Orphan Penumbrae: Structure and Dynamics

• Orphan penumbrae are filamentary photospheric structures that mimic sunspot penumbrae but lack an umbra and are typically located along polarity inversion lines.
• They show organized Evershed-like flows and nearly horizontal magnetic fields, with evidence supporting both siphon flow and magnetoconvective interpretations.
• Their magnetic topology ranges from flat Ω-shaped loops to segments of flux ropes, influencing the formation of active-region filaments.

Orphan penumbrae (OPUs) are photospheric filamentary structures in active regions that resemble regular sunspot penumbrae but are not connected to an umbra. Across case studies and statistical analyses, they are consistently found near polarity inversion lines (PILs), in strongly inclined or nearly horizontal magnetic fields, and with penumbra-like filament morphology and organized flows. Their magnetic interpretation varies with context: some observations identify OPUs with low-lying, flat $\Omega$-shaped flux systems that later submerge, whereas others place them in the photospheric segment of a filament-related flux rope, including cases in which OPU fields ascend and contribute directly to H$\alpha$ filament formation (Kuckein et al., 2011, Jurcak et al., 2014, Buehler et al., 2016, Durán et al., 10 Jul 2025).

1. Definition, identification, and morphology

The term “orphan penumbrae” refers to penumbral-like structures not connected to any umbra, following usage traced to Zirin. In continuum and G-band diagnostics, OPUs appear as elongated filamentary patches with bright heads and darker tails, closely resembling ordinary penumbrae. Jurčák, Bellot Rubio, and Sobotka emphasized that they are located between opposite-polarity magnetic patches, typically in or right on PILs, and that their filaments are aligned with the horizontal magnetic field (Jurcak et al., 2014).

Several morphological traits recur across the literature. In compact active regions, OPUs can form a bridge connecting different small groups of pores together, especially along a very narrow PIL, as in NOAA 10781 (Kuckein et al., 2011). In Hinode data they can appear almost identical to regular penumbrae in G band, with bright penumbral grains at the heads, darker filament bodies, and darker tails; however, they lack the darkest intensity values associated with an umbra, are slightly brighter in G band, and are clearly brighter in Ca II H. Jurčák, Bellot Rubio, and Sobotka reported a most frequent G-band intensity of about $0.65\,I_{\rm QS}$ for both regular and orphan penumbrae, while the Ca II H intensity distribution peaks near $0.85\,I_{\rm QS}$ for orphan penumbrae versus about $0.75\,I_{\rm QS}$ for regular penumbrae (Jurcak et al., 2014).

Later work broadened the morphological picture without altering its basic definition. Castellanos Durán et al. described a broad range of OPU shapes, from straight ribbon-like structures to fan-shaped and curved configurations, but stressed that they remain one filament thick, lack a central dark core, and do not show the spine–intra-spine pattern characteristic of sunspot penumbrae. Their filaments nevertheless retain the canonical head–body–tail organization of penumbral fine structure (Durán et al., 10 Jul 2025).

2. Magnetic structure and topology

The magnetic field of OPUs has been reconstructed primarily through spectropolarimetric inversions. In the multi-height study of NOAA 10781, Kuckein et al. used the chromospheric He I 10830 Å triplet and the photospheric Si I 10827 Å line to separate upper-chromospheric and upper-photospheric structure. In that framework, the magnetic components were expressed in the local frame as

$$B_\mathrm{vert} = f\,|B|\cos\gamma,\qquad B_\mathrm{hor} = \sqrt{f}\,|B|\sin\gamma,$$

with $\gamma=90^\circ$ corresponding to a horizontal field. The OPU region on 2005 July 5 showed photospheric horizontal fields reaching $1000{-}1100$ G in Si I and chromospheric horizontal fields up to about $800$ G in He I, concentrated at the pores and orphan penumbrae along the PIL (Kuckein et al., 2011).

A complementary single-height local-reference-frame analysis by Jurčák, Bellot Rubio, and Sobotka found that OPU filament bodies are typically within about $\pm 10^\circ$ of horizontal,

$\alpha$0

with fields becoming more inclined near heads and tails, where the loops emerge and submerge. In their two Hinode cases, the field crosses the PIL from positive to negative polarity in a normal configuration, and the overall geometry is that of $\alpha$1-shaped flux ropes whose filamentary apex lies in the photosphere (Jurcak et al., 2014).

Buehler et al. extended the magnetic diagnosis to three optical-depth nodes, $\alpha$2, using spatially coupled SPINOR inversions. In AR 10953, the OPU in region B had $\alpha$3 G at $\alpha$4, $\alpha$5 G at $\alpha$6, and $\alpha$7 G at $\alpha$8, with local inclinations $\alpha$9 and $0.65\,I_{\rm QS}$0, respectively. These values imply that the horizontal component $0.65\,I_{\rm QS}$1 dominates over the vertical component $0.65\,I_{\rm QS}$2 throughout the sampled layers. After ascent into higher layers, the filament channel exhibited an average horizontal field of about $0.65\,I_{\rm QS}$3 G at $0.65\,I_{\rm QS}$4, with local maxima up to about $0.65\,I_{\rm QS}$5 G, and the azimuth took on an inverse configuration relative to the PIL, suggesting a flux rope topology for the new H$0.65\,I_{\rm QS}$6 filament (Buehler et al., 2016).

The 2025 statistical study preserved the same topological conclusion in a broader sample. Castellanos Durán et al. found that OPUs are associated with typically $0.65\,I_{\rm QS}$7-shaped magnetic field configurations, where opposite-polarity fields predominate at the two ends of the OPU, and that a significant horizontal canopy field is often present at $0.65\,I_{\rm QS}$8, persisting even when the photospheric filaments have started to disappear (Durán et al., 10 Jul 2025).

3. Flows and fine-structure dynamics

OPUs are not merely static magnetic structures. Their filaments host organized penumbra-like flows whose kinematic pattern closely parallels the Evershed flow. Jurčák, Bellot Rubio, and Sobotka found that line-of-sight velocities are low near filament heads, increase along the filament body, and peak at the tails, where the magnetic field is stronger and the flow submerges. In their Hinode sample, maximum LOS velocities reached about $0.65\,I_{\rm QS}$9 in spectropolarimetric maps and up to $0.85\,I_{\rm QS}$0 in individual filaments (Jurcak et al., 2014).

A notable feature of their analysis is that neighboring OPU filaments can display opposite flow directions. This was interpreted as evidence for different magnetic connectivity between adjacent $0.85\,I_{\rm QS}$1-loops rather than a single coherent flow system. They further argued that the flow direction follows the weaker-field to stronger-field footpoint ordering, which is the defining signature of a siphon flow. On that basis they proposed that the fast flows in orphan penumbrae may be caused by the siphon flow mechanism, and that the Evershed flow in regular penumbrae may likewise be a manifestation of siphon flows (Jurcak et al., 2014).

Buehler et al. observed a comparable but somewhat lower-amplitude flow system in AR 10953. In the OPU of region B, LOS velocities were of order $0.85\,I_{\rm QS}$2 across $0.85\,I_{\rm QS}$3, while bright points tracked manually in Ca II H moved at about $0.85\,I_{\rm QS}$4 in the plane of the sky. During late stages of ascent, part of the flow appeared to be channeled above the photosphere, consistent with the rise of the field into higher layers (Buehler et al., 2016).

The 2025 survey added two further elements. First, Castellanos Durán et al. showed that OPU tail downflows can be very strong, up to about $0.85\,I_{\rm QS}$5 in the deepest node, comparable to the strongest penumbral tail downflows. Second, they identified counter-Evershed-like flows in a fraction of OPU filaments, supporting the view that OPUs are an unusually clean laboratory for studying the relation between inclined-field magnetoconvection, loop geometry, and filament-aligned flow without the organizing influence of a large umbral core (Durán et al., 10 Jul 2025). This suggests that the siphon-flow interpretation and the magnetoconvective interpretation are not always mutually exclusive: local pressure gradients and $0.85\,I_{\rm QS}$6-loop geometry can channel flows, but the broader flow system remains embedded in penumbral magnetoconvection.

4. Formation pathways, lifetimes, and decay

Two formation pathways recur in the literature. The first is emergence near the active-region PIL. In NOAA 10781, opposite polarities first separated and the PIL widened, then re-approached until the PIL became extremely narrow; only then did pores and orphan penumbrae appear along the PIL. Kuckein et al. explicitly suggested that orphan penumbrae are formed in very narrow PILs of compact active regions and are the photospheric manifestation of flux ropes in the photosphere (Kuckein et al., 2011).

The second pathway is detachment from an existing sunspot penumbra. Castellanos Durán et al. manually identified 80 individual OPUs in Hinode data and were able to classify the formation of 57 of them: 26 formed by a patch of a penumbra separating from a sunspot, 24 by new magnetic flux emerging close to the PIL of an active region, and 7 remained ambiguous because of complex surroundings. In separation events, the detached patch retains its filaments with essentially unchanged properties during and after separation; in emergence events, new bipolar flux and pores appear first, followed by the development of penumbral-like filaments (Durán et al., 10 Jul 2025).

Decay also differs by context. Jurčák, Bellot Rubio, and Sobotka described cases in which opposite-polarity patches flanking an orphan penumbra approach each other at about $0.85\,I_{\rm QS}$7, the filamentary structure shortens and shrinks over roughly $0.85\,I_{\rm QS}$8 hours, and the whole structure is interpreted as submerging. Assuming loop height comparable to width, they estimated a submergence speed of about $0.85\,I_{\rm QS}$9 (Jurcak et al., 2014).

The statistical study by Castellanos Durán et al. found lifetimes of a few hours up to roughly $0.75\,I_{\rm QS}$0 hours. In a detailed decay sequence, granulation invaded the OPU from the sides at $0.75\,I_{\rm QS}$1, while a strong horizontal canopy persisted above it at $0.75\,I_{\rm QS}$2; only later did the canopy weaken and retreat, leaving behind small plage-like flux concentrations (Durán et al., 10 Jul 2025). This pattern indicates that OPU decay can begin in deeper layers, with the upper-photospheric or low-chromospheric canopy outliving the visible filaments.

5. Relation to active-region filaments and chromospheric structure

The strongest connection between OPUs and filament physics comes from multi-height active-region studies. In NOAA 10781, the broader, diffuse part of the H$0.75\,I_{\rm QS}$3 filament lay directly above the orphan penumbrae and between pores, whereas the lower spine lay over only weak flux patches. Kuckein et al. inferred that the filament was divided in two parts, one lying in the chromosphere and another trapped in the photosphere, so that only the top of the helical structure was seen in He I. In that interpretation, the pores and orphan penumbrae at the PIL are the photospheric counterpart of an extremely low-lying filament (Kuckein et al., 2011).

Buehler et al. documented a more direct evolutionary link in AR 10953. There, an initial H$0.75\,I_{\rm QS}$4 filament existed above OPU-bearing regions A and B. After flux emergence and brightenings in region C, the initial filament was eroded and disappeared. Only then did the orphan penumbrae begin to evolve: their photospheric horizontal fields vacated $0.75\,I_{\rm QS}$5 and $0.75\,I_{\rm QS}$6, the local flux dropped by up to $0.75\,I_{\rm QS}$7, and the structures ascended with a derived vertical speed of about $0.75\,I_{\rm QS}$8. Co-spatial dark H$0.75\,I_{\rm QS}$9 seed fragments appeared at A, B, and C, elongated parallel to the PIL, and merged into a new continuous H$$B_\mathrm{vert} = f\,|B|\cos\gamma,\qquad B_\mathrm{hor} = \sqrt{f}\,|B|\sin\gamma,$$0 filament within $$B_\mathrm{vert} = f\,|B|\cos\gamma,\qquad B_\mathrm{hor} = \sqrt{f}\,|B|\sin\gamma,$$1 hours of the OPU rise (Buehler et al., 2016).

This filament-forming case does not imply that all OPUs produce large filaments. Castellanos Durán et al. noted that among 25 OPUs also observed in Hinode H$$B_\mathrm{vert} = f\,|B|\cos\gamma,\qquad B_\mathrm{hor} = \sqrt{f}\,|B|\sin\gamma,$$2 images, almost all show chromospheric fibrils above them, indicating that part of the magnetic field of the OPUs extends to the chromosphere. At the same time, only one OPU in their broader sample could be clearly associated with a larger active-region filament formation event of the type analyzed by Buehler et al. (Durán et al., 10 Jul 2025). A plausible implication is that chromospheric connectivity is common, whereas filament assembly on active-region scales requires additional topological and evolutionary conditions.

6. Comparison with regular penumbrae, interpretive disputes, and open problems

The basic comparison with regular penumbrae is stable across all four studies. OPUs share the filamentary head–body–tail morphology, nearly horizontal fields in filament bodies, stronger and more vertical fields at heads and tails, and Evershed-like flows. They differ in lacking an umbra and, therefore, lacking the large vertical background field and spine–intra-spine pattern of a sunspot penumbra. They are also typically brighter, smaller, and only one filament ring thick (Jurcak et al., 2014, Buehler et al., 2016, Durán et al., 10 Jul 2025).

Several quantitative comparisons support this view. In AR 10953, the OPU of region B was hotter than the sunspot penumbra by about $$B_\mathrm{vert} = f\,|B|\cos\gamma,\qquad B_\mathrm{hor} = \sqrt{f}\,|B|\sin\gamma,$$3 K at all three inversion nodes and weaker by about $$B_\mathrm{vert} = f\,|B|\cos\gamma,\qquad B_\mathrm{hor} = \sqrt{f}\,|B|\sin\gamma,$$4 G, while maintaining very similar inclination and azimuth. In the 2025 survey, OPU pixels occupied the same general intensity–field and velocity–polarity regimes as penumbral filaments, but the highest penumbral field strengths and darkest penumbral intensities were underrepresented or absent (Buehler et al., 2016, Durán et al., 10 Jul 2025).

The principal interpretive disagreements concern topology and dynamics rather than morphology. One line of work emphasizes flat $$B_\mathrm{vert} = f\,|B|\cos\gamma,\qquad B_\mathrm{hor} = \sqrt{f}\,|B|\sin\gamma,$$5-loops that submerge as opposite polarities converge (Jurcak et al., 2014); another identifies OPUs with the photospheric segment of a filament-related flux rope, either trapped below the chromosphere or later rising into it (Kuckein et al., 2011, Buehler et al., 2016). These views are not strictly incompatible, because they derive from different active-region contexts and different evolutionary stages. Likewise, the flow interpretation is not settled. Jurčák, Bellot Rubio, and Sobotka argued that the weaker-to-stronger-footpoint ordering strongly supports siphon flows, whereas Castellanos Durán et al. stressed that the prevalence of Evershed-like and counter-Evershed-like flows in OPUs supports penumbral magnetoconvection as the underlying generic mode, with local loop geometry enabling siphon-like behavior in specific filaments (Jurcak et al., 2014, Durán et al., 10 Jul 2025).

Methodological limitations remain important. Kuckein et al. did not perform magnetic extrapolations and lacked continuous coverage on July 4; Jurčák, Bellot Rubio, and Sobotka used single-component, height-independent inversions and could not directly detect the small submergence velocity; Buehler et al. inferred geometric rise from optical-depth evolution and had spectropolarimetric cadences of $$B_\mathrm{vert} = f\,|B|\cos\gamma,\qquad B_\mathrm{hor} = \sqrt{f}\,|B|\sin\gamma,$$6 hours; Castellanos Durán et al. analyzed magnetic vectors in the line-of-sight frame without azimuth disambiguation or transformation to the local frame (Kuckein et al., 2011, Jurcak et al., 2014, Buehler et al., 2016, Durán et al., 10 Jul 2025). The resulting open problems are therefore structural rather than definitional: how often OPU fields rise rather than submerge, what role overlying chromospheric canopies play in trapping or releasing them, how their internal fine structure maps onto MHD penumbral models, and under what circumstances their chromospheric extension becomes a filament-scale flux rope.