Pierdavide Coisson Antti Pulkkinen Jeffrey J. Love 2016 <p><span>Analysis is made of the long-term statistics of three different measures of ground level, storm time geomagnetic activity: instantaneous 1&nbsp;min first differences in horizontal intensity Δ</span><i>B</i>_<i>h</i><span>, the root-mean-square of 10 consecutive 1&nbsp;min differences </span><i>S</i><span>, and the ramp change </span><i>R</i><span> over 10&nbsp;min. Geomagnetic latitude maps of the cumulative exceedances of these three quantities are constructed, giving the threshold (nT/min) for which activity within a 24&nbsp;h period can be expected to occur once per year, decade, and century. Specifically, at geomagnetic 55°, we estimate once-per-century Δ</span><i>B</i>_<i>h</i><span>, </span><i>S</i><span>, and </span><i>R</i><span> exceedances and a site-to-site, proportional, 1 standard deviation range [1&nbsp;</span><i>σ</i><span>, lower and upper] to be, respectively, 1000, [690, 1450]; 500, [350, 720]; and 200, [140, 280]&nbsp;nT/min. At 40°, we estimate once-per-century Δ</span><i>B</i>_<i>h</i><span>, </span><i>S</i><span>, and </span><i>R</i><span> exceedances and 1&nbsp;</span><i>σ</i><span> values to be 200, [140, 290]; 100, [70, 140]; and 40, [30, 60] nT/min.</span></p> application/pdf 10.1002/2016GL068664 en American Geophysical Union Global statistical maps of extreme-event magnetic observatory 1 min first differences in horizontal intensity article