资源环境科技发展态势分析平台

Geomagnetic indices are routinely used to characterize space weather event intensity. The D-ST index is well resolved but is only available over five solar cycles. The aa index extends over 14 cycles but is highly discretized with poorly resolved extremes. We parameterize extreme aa activity by the annual-averaged top few percent of observed values, show that these are exponentially distributed, and they track annual DST index minima. This gives a 14-cycle average of similar to 4% chance of at least one great (D-ST < -500 nT) storm and similar to 28% chance of at least one severe (D-ST < -250 nT) storm per year. At least one D-ST = -809 [-663,-955] nT event in a given year would be a 1:151 year event. Carrington event estimate D-ST similar to -850 nT is within the same distribution as other extreme activity seen in aa since 1868 so that its likelihood can be deduced from that of more moderate events. Events with D-ST < -1, 000 nT are in a distinct class, requiring special conditions.

Plain Language Summary Here we use measurements of disturbances in the Earth's magnetic field that go back to 1868, and we present a novel way of analyzing the data to identify the largest magnetic storms going back some 80 years longer than has been done before. As a result, we are able to state the chance of at least one superstorm occurring in a year. We find that on average there is a 4% (28%) chance of at least one great (severe) storm per year and a 0.7% chance of a Carrington class storm per year, which can be used for planning the level of mitigation needed to protect critical national infrastructure.