Performance of Major Flare Watches from the Max Millennium Program (2001 - 2010)

D. S. Bloomfield, P. T. Gallagher, W. H. Marquette, R. O. Milligan, R. C. Canfield

Research output: Contribution to journalArticlepeer-review

12 Citations (Scopus)


The physical processes that trigger solar flares are not well understood, and significant debate remains around processes governing particle acceleration, energy partition, and particle and energy transport. Observations at high resolution in energy, time, and space are required in multiple energy ranges over the whole course of many flares to build an understanding of these processes. Obtaining high-quality, co-temporal data from ground- and space- based instruments is crucial to achieving this goal and was the primary motivation for starting the Max Millennium program and Major Flare Watch (MFW) alerts, aimed at coordinating observations of all flares ≥ X1 GOES X-ray classification (including those partially occulted by the limb). We present a review of the performance of MFWs from 1 February 2001 to 31 May 2010, inclusive, which finds that (1) 220 MFWs were issued in 3407 days considered (6.5 % duty cycle), with these occurring in 32 uninterrupted periods that typically last 2 - 8 days; (2) 56% of flares ≥ X1 were caught, occurring in 19 % of MFW days; (3) MFW periods ended at suitable times, but substantial gain could have been achieved in percentage of flares caught if periods had started 24 h earlier; (4) MFWs successfully forecast X-class flares with a true skill statistic (TSS) verification metric score of 0.500, that is comparable to a categorical flare/no-flare interpretation of the NOAA Space Weather Prediction Centre probabilistic forecasts (TSS = 0.488).
Original languageEnglish
Pages (from-to)411-427
Number of pages17
JournalSolar Physics
Early online date04 Jan 2016
Publication statusPublished - 01 Feb 2016


Dive into the research topics of 'Performance of Major Flare Watches from the Max Millennium Program (2001 - 2010)'. Together they form a unique fingerprint.

Cite this