The "middle corona" is a critical transition between the highly disparate physical regimes of the lower and outer solar corona. Nonetheless, it remains poorly understood due to the difficulty of observing this faint region (1.5-3 solar radii). New observations from the GOES Solar Ultraviolet Imager in August and September 2018 provide the first comprehensive look at this region's characteristics and long-term evolution in extreme ultraviolet (EUV). Our analysis shows that the dominant emission mechanism here is resonant scattering rather than collisional excitation, consistent with recent model predictions. Our observations highlight that solar wind structures in the heliosphere originate from complex dynamics manifesting in the middle corona that do not occur at lower heights. These data emphasize that low-coronal phenomena can be strongly influenced by inflows from above, not only by photospheric motion, a factor largely overlooked in current models of coronal evolution. This study reveals the full kinematic profile of the initiation of several coronal mass ejections, filling a crucial observational gap that has hindered understanding of the origins of solar eruptions. These new data uniquely demonstrate how EUV observations of the middle corona provide strong new constraints on models seeking to unify the corona and heliosphere.

ABSTRACT Hinode /Solar Optical Telescope (SOT) observations reveal two new dynamic modes in quiescent solar prominences: large-scale (20–50 Mm) “arches” or “bubbles” that “inflate” from below into prominences, and smaller-scale (2–6 Mm) dark turbulent upflows. These novel dynamics are related in that they are always dark in visible-light spectral bands, they rise through the bright prominence emission with approximately constant speeds, and the small-scale upflows are sometimes observed to emanate from the top of the larger bubbles. Here we present detailed kinematic measurements of the small-scale turbulent upflows seen in several prominences in the SOT database. The dark upflows typically initiate vertically from 5 to 10 Mm wide dark cavities between the bottom of the prominence and the top of the chromospheric spicule layer. Small perturbations on the order of 1 Mm or less in size grow on the upper boundaries of cavities to generate plumes up to 4–6 Mm across at their largest widths. All plumes develop highly turbulent profiles, including occasional Kelvin–Helmholtz vortex “roll-up” of the leading edge. The flows typically rise 10–15 Mm before decelerating to equilibrium. We measure the flowfield characteristics with a manual tracing method and with the Nonlinear Affine Velocity Estimator (NAVE) “optical flow” code to derive velocity, acceleration, lifetime, and height data for several representative plumes. Maximum initial speeds are in the range of 20–30 km s −1 , which is supersonic for a ∼10,000 K plasma. The plumes decelerate in the final few Mm of their trajectories resulting in mean ascent speeds of 13–17 km s −1 . Typical lifetimes range from 300 to 1000 s (∼5–15 minutes). The area growth rate of the plumes (observed as two-dimensional objects in the plane of the sky) is initially linear and ranges from 20,000 to 30,000 km 2 s −1 reaching maximum projected areas from 2 to 15 Mm 2 . Maximum contrast of the dark flows relative to the bright prominence plasma in SOT images is negative and ranges from −10% for smaller flows to −50% for larger flows. Passive scalar “cork movies” derived from NAVE measurements show that prominence plasma is entrained by the upflows, helping to counter the ubiquitous downflow streams in the prominence. Plume formation shows no clear temporal periodicity. However, it is common to find “active cavities” beneath prominences that can spawn many upflows in succession before going dormant. The mean flow recurrence time in these active locations is roughly 300–500 s (5–8 minutes). Locations remain active on timescales of tens of minutes up to several hours. Using a column density ratio measurement and reasonable assumptions on plume and prominence geometries, we estimate that the mass density in the dark cavities is at most 20% of the visible prominence density, implying that a single large plume could supply up to 1% of the mass of a typical quiescent prominence. We hypothesize that the plumes are generated from a Rayleigh–Taylor instability taking place on the boundary between the buoyant cavities and the overlying prominence. Characteristics, such as plume size and frequency, may be modulated by the strength and direction of the cavity magnetic field relative to the prominence magnetic field. We conclude that buoyant plumes are a source of quiescent prominence mass as well as a mechanism by which prominence plasma is advected upward, countering constant gravitational drainage.

In investigating the long-term variability in the solar soft X-ray flux, we discovered a sudden increase ("the step") at the onset of Cycle 23. This step occurred in August 1997 when the X-ray emission observed by the Yohkoh SXT instrument increased within two rotations of the Sun from some of the lowest values seen during solar minimum to about 30% of the rate seen at the following maximum. This was not the result of the emergence of one or two active regions but seemed to be a global phenomenon. Further, there was a north-south asymmetry in the step, activity in the northern hemisphere of the Sun emerging first followed a month later by activity in the south. The strong-field magnetic flux shows the same rapid increase in activity. Preliminary results from prior cycles indicate similar step functions but the data are sparse and will need considerably more careful relative calibration and analysis before we can draw firm conclusions. This result imposes new constraints on the prediction of solar activity from any combination of magnetic dynamo and flux transport models.