105. Transient inverse-FIP effect observed during a solar flare

Author: Deborah Baker, Lidia van Driel-Gesztelyi, David Long (UCL-MSSL, UK) and David H. Brooks (George Mason University, USA).

<< previous nuggetnext nugget >>


The elemental composition of astrophysical plasma and its variations are crucial to our understanding of the physical conditions and processes occurring within the plasma. Elemental abundance variations of solar and stellar coronae have mainly been linked to the surface temperatures of stars [1] and recent EUV spectroscopic observations of the Sun demonstrate that magnetic activity also has a major role to play [e.g. 2-5].

Coronae of the Sun and solar-type stars are over-abundant in elements with low first ionization potential (FIP effect), while cooler M dwarfs, which have large starspots and produce giant flares, have coronae either depleted of such elements or enhanced in high-FIP elements (inverse FIP or IFIP effect). Plasma fractionation takes place in stellar chromospheres where low-FIP elements are ionized and high-FIP elements are neutral.

For the first time, Hinode/EIS observed the IFIP effect on the Sun in highly localized patches near large sunspots during flares [6]. Since then, IFIP plasma has been observed in only eight active regions. In this nugget, we feature the most striking Hinode/EIS observations showing the spatial distribution and temporal evolution of IFIP plasma in AR 11429 during the decay phase of an M-class flare. This work was published in [7].

Hinode/EIS Observations

Hinode/EIS observed AR 11429 when operating in an autonomous observing mode during a Major Flare Watch campaign on 6 March 2012. The M2.2 flare triggered a high cadence response study and Hinode/EIS rastered the active region from the time of the flare’s peak to the end of the decay phase (Figure 1).

Figure 1. GOES soft X-ray light curve during M2.2 class flare on 6 March 2012. Dashed red lines correspond to the Hinode/EIS raster times in Figure 2.

We have selected a series of observations that highlight the rapid and extreme evolution of plasma composition. Figure 2 is composed of Ar XIV intensity maps (top panel) and line ratio composition maps of high-FIP Ar XIV to low-FIP Ca XIV (bottom panel). When the flare is at peak intensity (12:38 UT), only FIP effect is evident throughout the active region. Minutes into the decay phase (12:47 UT and 12:56 UT), IFIP plasma appears at the footpoints of bright flare loops within the active region while the loop tops exhibit enhanced FIP composition. By the time the soft X-ray intensity has returned to preflare levels (13:23 UT), the plasma at the eastern footpoints has evolved to photospheric plasma (ratio ~1) whereas IFIP composition still persists at the western footpoints. At the same time, the FIP effect has weakened along the loops. Hinode/EIS observed these extremes in plasma evolution in less than one hour after the decay phase of the flare. Seven hours later there was no IFIP plasma detected within the active region.

Figure 2. Selected Hinode/EIS Ar XIV 194.4 Å intensity maps (top) and Ar XIV/Ca XIV ratio maps (bottom) at 12:38, 12:47, 12:56, and 13:23 UT on 6 March 2012. The color bar scale shows FIP effect as blue/green, photospheric composition as orange, and IFIP effect as yellow.

What is special about the location of the IFIP patches?

Distinct IFIP patches occurred at very particular locations within the unusually complex magnetic configuration of AR 11429. Figure 3 displays an SDO/HMI continuum image during the flare’s decay phase at 12:47 UT overplotted with contours of IFIP plasma (green) and flare ribbons (orange). The IFIP patches are located in the highly sheared emerging flux over coalescing umbrae that are crossed by flare ribbons as indicated by the intersection of the contours. When the active region first rotated onto the solar disk, it was a mature sunspot group containing several common penumbra. For several days, major flux emergence of highly sheared field took place with 2–3 major bipoles still in emergence at the time of the Hinode/EIS observations on 6 March.

Figure 3. SDO/HMI continuum at 12:47 UT overplotted with contours of flare ribbons (orange) and IFIP (green). Arrows indicate the coalescing umbrae.

During the evolution of the active region, flux approached and collided with a pre-existing spot, forcing the coalescence of the smaller flux fragments into a growing, strongly coherent umbra surrounded by a common penumbra. Such field represents different strands of highly sheared field – evidenced by the presence of magnetic tongues [8] – that are converging towards each other to form sunspots, and therefore meet below the photosphere/chromosphere in the location of the coalescing umbrae. This is highly suggestive of subsurface/sub-chromospheric magnetic reconnection. Such reconnection leads to increased fast-mode wave flux from below the region of plasma fractionation in the chromosphere.

Our interpretation of the observations is consistent with the ponderomotive fractionation model for the creation of IFIP plasma [9]. The model invokes the ponderomotive force exerted by Alfvén waves when they refract from the high density gradient in the chromosphere. This gives rise to ion–neutral separation in the chromospheres of the Sun and other stars. The direction of the ponderomotive force determines whether low FIP elements become enhanced or depleted in stellar coronae. Alfvén waves originating in the corona produce the FIP effect and waves of sub-chromospheric origin create the IFIP effect. Sunspots are preferential locations for upward traveling acoustic waves to mode convert as the plasma β = 1 layer occurs at lower heights within the photosphere. Therefore the increased wave flux generated by the subsurface reconnection at coalescing umbrae will in turn preferentially create IFIP plasma above the umbrae. The IFIP plasma is only observed when the flare ribbons cross the umbrae and the IFIP plasma is evaporated into the flare loops. The flare reveals the IFIP plasma but does not create it.


We have shown that IFIP plasma is observed for a short time during the decay phase of a moderate flare in very particular locations within the unusually complex magnetic configuration of AR 11429. These highly localized regions of IFIP plasma appear over coalescing umbrae crossed by flare ribbons. We argue that the highly unusual plasma composition was created by increased fast mode wave flux that was generated by subsurface reconnection of the coalescing umbrae. According to the Laming fractionation model, fast mode waves coming from below the fractionation region of the chromosphere means that the ponderomotive force is directed downward so that low-FIP elements are depleted from the chromospheric plasma. The plasma is then evaporated into the corona in the flaring loops where it is observed by Hinode/EIS. This has implications for understanding the coronal composition of M dwarfs. The spatially resolved observations on the Sun may provide clues to the processes on M-stars which have IFIP-dominated coronae all the time, not only during large flares.


  • [1] Wood & Linsky 2010 ApJ 717, 1279
  • [2] Baker et al. 2015, ApJ 802, 104
  • [3] Brooks et al. 2015, Nat Commun, 6, E5947
  • [4] Brooks et al. 2017 Nat Commun 8, 183
  • [5] Baker et al. 2018, ApJ 856, 71
  • [6] Doschek & Warren 2015, ApJL 808, 7
  • [7] Baker et al. 2019, ApJ 875, 35
  • [8] Luoni et al. 2011, SoPh, 270, 45
  • [9] Laming 2015 Living Reviews in Solar Physics 12, 2