Statistical analysis and verification of 3‐hourly geomagnetic activity probability predictions
Authors:
Open Access
- 21 November 2015
- journal article
- research article
- Published by American Geophysical Union (AGU) in Space Weather
- Vol. 13 (12) , 831-852
Abstract
The Space Environment Prediction Center (SEPC) has classified geomagnetic activity into four levels: quiet to unsettled (Kp < 4), active (Kp = 4), minor to moderate storm (Kp = 5 or 6), and major to severe storm (Kp > 6). The 3‐hourly Kp index prediction product provided by the SEPC is updated half hourly. In this study, the statistical conditional forecast models for the 3‐hourly geomagnetic activity level were developed based on 10 years of data and applied to more than 3 years of data, using the previous Kp index, interplanetary magnetic field, and solar wind parameters measured by the Advanced Composition Explorer as conditional parameters. The quality of the forecast models was measured and compared against verifications of accuracy, reliability, discrimination capability, and skill of predicting all geomagnetic activity levels, especially the probability of reaching the storm level given a previous “calm” (nonstorm level) or “storm” (storm level) condition. It was found that the conditional models that used the previous Kp index, the peak value of BtV (the product of the total interplanetary magnetic field and speed), the average value of Bz (the southerly component of the interplanetary magnetic field), and BzV (the product of the southerly component of the interplanetary magnetic field and speed) over the last 6 h as conditional parameters provide a relative operating characteristic area of 0.64 and can be an appropriate predictor for the probability forecast of geomagnetic activity level.Keywords
Funding Information
- National Geophysical Data Center of the National Oceanic and Atmospheric Administration (NOAA)
- National Basic Research Programs of China (2012CB825606, 2011CB811406)
This publication has 18 references indexed in Scilit:
- Operational Space Weather Services in National Space Science Center of Chinese Academy of SciencesSpace Weather, 2015
- Validating the Rice neural network and the Wing Kp real‐time modelsSpace Weather, 2014
- Comparison of neural network and support vector machine methods for Kp forecastingJournal of Geophysical Research: Space Physics, 2013
- The Kp index and solar wind speed relationship: Insights for improving space weather forecastsSpace Weather, 2013
- Improvements in short‐term forecasting of geomagnetic activitySpace Weather, 2012
- A sensitive geomagnetic activity index for space weather operationSpace Weather, 2010
- Real‐time prediction of magnetospheric activity using the Boyle IndexSpace Weather, 2009
- Black box modeling of magnetospheric dynamics to forecast geomagnetic activitySpace Weather, 2004
- Probabilistic forecasting of geomagnetic indices using solar wind air mass analysisSpace Weather, 2004
- Further evidence on the correlation between transverse fluctuations in the interplanetary magnetic field andKpJournal of Geophysical Research, 1969