Persistent Magnetic Cutting

• Persistent magnetic cutting is the sustained reconfiguration of magnetic connectivity and current pathways across solar flares, type-II superconductors, and computational magnetic topology.
• It encompasses phenomena such as tether-cutting reconnection in solar flares, steady vortex cutting in superconductors, and topological segmentation via persistent homology.
• Researchers integrate high-resolution observations, nonlinear force-free extrapolations, and MHD simulations to quantify flux exchanges, reconnection rates, and structural transitions.

Searching arXiv for recent and foundational papers relevant to persistent magnetic cutting across solar physics, superconductivity, and magnetic-field-line topology. Persistent magnetic cutting denotes a sustained or irreversible reconfiguration of magnetic connectivity and current pathways. In solar-flare physics, it refers to tether-cutting magnetic reconnection that severs the inner legs of sheared core fields near a polarity inversion line, forms newly connected low-lying loops, and leaves a persistently more horizontal photospheric field together with a downward-relocated coronal current system (Liu et al., 2011, Liu et al., 2013). In filament-eruption studies, the term has also been used for numerous, spatially distributed, small-scale reconnection episodes that recur quasi-continuously over an extended period and cumulatively alter filament stability (Tan et al., 5 Sep 2025). In type-II superconductors, persistent cutting denotes the steady coexistence of flux-line cutting and flux transport just above the critical current, governed by distinct nonlinear effective resistivities (Clem et al., 2011, Clem, 2011). In magnetic-field-line analysis, “cutting” can further denote topological segmentation of Poincaré orbits by persistent homology into islands, chaotic layers, and invariant tori (Bohlsen et al., 2024).

1. Terminological scope and core physical idea

The phrase is not tied to a single formalism. In the solar literature, its central content is reconnection-driven reorganization of coronal connectivity and currents, usually localized near a PIL and coupled to a photospheric back reaction. In superconductivity, the emphasis shifts from coronal topology to vortex dynamics: nonparallel vortices intersect, cross-join, and reconnect while depinning-driven transport occurs simultaneously. In computational magnetic topology, the “cut” is not a physical reconnection event but a principled segmentation of orbit point clouds into topologically distinct classes.

Domain	Magnetic object	Persistent signature
Solar flares and eruptions	Sheared core fields, sigmoids, flux ropes, current channels	Lasting increase of photospheric horizontal field, footpoint exchange, current-system contraction, enduring loop/ribbon reconfiguration
Type-II superconductors	Vortex lines in the mixed state	Finite $E_\parallel$ and $E_\perp$ above $J_c$, simultaneous cutting and transport, steady voltage and magnetization response
Magnetic-field-line topology	Poincaré-map orbit point clouds	$H_1$ and $H_0$ persistence used to separate islands, chaotic layers, and invariant tori

A common misconception is that magnetic cutting necessarily refers to a single impulsive reconnection episode. The solar papers instead describe two persistent regimes: a flare-core regime in which reconnection over about 30 minutes leaves an irreversible reconfigured state (Liu et al., 2013), and a filament regime in which many small reconnections recur for roughly 1.5 hours and cumulatively affect stability (Tan et al., 5 Sep 2025). The superconducting papers likewise reject a strictly episodic view: above the coupled threshold, cutting and transport form a steady resistive state rather than isolated events (Clem et al., 2011).

2. Solar-flare formulation: irreversible restructuring in NOAA AR 11158

A quantitatively documented flare-core example is the 2011 February 13 M6.6 flare in NOAA AR 11158. The flare started at 17:28 UT, peaked at 17:38 UT, and ended at 17:47 UT. In a compact box $R$ centered on the main PIL at the center of the magnetic sigmoid, the mean horizontal photospheric field increased from $\langle B_h\rangle \approx 1160$ G at 17:24 UT to $\approx 1480$ G at 18:00 UT, a 320 G, $28\%$ increase in about 30 min, at a rate of $0.15\ \mathrm{G\ s^{-1}}$. The enhancement persisted permanently after the flare, in contrast to the much slower long-term evolution in $E_\perp$0, $E_\perp$1. Vector maps showed that the photospheric magnetic field in $E_\perp$2 became more horizontal after the flare (Liu et al., 2011).

The observational chain combined high-cadence, seeing-free HMI vector magnetograms with Hinode/SP-based nonlinear force-free field reconstructions and UV/HXR imaging. Hinode/SP scanned nearly the entire active region at 12:00–12:32 UT, 16:00–16:32 UT, 19:00–19:32 UT, and 21:30–22:02 UT. The NLFFF extrapolations used the weighted optimization method after preprocessing the photospheric boundary to better satisfy force-free conditions, on a $E_\perp$3 grid over $E_\perp$4, with surface flux balance of $E_\perp$5–$E_\perp$6. The method enforced $E_\perp$7 and $E_\perp$8, but specific force-free and divergence metrics were not reported (Liu et al., 2011).

The coronal-current reconstruction localized the strongest preflare horizontal current channel immediately above $E_\perp$9, with $J_c$0 at heights $J_c$1 Mm. Across the flare, the $J_c$2 contour in a vertical cross-section through $J_c$3 dropped from $J_c$4 Mm to $J_c$5 Mm, implying a $J_c$6 Mm downward collapse localized above $J_c$7. Altitude profiles showed that $J_c$8 increased below $J_c$9 Mm during the flare interval and decreased above, then changed little from 19:00 to 21:30 UT, indicating that the reconfigured current system persisted for hours. The inclination angle $H_1$0 dropped across the flare below $H_1$1 Mm, with a larger drop nearer the surface, again indicating a more horizontal low-altitude field (Liu et al., 2011).

The flare kernels and HXR sources tied the magnetic change to the central feet of the sigmoid. Four compact 1700 Å kernels FP1–FP4 appeared at onset and evolved into double J-shaped Ca II H ribbons. RHESSI 25–100 keV images showed early footpoint-like HXR sources cospatial with FP1 and FP2 before 17:32:30 UT, transitioning to ribbon-like HXR emission connecting FP1–FP4 later. The compact region $H_1$2 was sandwiched between the stronger FP2 and FP3 kernels and lay below the intense horizontal-current channel, while the initial strong UV/HXR sources occurred on the two sides of that channel rather than on the peak vertical current density $H_1$3. The vertical Lorentz-force change in $H_1$4,

$H_1$5

was estimated as $H_1$6 dynes, consistent with Hudson’s implosion hypothesis (Liu et al., 2011).

These measurements established a specific solar meaning of persistent magnetic cutting: the flare cut the “tethers” of sheared core fields, formed new, low-lying current-carrying loops, drove a net downward Lorentz-force impulse, and left a persistently more horizontal and reconfigured field (Liu et al., 2011).

3. Connectivity exchange, twist reduction, and eruptive amplification

The AR 11158 event was later recast in explicitly topological terms through preflare and postflare NLFFF comparisons at 17:22:12 UT and 17:58:12 UT. Before the flare, the inner footpoints FP2 and FP3 lay on opposite sides of the PIL and were co-spatial with local maxima of magnetic twist. Field lines traced from FP2 and FP3 formed two elongated, sheared flux bundles reaching toward FP1 and FP4, respectively; the FP3–FP4 bundle was the most twisted, up to $H_1$7 turns. Immediately after the flare, regions with $H_1$8 were greatly diminished, the mean surface twist in FP2 and FP3 decreased by $H_1$9, and the new FP2–FP3 field lines became shorter and less twisted, $H_0$0. Overarching post-flare loops rooted in the ribbons had $H_0$1, matching the post-flare arcade (Liu et al., 2013).

The key quantitative connectivity result was footpoint exchange. Of 66 post-flare field lines traced from FP3, 11 terminated within FP2 and carried $H_0$2 Mx, about $H_0$3 of the FP3-region flux. Conversely, 11 of 51 FP2 lines exchanged to FP3, again $H_0$4 Mx, about $H_0$5 of the FP2-region flux. Conservatively, about $H_0$6 of inner-footpoint flux underwent exchange during the flare. Treated as a $H_0$7 min reconnection interval, the mean flux-transfer rate was

$H_0$8

The reconnected lines were rooted in the edge regions of the inner footpoints that brightened first and had among the highest preflare twist, $H_0$9 (Liu et al., 2013).

This formulation sharpened the meaning of persistence. It was no longer only the persistence of an enhanced $R$0, but also the persistence of a new connectivity graph: short, low-lying FP2–FP3 loops across the PIL; erupting longer FP1–FP4 loops; diminished twist at the original footpoints; and a lasting photospheric horizontal-field enhancement in the region between FP2 and FP3 (Liu et al., 2013).

A related MHD simulation extended the tether-cutting sequence into a flux-rope-formation framework. In NOAA AR 12241, a non-force-free extrapolation initialized a low-lying, strongly sheared arcade enveloping the main PIL. The finite Lorentz force was largest near the PIL at low heights and drove the oppositely directed legs of the arcade toward one another, producing an X-type geometry and, in three dimensions, a hyperbolic flux tube with high squashing factor $R$1 ($R$2) and locally enhanced $R$3. Tether-cutting reconnection at the low-lying HFT produced newly formed post-flare arcade loops below and a magnetic flux rope above, with twist number exceeding unity and reaching $R$4 by $R$5 min. The rope then rose at an average speed of $R$6 through a region with decay index $R$7, while slip-running reconnection moved its eastern footpoint along high-$R$8 contours and extended the rope along the PIL (Prasad et al., 2023).

The simulation identified a two-way persistence mechanism: converging Lorentz-force inflow repeatedly drove nonparallel arcade legs together at the HFT, and the rising rope further thinned and strengthened the current layer beneath it, enhancing reconnection. A plausible implication is that solar persistent magnetic cutting can operate both as a localized connectivity exchange near a flare core and as a positive-feedback eruptive pathway when the forming rope enters a torus-unstable domain (Prasad et al., 2023).

4. Distributed persistent reconnection in filament eruptions

Solar Orbiter observations introduced a broader eruptive-filament usage of the term. In a failed medium-scale filament eruption observed on 5 April 2024, persistent magnetic cutting denoted a regime in which numerous, spatially distributed, small-scale reconnection episodes occurred quasi-continuously over an extended period preceding and during the eruption. The event was observed by the Extreme Ultraviolet Imager High Resolution Imager at 174 Å with $R$9 s cadence for four hours, 19:59–23:59 UT, at $\langle B_h\rangle \approx 1160$0–$\langle B_h\rangle \approx 1160$1, while PHI provided eight line-of-sight magnetograms at 30-min cadence between 20:00 and 23:30 UT. The target filament, W150S18, was $\langle B_h\rangle \approx 1160$2 Mm in length and exhibited a double-decker configuration with upper and lower filaments separated by $\langle B_h\rangle \approx 1160$3 Mm along the slice; the upper filament thickness was $\langle B_h\rangle \approx 1160$4–$\langle B_h\rangle \approx 1160$5 Mm (Tan et al., 5 Sep 2025).

The time–distance analysis yielded projected rise speeds of $\langle B_h\rangle \approx 1160$6 for the upper filament and $\langle B_h\rangle \approx 1160$7 for the lower filament; correcting for the $\langle B_h\rangle \approx 1160$8 separation with an assumed radial trajectory increased these to $\langle B_h\rangle \approx 1160$9 and $\approx 1480$0, respectively. Activity spanned roughly $\approx 1480$1 hours, with jets and brightenings before and during the eruption. The paper identified five distinct reconnection contexts: flux emergence and recurrent small jets along the filament; left-footpoint reconnection with nearby open or large-scale field; reconnection between the upper filament and wrap-around fields; reconnection between the lower filament, the upper structure, and the overlying closed field; and impact-driven reconnection at low altitude during fallback (Tan et al., 5 Sep 2025).

Flux emergence beneath or near the filament occurred at a rate of $\approx 1480$2 at 20:30 UT, implying cumulative emerged flux of $\approx 1480$3 Mx over $\approx 1480$4 hours. A bidirectional jet near the left footpoint evolved into a transient collimated coronal jet. As the upper filament lifted, bidirectional jets propagated along its threads and drained upper-filament material toward the footpoints, implicating reconnection with enveloping wrap-around fields. The lower filament rose about $\approx 1480$5 min later, showed pronounced twist and rotation, and developed an elongated strip-like brightening along its outer edge, indicating reconnection with the higher structure and later with the stronger overlying closed field on the right. A compact post-flare loop appeared above the left part of the lower filament, while a larger loop system became visible on the right and was gradually filled by evaporated plasma heated at the footpoints by the impact and drainage of falling filament material. The compact loop brightened more than $\approx 1480$6 min earlier than the right footpoint of the large loop (Tan et al., 5 Sep 2025).

This regime differs from the AR 11158 flare-core picture in scale and multiplicity. The reconnections were at sub-megammeter to few-megammeter scales, produced narrow jets, thin brightenings, and small loops rather than a dominant one-off current sheet, and occurred in quick succession and partly overlapped in time and space. The paper’s conceptual claim was that the filament’s fate emerged from the sum of cuts and reattachments rather than from any single catastrophic event. It explicitly linked persistent cutting to the cumulative modulation of mass loading, twist distribution, and magnetic connectivity in both the filament core and the strapping field (Tan et al., 5 Sep 2025).

A common misconception is that persistent cutting should invariably favor an ejective eruption. The failed-eruption case showed the opposite possibility: repeated reconnections drained mass, redirected flows, and reattached the rising core to strong overlying loops. The paper argued that when the external decay index $\approx 1480$7 remains below the torus threshold up to the achieved heights, or when reconnection repeatedly reattaches the rising core to strong overlying loops, the eruption remains confined (Tan et al., 5 Sep 2025).

5. Flux-line cutting in type-II superconductors

In type-II superconductors, persistent magnetic cutting is a steady mixed-state process involving both flux-line cutting and flux transport. Flux transport, or depinning, is the motion of intact vortices under the Lorentz force when the component of current density perpendicular to the local flux density exceeds the depinning threshold; it produces an electric field perpendicular to $\approx 1480$8, with $\approx 1480$9. Flux-line cutting is the intersection, cross-joining, and reconnection of locally nonparallel vortices when the component of current density parallel to $28\%$0 exceeds the cutting threshold; it produces $28\%$1 and can be triggered by a helical vortex expansion instability (Clem et al., 2011).

The central constitutive description is the extended elliptic critical-state model. In slab geometry, with $28\%$2 making angle $28\%$3 to $28\%$4, the critical current obeys an elliptical dependence

$28\%$5

and the resistive state just above $28\%$6 is described by

$28\%$7

The ratio $28\%$8 controls the partition of $28\%$9 between components parallel and perpendicular to $0.15\ \mathrm{G\ s^{-1}}$0, and for the experimental geometry the model predicts

$0.15\ \mathrm{G\ s^{-1}}$1

Within the elliptic models, when $0.15\ \mathrm{G\ s^{-1}}$2 is neither parallel nor perpendicular to $0.15\ \mathrm{G\ s^{-1}}$3, both cutting and transport turn on simultaneously at the coupled threshold $0.15\ \mathrm{G\ s^{-1}}$4. The papers explicitly describe an intimate relationship between flux cutting and depinning (Clem et al., 2011).

The experimental test used an epitaxially grown YBCO thin film of thickness $0.15\ \mathrm{G\ s^{-1}}$5 nm, patterned track width $0.15\ \mathrm{G\ s^{-1}}$6m, and voltage-tap separation $0.15\ \mathrm{G\ s^{-1}}$7m. At $0.15\ \mathrm{G\ s^{-1}}$8 K and $0.15\ \mathrm{G\ s^{-1}}$9 T, the angular dependence of $E_\perp$00 was well fitted by the elliptical model with $E_\perp$01, $E_\perp$02, and $E_\perp$03. At $E_\perp$04 K and $E_\perp$05 T, the fitted values were $E_\perp$06, $E_\perp$07, and $E_\perp$08. The measured $E_\perp$09 at $E_\perp$10 K and $E_\perp$11 T favored the extended elliptic critical-state model with $E_\perp$12; at $E_\perp$13 K and $E_\perp$14 T, none of the five base models fit the shape, although the EECSM captured the magnitude scale with small $E_\perp$15 values and an empirical modification improved the fit with $E_\perp$16 and $E_\perp$17 (Clem et al., 2011).

The cylindrical-wire theory gave a complementary steady-state picture. In a long cylinder at its critical current in a longitudinal applied field $E_\perp$18, the local basis is set by the pitch angle $E_\perp$19 of the magnetic induction, and the constitutive relations are

$E_\perp$20

A uniform axial electric field follows from symmetry and $E_\perp$21, yielding the linkage

$E_\perp$22

The exact balance between transport-induced growth of $E_\perp$23 and cutting-induced consumption of $E_\perp$24 is

$E_\perp$25

Steady state requires $E_\perp$26. In this regime, persistent magnetic cutting means sustained, cyclic reconnection and reconfiguration of vortices that continuously produce $E_\perp$27 and consume internal $E_\perp$28, allowing a finite, uniform axial electric field at or just above the critical current without secular growth of longitudinal flux (Clem, 2011).

Another misconception addressed by the superconducting literature is that cutting and transport are independent thresholds that can be treated separately. The helical vortex arc model instead showed that the same instability simultaneously initiates flux cutting and flux transport at $E_\perp$29, with limiting cases

$E_\perp$30

thereby linking the longitudinal and transverse critical currents (Clem, 2011).

6. Topological segmentation by persistent homology

A different, explicitly topological use of the phrase appears in magnetic-field-line analysis. Here the objects of study are field-line orbits on a Poincaré section $E_\perp$31, represented as finite orbit point clouds

$E_\perp$32

Using either Euclidean distance in geometric coordinates $E_\perp$33 or the cylindrical geodesic metric in straight field-line coordinates $E_\perp$34, one constructs the Vietoris–Rips complex $E_\perp$35 over increasing scales $E_\perp$36. Persistent homology then records births and deaths of homology classes through the filtration $E_\perp$37. In this setting, $E_\perp$38 captures loop-like holes formed by islands and global enclosure, while $E_\perp$39 encodes connectivity and nearest-neighbor spacing (Bohlsen et al., 2024).

The classification rules are quantitative. The enclosure of the last-to-die $E_\perp$40 feature is

$E_\perp$41

If $E_\perp$42, the orbit is classified as a magnetic island or island chain. In the toy tokamak, with scans along $E_\perp$43 and $E_\perp$44, $E_\perp$45 was either near zero for islands or $E_\perp$46 for non-islands, so $E_\perp$47 separated islands from invariant tori and chaotic layers. Large chaotic layers were then detected by the internal hole count

$E_\perp$48

with the rule that among non-islands, $E_\perp$49 implies a large chaotic layer, while the remaining cases are thin chaotic layers or invariant tori. The approximate island count was obtained from

$E_\perp$50

For islands, $E_\perp$51 at $E_\perp$52 estimated the number of major islands in the chain (Bohlsen et al., 2024).

The toy tokamak benchmark used $E_\perp$53, $E_\perp$54, section $E_\perp$55, and $E_\perp$56 intersections per orbit. In geometric coordinates, islands on the inboard side could deform into thin lines and lose associated $E_\perp$57 loops early in the filtration, so only $E_\perp$58 of $E_\perp$59 islands in a chaotic orbit were detected. In straight field-line coordinates, the same $E_\perp$60 cluster contained $E_\perp$61 classes, recovering the correct count, and a secondary $E_\perp$62-island chain also emerged. The global loop class had $E_\perp$63, approaching the known VR circle value $E_\perp$64 in the integrable limit (Bohlsen et al., 2024).

In this computational usage, persistent magnetic cutting is a segmentation protocol rather than a physical reconnection mechanism. The method is sufficiently general to apply to generic $E_\perp$65D Hamiltonian systems, but it also has a stated limitation: persistent homology alone cannot cleanly distinguish thin chaotic layers from invariant tori, so complementary diagnostics such as weighted Birkhoff averages are recommended for a universal classifier (Bohlsen et al., 2024).