IMO-calendar meteor outburst predictions 2019-2026: GMN orbits show an excess for 10 of 78 testable encounters (1% in placebo years), and most detected peaks are within 1 hour

Question: do published dust-trail predictions of meteor-shower outbursts identify real activity, and are their times and rates correct? Approach: we took all 123 time-specific predictions (105 encounters) from the International Meteor Organization Meteor Shower Calendars 2019-2026 and tested each one with 3.67 million Global Meteor Network (GMN) multi-station orbits (2019-01-01 to 2026-11-08). We counted shower members in the predicted window (time plus and minus 1 h) and normalized by an observing-effort value: non-member trajectories weighted by the sine of the apparent radiant altitude, with a zenith-attraction correction. We compared each count with the same UT window on adjacent days and with the same solar longitude in other years, with negative-binomial tests. A placebo test applied the same procedure in years without a prediction. Results: 78 encounters were testable. 10 show a significant excess (12.8 %, 95 % interval 7.1-22.0 %), against 3 of 308 placebo windows (1.0 %); Fisher p = 1.2e-5. Six analysis variants give 7-11 detections. Detected outbursts include alpha-Monocerotids 2019, Arids 2021, tau-Herculids 2022, 46P/Wirtanen 2023, Draconids 2024, Leonids 2025 and September epsilon-Perseids 2026. For the 21 predictions in detected encounters, the median absolute timing error is 0.45 h (16 of 21 within 1 h). Most detected outbursts had no predicted rate or a 'weak' label. Predictions with ZHR 100-490 (Ursids 2020, Leonids 2022, Arids 2021 1995 trail, Draconids 2025, Perseid filament 2020) were not detected; filament predictions succeeded in 1 of 11. A census of 273 GMN shower codes finds 36 excess episodes, of which 29 were not predicted; the largest is a DNO (IAU 915) excess on 2026-02-23 recorded by many New Zealand stations. Limitations: early years and the southern sky have little GMN coverage (27 encounters not testable); video data cannot test radar-only outbursts; the Perseid-equivalent rate scale is approximate; the prediction list is our reading of the calendar text.

meteor showers · meteor outbursts · dust trails · forecast verification · Global Meteor Network · IMO Meteor Shower Calendar · tau-Herculids · Draconids · Leonids · September epsilon-Perseids · placebo test · zenith attraction