SUMER observations of the inverse Evershed effect in the transition region above a sunspot

Aims. We analyse SUMER spectral scans of a large sunspot within active region NOAA 10923, obtained on 14-15 November 2006, to determine the morphology and dynamics of the sunspot atmosphere at different heights/temperatures. Methods: The data analysed here consist of spectroheliograms in the continuum around 142.0 nm and in the Si iv 140.2 nm, O iii 70.3 nm, N iv 76.5 nm, and O iv 79.0 nm spectral lines. Gaussian-fitting of the observed profiles provides line-of-sight velocity and Doppler-width maps. Results: The data show an asymmetric downflow pattern compatible with the presence of the inverse Evershed flow in a region within roughly twice the penumbral radius at transition-region temperatures up to 0.18 MK. The motions, highly inhomogeneous on small scales, seem to occur in a collar of radially directed filamentary structures, with an average width less than the 1 Mm spatial resolution of SUMER and characterised by different plasma speeds. Assuming that the flows are directed along the field lines, we deduce that such field lines are inclined by 10 deg to 25 deg with respect to the solar surface.

💡 Research Summary

This paper presents a detailed investigation of the inverse Evershed flow in the solar transition region (TR) above a large sunspot belonging to active region NOAA 10923, using spectroheliograms obtained with the SUMER instrument on board SOHO on 14–15 November 2006. The authors selected a set of spectral lines that sample a range of TR temperatures: the Si IV 140.2 nm line (≈8×10⁴ K), O III 70.3 nm (≈5×10⁴ K), N IV 76.5 nm (≈1.2×10⁵ K), and O IV 79.0 nm (≈1.4×10⁵ K), together with a continuum window around 142 nm used for context imaging. For each raster position, Gaussian profiles were fitted to the observed line shapes, yielding line‑of‑sight (LOS) Doppler shifts (interpreted as plasma velocities) and Doppler widths (providing information on thermal and non‑thermal motions).

The velocity maps reveal a strikingly asymmetric pattern of downflows surrounding the sunspot. While the umbral core shows almost zero LOS velocity, a pronounced red‑shifted (downward) signal appears at radial distances of roughly 1.5–2 times the penumbral radius. The magnitude of the downflows ranges from 5 km s⁻¹ up to about 15 km s⁻¹, depending on the line and the exact location. This spatial distribution is consistent with the inverse Evershed effect, i.e., a flow that, unlike the classical photospheric Evershed outflow, moves inward toward the sunspot at higher atmospheric layers.

A key finding is that the downflows are not smooth, large‑scale motions but are highly fragmented. The authors identify a “collar” of radially oriented filamentary structures whose individual widths are below the 1 Mm spatial resolution of SUMER. These filaments are evident in all four TR lines, yet each line shows slightly different velocities, indicating that the same magnetic structures host plasma at different temperatures moving at different speeds. For example, Si IV typically shows average downflows of ~8 km s⁻¹, whereas O IV exhibits speeds up to ~12 km s⁻¹. This temperature‑dependent acceleration suggests that the flow is guided along magnetic field lines that become increasingly inclined as they rise into the TR.

Doppler‑width analysis further reveals that the filaments contain significant non‑thermal broadening, especially in the hotter O IV line, implying enhanced turbulence or unresolved multi‑component flows within the structures. The authors interpret the observed LOS velocities as the projection of flows aligned with magnetic field lines. Assuming a simple geometry where the flow follows a loop that rises from the photosphere, turns over near the penumbral edge, and then descends, they infer that the field lines are inclined by roughly 10°–25° relative to the solar surface. This modest inclination is compatible with previous magnetic extrapolations for sunspot penumbrae and supports the notion that the inverse Evershed flow in the TR is guided by low‑lying, nearly horizontal magnetic arches.

Importantly, the inverse Evershed signature extends to about twice the penumbral radius, indicating that the TR flow occupies a larger area than the classical chromospheric inverse Evershed flow, which is typically confined within the penumbra. This broader reach implies that the TR may act as a conduit for mass and energy exchange between the outer penumbral atmosphere and the overlying corona, potentially influencing coronal heating processes above sunspots.

In summary, the study provides the first high‑resolution spectroscopic evidence that the inverse Evershed flow in the transition region is highly structured, filamentary, and temperature‑dependent. The flow occurs in narrow, radially oriented channels that are inclined by 10°–25° to the solar surface, and it extends well beyond the visible penumbra. These observations place strong constraints on models of sunspot magnetic topology and on the mechanisms that drive mass transport from the chromosphere into the corona. Future work combining SUMER‑type observations with higher‑resolution instruments such as IRIS, together with three‑dimensional magnetohydrodynamic simulations, will be essential to unravel the detailed physics of these filamentary inverse Evershed channels and their role in the overall energy balance of active regions.