Coronal Null Points in Solar Corona

• Coronal null points are regions where the solar magnetic field vanishes, leading to distinct spine and fan separatrix structures.
• They serve as focal points for energy accumulation, enabling wave focusing, reconnection processes, and influencing flare and jet dynamics.
• Embedded in complex magnetic skeletons, null points regulate large-scale connectivity and are key to understanding plasma behavior in active regions.

Coronal null points are locations in the solar corona where the magnetic field vanishes, $\mathbf{B}(\mathbf{r}_0)=\mathbf{0}$. In three dimensions they are topological singularities whose local field-line geometry is organized into a spine and a fan separatrix surface, and in realistic coronal fields they are embedded in larger magnetic skeletons containing separators and quasi-separatrix layers. Because the local Alfvén speed also tends to zero at a null, coronal nulls are simultaneously singular points of magnetic connectivity and singular points for MHD wave propagation, which makes them central to reconnection, wave focusing, flare ribbons, jets, coronal bright points, coronal rain, and open–closed flux boundaries in the heliosphere (McLaughlin et al., 2010, 0901.0865).

1. Definition and local magnetic structure

A coronal magnetic null point is defined by the condition

$\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$

Near the null, the field is linearized as

$$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$

or, equivalently,

$$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$

with the local topology determined by the eigenvalues and eigenvectors of the Jacobian. Because $\nabla\cdot\mathbf{B}=0$, the eigenvalues satisfy $\lambda_1+\lambda_2+\lambda_3 \approx 0$. In the generic 3D case, one eigenvalue has sign opposite to the other two: its eigenvector defines the spine, while the remaining two span the fan plane, a separatrix surface that divides distinct connectivity domains (0901.0865, Sun et al., 2012).

This local structure is the basis of the standard fan–spine description. The fan is a two-dimensional separatrix surface of field lines that emanate from, or converge into, the null, and the spine is a pair of singular field lines passing through the null. Where fan surfaces intersect, a separator field line is formed. In 2D the corresponding local structure is an X-type null; in 3D the same singular behavior persists, but with spine and fan replacing the planar X-point language (Meszarosova et al., 2013, McLaughlin et al., 2010).

Field-line mapping is singular on both the fan and the spine. This singularity is quantified by the squashing factor $Q$: $Q$ is formally infinite on the spine and fan, while adjacent volumes with large but finite $Q$ define quasi-separatrix layers (QSLs). For the generic case of a non-radially-symmetric null, $Q$ decays most slowly away from the spine or fan in the direction in which $\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$0 increases most slowly; extended, elliptical high-$\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$1 halos around spine footpoints are therefore generic rather than exceptional (Pontin et al., 2016).

Analytic model fields make these anisotropies explicit. A standard linear potential null is

$\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$2

while a frequently used 3D null for wave studies is

$\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$3

The first makes clear that proper radial symmetry is special ($\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$4); the second distinguishes proper ($\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$5) and improper ($\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$6) nulls and underlies much of the modern wave literature (Pontin et al., 2016, Thurgood et al., 2012).

2. Occurrence, statistics, and large-scale magnetic skeletons

Potential-field extrapolations and statistical methods both indicate that coronal null points are common but not uniformly distributed with height. In the quiet Sun, direct null finding and Fourier-spectrum methods agree on a super-chromospheric column density above $\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$7 of approximately $\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$8 null points per square megameter of solar surface. Higher-resolution data produce many more nulls very near the photosphere, but do not significantly alter the coronal null count; the majority of nulls reside in the low atmosphere, while the coronal population is comparatively sparse (0901.0865).

A three-year survey using PFSS modeling predicted 582 potential coronal null points between Carrington Rotations 2098 and 2139, and those predicted locations were manually inspected in contrast-enhanced SDO/AIA 171 Å images at the east and west limb. The same survey reported a statistically significant difference between observed and predicted latitudinal distributions of null points by a Kolmogorov–Smirnov test, which indicates that null observability is itself a structured selection problem rather than a trivial reflection of model counts (Freed et al., 2014).

At larger scales, a single parasitic polarity can generate more than one coronal null and can also generate separator lines connecting those nulls. In analytical source-surface models of coronal-hole linkages, one parasitic polarity region can produce multiple null points in the corona and, more importantly, separator lines connecting them; the separatrix dome of one null and the separatrix curtain of another can join stably, and apparently disconnected coronal holes can be linked not by an open corridor of finite width but by a singular line coincident with the separatrix footprint of the parasitic polarity (Titov et al., 2010).

This larger-scale topological view suggests that coronal nulls should not be treated only as isolated singular points. They belong to magnetic skeletons that include domes, curtains, separators, and QSLs, and those skeletons regulate how active regions, coronal holes, and the slow-wind source region are magnetically connected. A plausible implication is that null-point physics is often inseparable from separator and QSL physics, especially when open and closed flux systems interact (Titov et al., 2010).

3. Observational signatures across coronal phenomena

Circular-ribbon flares provide some of the clearest direct signatures of coronal null topology. In a GOES B-class flare on 2015 May 20, a chromospheric circular ribbon corresponded to the fan footpoints, while compact kernels marked the inner and outer spine footpoints; RHESSI $\mathbf{B}(\mathbf{r}_0)=\mathbf{0}.$9–$$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$0 keV detected both the site of the 3D null point and the shape of the outer spine. The circular ribbon was formed by adjacent compact kernels of size $$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$1–$$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$2, and those kernels brightened sequentially in the clockwise direction (Romano et al., 2017). The morphology of elongated spine-associated ribbons in such events is consistent with the generic elongated high-$$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$3 halos predicted for asymmetric nulls (Pontin et al., 2016).

Coronal bright points (BPs) are another major observational class. In a topological survey of 10 BPs, 9 out of 10 were found above an opposite-polarity intrusion defining a coronal null point. All bright points in coronal holes and two out of three bright points in quiet-Sun regions resided in regions containing a magnetic null point. In these 9 cases the X-ray bright point occupied only a limited part of the projected fan-dome area, either fully inside the dome or expanding over a limited area below which typically a dominant flux concentration resided (Galsgaard et al., 2017). In a separate CBP study, the overlying corona exhibited a fan–spine configuration with a coronal null at a height of $$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$4 Mm, and the chromospheric response lagged the coronal heating episode by less than 3 min, which the authors interpreted as evidence that the heating may have occurred at coronal heights (Madjarska et al., 2020).

Direct imaging can also reveal nulls through loop kinematics. A saddle-like coronal structure observed above NOAA 2666 showed loops converging toward the center of the saddle in the vertical direction and diverging in the horizontal direction. Potential-field calculations placed a null point at a height of $$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$5 above the photosphere at the center of the saddle structure, and the event was interpreted as smooth coronal magnetic field reconnection without detectable heating in the reconnection region or magnetically connected areas (Filippov, 2018).

Null-point topologies also host thermodynamic phenomena. Observations of “raining null-point topologies” showed coronal rain within the legs of closed loops well under the fan surface and preferentially near separatrices of the topology: the spine lines, null point, and fan surface. At least 15 such topologies were identified, and the frequency of rain formation and the ease with which it is observed strongly suggest that this phenomenon is generally present in null-point topologies of this size scale (Mason et al., 2019).

4. Wave propagation, mode conversion, and null-point seismology

Wave behavior near nulls is dictated by the rapid spatial variation of the characteristic speeds. The review literature established that, in a $$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$6 plasma, fast magnetoacoustic waves are focused toward the null by a refraction effect, so that wave energy and current density accumulate close to the null, while Alfvén waves remain tied to the field lines on which they are generated and accumulate along separatrices in 2D or along the spine or fan plane in 3D (McLaughlin et al., 2010). Fully 3D $$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$7 simulations confirmed that an initially pure fast wave remains permanently decoupled from the Alfvén mode, both linearly and nonlinearly, for proper and improper 3D nulls; the pure fast mode generates a nonlinear field-aligned disturbance, but refraction still focuses all wave energy toward the null point (Thurgood et al., 2012).

Observational evidence now links these theoretical results to specific null topologies. In a 26 November 2005 solar radio event, two radio sources at 244 and 611 MHz were shown, via extrapolated coronal topology, to lie in the fan structure of a coronal magnetic null point. Wavelet analysis found tadpoles with periods $$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$8–$$\mathbf{B}(\mathbf{r}) \approx \mathbf{J}\,(\mathbf{r}-\mathbf{r}_0), \qquad \mathbf{J}=\nabla\mathbf{B}|_{\mathbf{r}_0},$$9, interpreted as fast magnetoacoustic wave trains propagating in the fan. For the two correlated sources, the inferred Alfvén speeds were about $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$0 and $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$1, and the longest-period tadpoles implied waveguide widths of about $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$2 and $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$3 Mm, consistent with the thickness of the extrapolated fan (Meszarosova et al., 2013).

Nonlinear wave interactions introduce additional regimes. In a 2D coronal null embedded in a stratified atmosphere, slow magneto-acoustic shock waves generated in the chromosphere propagated through the null and produced a train of secondary slow shocks escaping along field lines. The resulting jet-like acoustic fronts propagated at approximately $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$4, and the characteristic frequency of the shock train was about $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$5, which was interpreted as a direct consequence of the thermodynamic structure around the null (Santamaria et al., 2017). In a more realistic 3D stratified null-point atmosphere driven by p-mode-like oscillations, most of the vertical velocity transmitted through the lower equipartition layer maintaining acoustic character, a small fraction converted to fast waves, and a second mode-conversion episode around the null produced fast-wave focusing with oscillatory current-density signatures compatible with the second harmonic of the driver frequency (Yadav et al., 2023).

The simple focusing picture is modified when finite-amplitude fast waves steepen before they reach the null. In a 2D potential X-type null without guide field, the decrease in fast speed toward the null amplifies nonlinear deformation, so a fast wave can become subject to nonlinear dissipation at a distance from the null and not reach it. In that model the steepening distance obeyed the empirical scaling

$$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$6

showing that the location of current spikes and heating relative to the null depends sensitively on incident amplitude and width (Zhong et al., 6 May 2026). This suggests that nulls are not only wave focussing sites but also nonlinear filters.

5. Reconnection, eruptions, and plasma equilibrium near nulls

Null-point reconnection spans a wide range of dynamical regimes. In the quadrupolar active region NOAA 11158, a coronal null at $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$7 Mm was identified in NLFFF extrapolations above a highly sheared emerging bipole. During eruption, multiple pairs of flare ribbons brightened simultaneously, a coronal HXR source appeared near the inferred null, and the ejecta followed a trajectory $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$8 away from the local radial direction. The authors argued that the asymmetrical photospheric flux distribution caused the confining magnetic pressure to decrease much faster horizontally than upward, so the null-related topology guided the non-radial eruption and the observed inverted-Y structure outlined the fan–spine system (Sun et al., 2012).

Not all null-point reconnection is explosive. The NOAA 2666 saddle event showed inflows of $$B_i = M_{ij}(x_j-x'_j), \qquad M_{ij}=\frac{\partial B_i}{\partial x_j}\bigg|_{\mathbf{x}'},$$9–$\nabla\cdot\mathbf{B}=0$0 into the null region and outflows of order $\nabla\cdot\mathbf{B}=0$1–$\nabla\cdot\mathbf{B}=0$2 away from it, but no obvious heating manifestations in hot EUV channels or X-rays (Filippov, 2018). This demonstrates that coronal nulls can host low-heating, quasi-steady reconnection as well as flare-associated energy release.

Null-adjacent closed loops have distinctive equilibrium properties. HYDRAD simulations of a loop passing close to a null point, with a near-singular expansion factor, showed that a siphon flow establishes itself within 4 hours of simulation time, flowing from the smaller-area footpoint to the larger-area footpoint. The modeled signature was strong upflows on the order of $\nabla\cdot\mathbf{B}=0$3 from the footpoint rooted in the localized minority polarity, weak downflows on the order of a few $\nabla\cdot\mathbf{B}=0$4 from the fan-surface footpoint, and near stationary plasma near the null region; analogous Hinode/EIS observations showed Doppler shifts that correlated well in both direction and magnitude (Mason et al., 2022).

A persistent misconception concerns flux-rope equilibria with a null below the rope. In 2D line-current models of a simple bipolar background field, a hyperbolic flux-tube or figure-of-eight configuration containing a null point below the flux rope is unstable and cannot exist for a long time in the solar corona. Stable pre-eruptive equilibria imply separatrices in the chromospheric horizontal-field distribution, recognizable as “herring-bone structures” in chromospheric fibril patterns; the null-below-rope configuration is instead a dynamic stage of eruption rather than a long-lived equilibrium (Filippov, 2013).

6. Modelling assumptions, controversies, and open problems

The dominant modeling tension concerns how much of null topology is captured by current-free extrapolations. Potential fields are often effective at recovering large-scale topology and the locations of nulls, fans, separators, and separatrix domes, as shown by the quiet-Sun density study and by bright-point topology surveys (0901.0865, Galsgaard et al., 2017). However, in the sheared quadrupolar eruption of AR 11158 the high-lying coronal null was present in NLFFF extrapolations but not in the corresponding potential field, which produced only low-lying nulls in weak fields (Sun et al., 2012). A conservative reading is that large-scale null statistics and many fan–spine topologies are robust in potential models, while the detailed geometry, height, and energetics of nulls in current-carrying active regions can require non-potential modeling.

Wave models remain similarly idealized. Many foundational studies use $\nabla\cdot\mathbf{B}=0$5, 2D geometry, or ideal MHD; some omit thermal conduction, radiation, and physical resistivity entirely (Thurgood et al., 2012, Santamaria et al., 2017, Zhong et al., 6 May 2026). This is sufficient to isolate refraction, phase mixing, and mode conversion, but it cannot determine how much of the focused energy is actually dissipated, how oscillatory current sheets reconnect, or how finite-$\nabla\cdot\mathbf{B}=0$6 mode coupling modifies the picture close to the null. The p-mode study therefore treats enhanced intensity near the null as a conjectured consequence of finite resistivity rather than a directly modeled result (Yadav et al., 2023).

There are also unresolved ambiguities in the thermodynamic interpretation of null-point observations. In raining null-point topologies, thermal nonequilibrium on long closed loops and interchange reconnection at the null were both proposed as viable explanations; the observations did not uniquely discriminate between them, and the authors explicitly called for higher spatial resolution data and MHD simulations (Mason et al., 2019). Likewise, bright-point studies established that nulls are common but also found one clear counterexample without an overlying coronal null, implying that null-point reconnection is important but not universal (Galsgaard et al., 2017).

Current research directions are correspondingly technical. The literature calls for higher-resolution, wide-field vector magnetograms; time-dependent null tracking; NLFFF and magneto-hydrostatic extrapolations; realistic 3D MHD simulations with conduction, radiation, and partial ionization effects; and direct observational tests linking high-$\nabla\cdot\mathbf{B}=0$7 null topology to Doppler signatures, heating delays, coronal rain, and quasi-periodic pulsations (0901.0865, Madjarska et al., 2020, Mason et al., 2022). Taken together, these studies suggest that coronal null points are best understood not as isolated mathematical curiosities but as dynamically evolving nodes in a multi-scale magnetic skeleton that governs where the corona reconnects, oscillates, condenses, and heats.