K-Cor Coronagraph Tracks Coronal Mass Ejections Linked to Solar Energetic Particles

The Sun’s most dangerous eruptions may announce themselves earlier than anyone thought possible. A team of scientists led by O. C. St. Cyr of NASA’s Goddard Space Flight Center has completed the most comprehensive study to date of coronal mass ejections, or CMEs, observed in the innermost corona by the ground-based K-Coronagraph at the Mauna Loa Solar Observatory in Hawaii, and their findings carry a striking message for the future of astronaut safety: the eruptions that produce hazardous solar energetic particle events look measurably different, and behave measurably faster, in the very lowest reaches of the solar atmosphere than the benign eruptions that do not. The research, published in the journal Solar Physics, analyzed 27 CMEs observed by K-Cor between late 2013 and November 2022, when the eruption of the Mauna Loa volcano forced the suspension of observatory operations, and compared them against a carefully constructed control sample of 52 eruptions that produced no detectable energetic particles.

Solar energetic particle events, or SEPs, are sudden surges in the population of electrons, protons, and heavy ions speeding through interplanetary space. They are widely understood to arise from particles accelerated by magnetic reconnection in solar flares and at shock waves that form ahead of fast CMEs. The most intense events can deliver radiation doses lethal to unshielded astronauts and can damage spacecraft electronics, which is why predicting them has become a priority as agencies prepare for sustained human missions to the Moon and Mars. For decades, forecasters have relied on space-based coronagraphs such as the Large Angle and Spectrometric Coronagraph on the Solar and Heliospheric Observatory, which images the corona only above roughly 2.3 solar radii from the Sun’s center. That limitation matters because the fastest SEP particles can reach Earth in a matter of tens of minutes, arriving before the parent CME has even entered the field of view of such instruments. By the time a space-based coronagraph confirms a fast eruption, the radiation event may already be underway.

The K-Coronagraph, installed at the 3,440-meter altitude Mauna Loa observatory in late 2013, was designed to close precisely this gap. As an internally occulted, ground-based instrument, it images the polarization brightness of the corona, produced by Thomson scattering of photospheric light off free electrons, in a near-infrared passband spanning 720 to 750 nanometers, and it does so starting at just 1.05 solar radii and extending to 3.0. Its 15-second cadence represented a dramatic improvement over the 3-minute cadence of its predecessors, the Mark-3 and Mark-4 coronameters, and a fully automated processing pipeline reduced data latency to 2.5 minutes, compared with as much as 24 hours previously. The pipeline includes an automatic CME detection algorithm that scans processed images and dispatches email alerts containing the eruption’s detection time, height, position angle, and speed, along with a “heartbeat” signal that tells forecasters the observatory is watching and sees nothing alarming.

To identify which of the roughly 600 CMEs detected by K-Cor since 2013 were accompanied by SEPs, the team drew on an updated catalog of events including protons of approximately 25 MeV observed by the High Energy Telescopes aboard the twin STEREO spacecraft and by instruments on near-Earth satellites such as SOHO’s Electron Proton Helium Instrument and Energetic and Relativistic Nuclei and Electron experiment. The combination proved powerful because of orbital geometry: at the start of K-Cor operations near the maximum of Solar Cycle 24, STEREO-B sat on the far side of the Sun, magnetically well connected to the Sun’s eastern limb, a region from which energetic particles are difficult to detect from Earth. After contact with STEREO-B was lost in October 2014, STEREO-A drifted into a similarly advantageous position. Of 215 individual SEP events catalogued during the study window, most occurred during the observatory’s daily data gaps, and in five cases K-Cor was observing but saw no CME, generally because the eruption was far from the plane of the sky where the instrument’s sensitivity falls off, or because elevated instrumental stray light degraded early observations. That left 27 confirmed SEP-associated CMEs.

The comparison with the control sample, drawn from a 2023 catalog of near-limb K-Cor events compiled by Song and colleagues, produced differences that were striking in their consistency. The average and median angular widths of the SEP-producing CME drivers were a factor of two larger than those of the non-SEP events, both in the inner corona and in the middle corona where space-based instruments took over. The average speed of SEP CMEs in the inner corona was more than twice that of their quiet counterparts, and nearly twice as fast in the middle corona. More than 80 percent of all the CMEs in both samples were accelerating as they climbed through the inner corona, but the average acceleration for the SEP events was almost a factor of three larger. Mass estimates from the Coordinated Data Analysis Workshops catalog, derived from LASCO brightness measurements, told the same story: the average SEP CME was nearly three times more massive than the average non-SEP event, with median masses of 8.7 times ten to the fifteenth grams versus 2.1 times ten to the fifteenth.

The study also found a possible new telltale signature in extreme-ultraviolet images from the Solar Dynamics Observatory’s Atmospheric Imaging Assembly. Among 64 events whose low-coronal morphology could be classified, nine displayed what the authors call very hot flux ropes, structures visible only in the hottest coronal channels and often trailing the cooler erupting bubble by roughly 0.1 solar radii. Seven of those nine, or 78 percent, were associated with SEP events, making the hot flux rope one of the strongest indicators of an impending radiation storm, though the authors caution it is not a necessary one, since eight of the fifteen classified SEP CMEs lacked the feature. The team further tied CME kinematics to flare timing, showing directly that over 70 percent of CMEs in both samples stood below five solar radii at the moment of peak soft X-ray flux, a relationship previously established only through extrapolation, and one that SEP prediction models may be able to exploit.

Perhaps the most operationally significant number in the study is 28 minutes. That is the average interval between K-Cor’s first detection of an SEP-associated CME and LASCO C2’s first detection of the same event, a head start delivered before any accounting of the hours of telemetry latency that can delay space-based imagery. K-Cor’s 84 percent detection rate of SEP CMEs occurring during its observing windows, 27 of a possible 32, was commensurate with the 92 percent rate achieved by the earlier Mk3 and Mk4 instruments over three decades. Only one of the 27 associated SEP events registered at the S1 level on NOAA’s solar radiation storm scale in the near-Earth environment, and three more would have exceeded that threshold as measured by STEREO-A far from Earth, but the authors note that the K-Cor era coincided with unusually weak solar activity, with only ten CMEs exceeding 1,800 kilometers per second compared with 37 in an equivalent span during the Mark-4 era. The Sun has produced 80 percent more fast CMEs in the less than three years since the 2022 hiatus than during the nine years of K-Cor observations themselves.

The timing of this work could hardly be more pointed. SOHO, launched three decades ago, will eventually be decommissioned, and its CME monitoring role is passing to NOAA’s new Compact Coronagraphs. CCOR-1, launched in June 2024 aboard GOES-19, and CCOR-2, launched in September 2025 on the Space Weather Follow-On mission now stationed at the Lagrange 1 point, both begin their fields of view at 3.5 solar radii or beyond, image every 15 minutes, and require a second image before a speed can be computed. The authors’ arithmetic is sobering: for a CME traveling at the study’s average SEP-associated speed of just over 800 kilometers per second, a 25 MeV proton following the nominal Parker spiral path reaches 1 astronomical unit in about 43 minutes, meaning the radiation event would already be underway by the time a CCOR speed estimate became available. Higher-energy particles, the ones that matter most for aviation and human spaceflight, arrive even sooner.

With Mauna Loa’s access road repaired, routine K-Cor observations resumed on March 31, 2026, and the instrument once again feeds near-real-time alerts to NASA’s Community Coordinated Modeling Center SEP Scoreboard. The study’s authors are careful to acknowledge the limitations of a single ground-based observatory, hemmed in by weather, daylight, staffing, and geography, and their sample excludes the rare, most extreme events simply because none occurred during an observing window. Yet their conclusion is unambiguous: ground-based coronagraphy, with its low observational threshold, rapid cadence, and 2.5-minute data latency, offers a warning capability that no current or planned space-based instrument can match on its own, and the distinctive properties of SEP-producing eruptions in the inner corona, their greater width, speed, acceleration, and mass, provide forecasters with physical signatures to watch for. As humanity prepares to send crews beyond Earth’s protective magnetosphere, the view from a mountaintop in Hawaii may prove to be one of the best early-warning systems available.