estimated Kp Index. Simple statistics are given at the top of the plot for the 1-hour model that characterize the model's performance in terms of accuracy (rms error), bias (mean error), association (correlation), and skill or prediction efficiency (relative error).

The bottom Lead Time panel plots the lead time of the prediction. The lead time is the time it takes for the solar wind to propagate from the ACE satellite, at L1, to the Earth. The actual lead time depends on solar wind speed and can be greater than or less than one hour depending on the solar wind speed. The left plot legend shows the 1-hour lead time and the 4-hour lead time is shown on the right.

For further information see: Kp Forecast Models, Wing, Simon (2005) et al, J. Geophy. Res, Vol 110, A04203, doi:10.1029/2004JA010500.