arXiv:1501.07397v1 [astro

Draft version January 30, 2015
Preprint typeset using LATEX style emulateapj v. 08/22/09
´moulin3 , S. L. Yardley1 , L. van Driel-Gesztelyi1,3,4 , D. M. Long1 , L. M. Green
D. Baker , D. H. Brooks , P. De
arXiv:1501.07397v1 [astro-ph.SR] 29 Jan 2015
Draft version January 30, 2015
Solar coronal plasma composition is typically characterized by first ionization potential (FIP) bias.
Using spectra obtained by Hinode’s EUV Imaging Spectrometer (EIS) instrument, we present a series
of large-scale, spatially resolved composition maps of active region (AR) 11389. The composition
maps show how FIP bias evolves within the decaying AR from 2012 January 4–6. Globally, FIP bias
decreases throughout the AR. We analyzed areas of significant plasma composition changes within
the decaying AR and found that small-scale evolution in the photospheric magnetic field is closely
linked to the FIP bias evolution observed in the corona. During the AR’s decay phase, small bipoles
emerging within supergranular cells reconnect with the pre-existing AR field, creating a pathway
along which photospheric and coronal plasmas can mix. The mixing time scales are shorter than
those of plasma enrichment processes. Eruptive activity also results in shifting the FIP bias closer to
photospheric in the affected areas. Finally, the FIP bias still remains dominantly coronal only in a
part of the AR’s high-flux density core. We conclude that in the decay phase of an AR’s lifetime, the
FIP bias is becoming increasingly modulated by episodes of small-scale flux emergence, i.e. decreasing
the AR’s overall FIP bias. Our results show that magnetic field evolution plays an important role in
compositional changes during AR development, revealing a more complex relationship than expected
from previous well-known Skylab results showing that FIP bias increases almost linearly with age in
young ARs (Widing & Feldman, 2001, ApJ, 555, 426).
Subject headings: Sun:abundances—Sun:active region—solar wind
Knowledge of plasma composition is key to understanding the physical processes associated with how solar plasma is transported, heated, and accelerated in
the solar atmosphere. In order to characterize composition variations, the first ionization potential (FIP) bias
(FIPBias ) is used to define the ratio of the elemental
abundance in the solar atmosphere (ASA ) to the elemental abundance in the photosphere (AP h ) such that
FIPBias = ASA /AP h . Composition is a critical property
that distinguishes the slow from fast solar wind (SW;
Geiss et al. 1995; von Steiger 1998; Antiochos et al. 2011).
The fast, steady wind has stable photospheric composition i.e. FIP bias ∼1 (von Steiger 1998; Zurbuchen et al.
1998; von Steiger et al. 2001; Zurbuchen et al. 2002), and
the slow, non-steady wind has variable coronal composition characteristic of AR plasma or high FIP bias >3–4
(e.g. Zurbuchen & von Steiger 2006). Further results of
abundance variations in discrete coronal structures were
summarized in Table 1 of Baker et al. (2013) and in review papers, such as Feldman & Widing (2003).
Plasma confined in closed-loop structures at the time of
emergence from beneath the photosphere has been determined to have photospheric composition (Sheeley 1995,
1996; Widing 1997; Widing & Feldman 2001). However,
once emerged, the plasma composition appears to be
1 University College London, Mullard Space Science Laboratory,
Holmbury St Mary, Dorking, Surrey, RH5 6NT, UK
2 College of Science, George Mason University, 4400 University
Drive, Fairfax, VA 22030, U.S.A.
3 Observatoire de Paris, LESIA, CNRS, UPMC Univ. Paris 06,
Univ. Paris-Diderot, Meudon, France
4 Konkoly Observatory, Hungarian Academy of Sciences, Budapest, Hungary
modified within days. McKenzie & Feldman (1992) and
Saba & Strong (1993) analyzed more established ARs in
soft X-rays and found variations in Fe/Ne (low/high FIP
elements) abundance ratios ranging from 2–7. In EUV,
Widing & Feldman (1995) derived FIP bias values of 4.8,
5.4 and 5.9 for three ARs.
FIP bias levels in some older ARs have been measured
from 8–16 (Young & Mason 1997; Dwivedi et al. 1999;
Widing & Feldman 2001; Feldman & Widing 2003). Widing & Feldman (2001) presented a systematic study of
compositional evolution in ARs using the Skylab spectroheliograph observations. Four ARs were identified
from 1973/4, during the minimum of Solar Cycle 20,
which emerged on the visible side of the Sun and could
be followed for 3–7 days. Using Mg VI/Ne VI intensity ratios, Widing & Feldman (2001) demonstrated that
plasma trapped in newly emerged loops has photospheric
composition and from thereon the AR plasma becomes
enriched at an almost constant rate per day so that after 2–3 days coronal and slow SW abundances of 3–4
were reached. After 4–5 days, FIP bias had risen to
8–9. Based on their study, Widing & Feldman (2001)
concluded that there is a close relationship between the
duration of plasma confinement and the magnitude of the
FIP bias. Surprisingly, few studies have been conducted
on the evolution of FIP bias of ARs after emergence beyond that of Widing & Feldman (2001).
In this paper, we examine the evolution of FIP bias
in a decaying AR that is older than those studied in
Widing & Feldman (2001). AR 11389 was well observed
from 2012 January 4 to 6 by several instruments (Section 2). In Section 3 we first describe the photospheric
and coronal observations of AR 11389 and its nearby
Baker et al.
coronal hole (CH). This is followed in Section 4 by a detailed investigation of the large-scale changes in FIP bias
within the supergranular cell forming part of the AR’s
negative polarity as well as a consideration of the smallscale, localized changes in FIP bias in and around the
AR. We finish in Section 5 with a discussion of the implications of our findings for the long-held view that AR
FIP bias increases linearly with time before concluding
in Section 6.
The observations presented here were obtained during
2012 January 4–6 using the Hinode Extreme-ultraviolet
Imaging Spectrograph (EIS: Culhane et al. 2007),
the Solar Dynamics Observatory (SDO)/Atmospheric
Imaging Assembly (AIA: Lemen et al. 2012) and
SDO/Helioseismic and Magnetic Imager (HMI: Schou
et al. 2012; Scherrer et al. 2012) instruments. Figure 1a,b
show full-disk HMI and SDO/AIA 193 ˚
A context images
on the 4th at 08:00 UT. The CH is located close to solar central meridian (CM) with AR 11389 on its southwestern boundary at approximately 35000 south of the
solar equator. AR 11388 is west of the AR-CH complex.
Zoomed SDO/HMI, SDO/AIA 171 ˚
A and 193 ˚
A images
of AR 11389 overplotted with the EIS field of view (FOV)
are shown at the bottom of Figure 1.
Hinode/EIS observed AR 11389 using the slit scanning
mode with the 200 slit and 200 scan step size for 180 pointing positions to build up a raster with a large FOV of
36000 ×51200 on January 4 and 6. A smaller sparse raster
FOV of 19500 ×28000 was constructed using the same slit
but with a scan step size of 300 for 65 pointing positions
on January 5.
Figure 2 shows Fe xii 195.12 ˚
A intensity maps and S
x 264.223 ˚
A – Si x 258.375 ˚
A composition (or FIP bias)
maps, with and without SDO/HMI magnetogram contours of ±200 G overplotted. Standard SolarSoft EIS
procedures were used to correct the raw data for dark
current, cosmic rays, hot, warm, and dusty pixels and
to remove instrumental effects of orbital variation, CCD
detector offset, and slit tilt. Single Gaussian functions
were used to fit the unblended calibrated spectra for Fe
viii 185.213 ˚
A, Fe ix 188.497 ˚
A, Fe x 184.536˚
A, Fe xiii
202.044 A, Fe xiv 264.787 A, Fe xv 284.160 ˚
A, Fe xvi
262.984 ˚
A, S x 264.233 ˚
A, and Si x 258.375 ˚
A, while double Gaussian functions were used to separate the blended
Fe xi 188.216 ˚
A, Fe xii 195.119 ˚
A, and Fe xiii 203.826
A lines. All these lines were used in the construction of
the composition maps. A discussion of the method used
to make these maps may be found in Brooks & Warren
(2011) and Baker et al. (2013). One caveat with this
study is that the composition maps reduce the data in
an identical fashion for all features in the AR. We have
not, for example, isolated the loops from the background
which must be kept in mind when analyzing changes in
composition maps from January 4–6.
AR 11389 appeared at the east limb on 2011 December 28. The AR was comprised of two positive polarity
spots and dispersed following negative field for the first
few days of its solar disk transit. By the time it approached the CM, the southern spot had decayed. The
magnetic field configuration at CM is displayed in Figure 1c. Five M-class GOES X-ray flares were attributable
to the AR prior to January 2. Activity then decreased
to intermittent C-class flares.
The precise age of AR 11389 can not be determined
from direct observation of its emergence. However, based
on a number of factors, we can reasonably state that the
AR is significantly older than those selected by Widing &
Feldman (2001) in their analysis of FIP bias evolution in
young ARs. First, AR 11389 is visible in STEREO B at
the east limb on 2011 December 20, 15 days prior to the
first Hinode/EIS observation on January 4. It is visible
in STEREO A approximately one week earlier. Second,
during its on-disk transit, the level of flaring in the AR
had reduced to very low levels, suggesting its age was beyond that of the flux emergence phase. Typically, flare
activity peaks when an AR’s sunspot area is at its maximum development just before the decay phase (Waldmeier 1955; Choudhary et al. 2013; van Driel-Gesztelyi
& Green 2014). Third, the lifetimes of ARs are approximately proportional to their peak magnetic flux (Schrijver & Zwaan 2000). AR 11389 is a large AR with total
negative magnetic flux measured to be ∼1.3×1022 Mx at
CM. Its lifetime is likely to be weeks–months rather than
days–weeks (van Driel-Gesztelyi & Green 2014). All of
these points imply that AR 11389 is expected to be older
than two weeks but less than one month as it was not
present during the previous solar rotation.
We give a detailed account of FIP bias evolution within
AR 11389 in the following subsections. Based on the
composition maps, the AR has been divided into regions
R1–R5, indicated by the dashed contours in Figure 3.
The positions of the contours within the EIS FOV change
slightly from January 4–6 as the AR evolves so the polarity inversion line (PIL) was used as a fixed reference
for the placement of the contours on the 6th. The underlying magnetic field and SDO/AIA multi-wavelength
images are used to understand the changes in plasma
composition observed in the upper atmosphere from the
EIS raster at 09:40 UT on January 4 to 13:40 UT
on January 6. Frequent references are made to the
two movies included as supplemental on-line material:
FIP EIS and SDO multi
4.1. Composition Evolution: Supergranular Cell
The most striking change observed in the composition
maps of AR 11389 over two days is the decrease in FIP
bias in R1 within the negative field of the AR (Figure 3).
The probability distribution functions (PDFs) of FIP
bias pixel−1 within R1 for January 4th (black PDF) and
6th (red PDF) are shown in Figure 4, as are the PDFs
for all regions. In R1, the PDF of FIP bias pixel−1 has
shifted to lower levels, i.e. towards photospheric composition, from January 4–6 and the mean value has decreased
from 2.30 to 2.04.
As is evident in the SDO/HMI magnetograms in the
accompanying SDO multi, evolution of the
magnetic field within R1 is driven largely by small
scale flux emergence causing fragmentation and merging
around the supergranular cell’s periphery. We identified
the primary sites of flux emergence events for the twoday period using an automated magnetic flux detection
Active region FIP bias evolution
algorithm. Contours were fitted around the emerging
positive flux fragments employing a clumping identification method with a contour level of 30 G (3σ accuracy)
(DeForest et al. 2007). Only the positive flux is measured since the negative flux is difficult to separate from
the negative background flux. However, the magnetic
field is divergence free, therefore, the same amount of
negative and positive flux is expected to emerge.
Emerging positive magnetic fragments were deemed to
be an ‘event’ if the fragments had magnetic flux greater
than 1018 Mx (singular or collective fragments with a
separation of less than 2500 ), had a duration longer than
30 minutes, and were observed to interact and undergo
flux cancellation with the supergranular cell of AR 11389.
Eight events met this criteria. They are shown in a series
of SDO/HMI zoomed magnetograms in Figure 5. Starting/ending times and maximum magnetic flux for each
event are given in Table 1.
within dispersed positive field of the nearby AR 11388 located towards the west of AR 11389 (Figure 1a). Shorter
loops rooted nearby connect within AR 11389. Repeated
flux emergence and the subsequent field cancellation has
eroded the photospheric magnetic field and forced these
coronal loops to reconnect elsewhere by the 6th. The
SDO multi movie shows the forced reorganization of the coronal loops rooted along the inverted-Y
field and the periphery of the supergranular cell in five
We conclude that in R1, supergranular cell evolution,
characteristic of all decaying ARs, led to the observed
decrease in FIP bias. Small bipoles emerging in and
around the supergranular cell reconnected with the preexisting coronal field. As a consequence of the reconnection, the photospheric plasma contained within these
bipoles mixed with high-FIP bias plasma of the coronal
4.2. Composition Evolution: Emergence and Spot Decay
Table 1
Positive flux emergence events within R1 from 2012 January 4–6.
Start Time (UT)
End Time (UT)
Maximum Positive
Flux (Mx)
4–Jan 09:30
4–Jan 13:30
0.9 × 1019
4–Jan 10:30
4–Jan 14:30
0.4 × 1019
4–Jan 15:30
4–Jan 18:30
0.2 × 1019
4–Jan 18:00
5–Jan 05:30
1.2 × 1019
4–Jan 23:30
5–Jan 11:00
0.3 × 1019
4–Jan 22:30
5–Jan 03:30
2.0 × 1019
5–Jan 12:30
5–Jan 17:30
0.7 × 1019
6–Jan 08:00
6–Jan 10:30
1.2 × 1019
Total negative flux within the contour of R1 at the time
of the Hinode/EIS raster on the 4th is 2.1×1021 Mx.
Over two days, the eight flux emergence events added
6.9×1019 Mx of positive flux to R1. Hence, photospheric
flux emergence added ∼3% to the total flux around the
supergranular cell within R1 where we observed an increase in photospheric composition plasma within coronal loops. This represents ∼0.5% of total flux within
AR 11389.
We suggest that the evolution of the photospheric magnetic field led to reconnection with the overlying coronal field. This is particularly evident on the western
side where the negative magnetic field in the photosphere forms an inverted-Y on January 4 (see arrows
in Figure 5a). However, by the 6th, the right branch
of the inverted Y-shaped field is no longer intact (see
the evolution in the bottom panels of Figure 5 and in
the upper left panel of SDO multi The left
panel of Figure 6 highlights some long loops rooted on
the right side of the inverted Y-shaped negative field on
the 4th. These loops observed in the SDO/AIA 171 ˚
passband connect outside the Hinode/EIS FOV shown,
Changes in FIP bias occur in other localized regions
(R2–R4) within AR 11389 and along its north-eastern
edge (R5), adjacent to the coronal hole. The corresponding PDFs of FIP bias pixel−1 for R2–R5 are shown in the
top panels of Figure 4 and PDFs for subregions within
R3 are shown in the bottom panels. PDFs of all major
regions shift to lower levels of FIP bias i.e. towards photospheric composition and mean values pixel−1 decrease
by 0.21–0.31 as plasma evolved from the 4th to the 6th.
Paths of low-FIP bias in R2 on January 6 appear to
trace the new loops that formed as a consequence of
reconnection along the supergranular cell boundary described in Section 4.1 (Figure 3b). Furthermore, the
western-most low-FIP bias patch within R2 is cospatial
with the location of persistent and highly concentrated
activity in the dispersed positive field at Y∼-35000 to 38000 . Moving magnetic features (MMFs; Harvey & Harvey 1973) are evident here on the 4th, prior to repeated
flux emergence beginning later on the 4th and continuing
until the early hours of January 6. Zoomed SDO/HMI
magnetograms showing the evolution of the field in the
red boxes are displayed in the top panel of Figure 7.
Within region R4, small scale magnetic field evolution takes place near the eastern/southeastern side of
the large positive spot. Flux cancellation occurs along
the fragmented PIL concentrated in the areas designated
by the yellow arrows in the bottom panels in Figure 7.
MMFs feed the process of cancellation as the spot decays
throughout the period. The magnetic field evolution is
mostly due to the formation of a supergranular cell (well
developed on January 6, Figure 7), although the change
observed in FIP bias is not as prominent as in the supergranular cell in the AR’s negative field (Section 4.1).
4.3. Composition Evolution: Eruptions
FIP bias evolution along the AR’s northeastern boundary in R5 is quite stark in the composition maps. Over
two days, the spatial extent of the blue/gray area where
FIP bias ∼1 has significantly increased, especially towards the south and northwest (Figure 3). Once again,
flux emergence has occurred where we have observed an
increase in plasma of photospheric composition. Two
small bipoles, BP1 and BP2, have emerged in R5. Figure 6 and SDO multi at 03:00 UT on the
Baker et al.
5th show the locations of the bipoles and their overlying/surrounding arcade of loops.
BP1 is visible in HMI magnetograms a few hours prior
to the Hinode/EIS observation on January 4. Relevant
frames of SDO multi have been marked to
indicate the sequence of episodes described here. At
∼03:00 UT on the 5th, a confined eruption of BP1 begins. The erupting field subsequently reconnects with the
overlying arcade, supplying plasma that is likely to be
of photospheric composition and cooler than the plasma
contained in the pre-reconnection coronal loops. This is
evident by a darkening in part of the arcade and in the
area between BP1 and the northern footpoints of the arcade loops. The reconnected loops disappear from the
SDO/AIA 171 ˚
A images though their ‘shape’ remains.
By 13:00 UT, the system has relaxed and dense coronal
loops have reformed and reappear in SDO/AIA 171 ˚
images as before the confined eruption.
Similarly, BP2 goes through a confined eruption and
and subsequent reconnection with the surrounding arcade field on January 6. Reconnection is dominantly
between closed connectivities (the emerging bipoles and
arcade loops) and does not appear to involve the open
field of the nearby CH.
The impact of the confined eruptions of BP1 and BP2
on FIP bias evolution in R5 is noticeable in the composition maps on January 5 and 6 in the middle panels
of Figure 2 (see also red arrows in the first image of
FIP EIS Low-FIP bias plasma extends further south on the eastern side of the AR immediately
after the eruption of BP1 on the 5th and expands further south on the 6th after the eruption of BP2. On the
6th, low-FIP bias plasma also extends along the northern edge of the EIS FOV where reconnected arcade loops
are rooted on their northern side. Clearly, BP1 and BP2
reconnecting with the overlying arcade has provided low
FIP-bias material to the arcade loops observed in the
composition maps.
4.4. Composition Evolution: AR Core
PDFs of FIP bias in regions R1, R2, R4, and R5 show
clear shifts towards photospheric composition after two
days (Figure 4). However, a more complex case is present
within region R3 covering part of the AR core. Significant localized FIP bias decreases and increases are in
fact offsetting each other. We divided region R3 into
four subregions, 3a–3d, for further investigation.
Contours 3a-3b (red) are located at the footpoints of
one set of connecting sheared coronal loops and 3c-3d
(blue) are located at the footpoints of a different set of
less sheared loops (c.f. locations of the subregion contour pairs with the PIL in Figure 3). In the case of
3a-3b, FIP bias substantially decreases with mean values falling 0.48 and 0.45, respectively (Figure 4, bottom
left panels). Flux emergence occurs within 3a and within
3b the underlying magnetic flux decreases/disperses as a
supergranular cell is evolving in the positive field. This
is consistent with what has been observed in the main
regions R4 and R5.
In contrast, FIP bias increases at the footpoints of
the coronal loops connecting 3c-3d (see Figure 4, bottom right panels). Magnetic flux density in contours 3c3d either remains approximately unchanged (3d) or be-
comes more concentrated (3c). These regions, 3c-3d, are
the only locations within AR 11389 and its surroundings
where the FIP-bias enhancement process is not dominated by the faster emergence process.
Comparison of a series of Hinode/EIS composition
maps of AR 11389 from 2012 January 4–6 revealed that
plasma composition evolved to lower FIP bias within the
AR. In the major regions, R1, R2, R4, and R5, PDFs
of FIP bias pixel−1 shifted towards photospheric composition and mean values pixel−1 decreased by 0.21–0.31.
Within the AR core (R3) where there was also an observed decrease in FIP bias, substantial localized decreases of 0.48 and 0.45 (subregions 3a–3b) were offset
by increases of 0.29 and 0.19 (subregions 3c–3d) in mean
values pixel−1 . Changes in mean values within one of the
main regions, R2, and 2 of the subregions of R3 (3a, 3d)
were in excess of half of their full-width-half-maximum
Overall evolution of plasma composition in AR 11389
differed from that of four ARs in Skylab spectroheliograph observations reported by Widing & Feldman
(2001). In their sample of newly emerged ARs, FIP
bias increased from within hours of flux emergence to
reach coronal abundances after two days. The fastest increase in the FIP bias occured during the phase of rapid
sunspot growth, from 1–3 days after emergence. Widing
& Feldman (2001) concluded that the FIP-effect mechanism is more effective during the emergence phase. Unlike the young ARs observed by Skylab, AR 11389 was
a mature AR in its decay phase. As expected in decaying ARs, the evolution of underlying magnetic field was
dominated by the formation of supergranular cells and
small-scale emergence/cancellation/MMFs which we observed in major regions R1, R2, R4, and R5, and subregion 3a–3b, the same regions where FIP bias decreased
over two days.
In contrast, FIP bias increased in contours 3c–3d where
the magnetic field evolved differently compared to the
other areas in the AR. Magnetic flux density remained
concentrated at each loop footpoint rather than decreasing by dispersion and cancellation. The fact that both
FIP bias and the magnetic field evolved in opposite directions in these locations compared to regions 3a–3b,
also in the AR core, provides a reasonable degree of confidence that systematic errors are not significant. However, as was stated in Baker et al. (2013), Si, with FIP
of 8.2 eV, and S, with FIP of 10.4 eV, are close to the
boundary between high and low FIP elements on the
Sun at 10 eV. Lanzafame et al. (2002) showed that S
behaves like a high-FIP element in ARs but it may overfractionate in quiet Sun regions (Lanzafame et al. 2005;
Brooks et al. 2009). Thus the behavior of S may be sensitive to the region on the Sun being investigated. For a
discussion of caveats associated with the S x 264.223 ˚
– Si x 258.375 ˚
A and a test of the validity of the method
please see Baker et al. (2013) and Brooks et al. (2015).
Sheeley (1995); Widing (1997); Young & Mason (1997);
Widing & Feldman (2001) found photospheric plasma
composition in newly emerged loops, therefore, flux
emergence provides a potential reservoir of low-FIP bias
plasma to mix with the high-FIP bias plasma contained
within AR coronal loops. The key questions are how does
Active region FIP bias evolution
this material make its way into coronal loops of the AR
and on what time scales does the plasma mixing occur?
As was demonstrated in Baker et al. (2013), reconnection can provide the pathways for mixing of magnetic
field containing plasma of different compositions.
When magnetic flux emerges it displaces the preexisting field in the AR, a current sheet generally forms
at the boundary between the different magnetic domains,
and reconnection takes place forming new coronal magnetic connections (e.g. Guo et al. 2013; Tarr et al. 2014,
and references therein). This implies that small scale
flux emergence within supergranular cells and flux cancellation along their boundaries will involve reconnection.
The effect of MMFs is similar to that of small scale flux
emergence episodes. Both involve the interaction between new, small-scale loops and large, pre-existing AR
loops. In this way, photospheric plasma from the flux
emergence episodes can be injected into the pre-existing
loops of the AR and surrounding arcade containing highFIP bias plasma. In AR 11389, evolution of supergranular cells, especially in the dispersed negative field, leads
to the reorganization of the overlying field from January
4–6. SDO/AIA and SDO/HMI observations show that
dispersing/disappearing field forces loops to reconnect
within the AR, the surrounding arcade, and network.
Then, low-FIP bias plasma fills the reconnected arcade
The plasma composition of AR 11389 is likely to be
determined by the relative time scales of plasma mixing resulting from small-scale flux emergence and FIP
bias enrichment processes. Widing & Feldman (2001) observationally determined that such enrichment in newly
emerged ARs occurs over days. This is consistent with
the time scales of the ponderomotive force FIP effect
model of Laming (2004, 2009, 2012). The ponderomotive force creates local abundance variation in relatively
short time scales (∼103 s) at chromospheric heights. As
a result, observed coronal abundance anomalies require
either transport or diffusion processes in order to supply
low FIP ions to the corona. Typical timescales for these
processes are of order of days (Laming 2014).
Plasma mixing time-scales due to small-scale flux
emergence and subsequent cancellation episodes within
the evolving magnetic field of supergranular cells are
likely to be short compared to those of FIP bias enrichment and transport processes. These mixed polarity features are highly dynamic and evolve on a time scale of
few hours (Table 1). Within the supergranular cell of
the AR’s negative field, we observed numerous episodes
with lifetimes of 1–12 h and maximum flux emergence
per episode in the range ∼0.2–2×1019 Mx. Despite their
small spatial scales and short lifetimes, they contribute
a significant fraction to the total solar magnetic flux (see
Thornton & Parnell 2011, and references therein). For
BP1 and BP2, the process is similar but more efficient
as the eruptions forced more reconnection for longer time
In this paper, we provide clear evidence of FIP bias
evolution in a decaying AR near a CH from 2012 January
4–6 using SDO/AIA, SDO/HMI, and Hinode/EIS observations. A sequence of large FOV, high spatial resolution
S x–Si x EIS composition maps shows a global increase
in plasma of photospheric composition in AR 11389. We
analyzed in detail the locations of significant changes in
FIP bias and found that small-scale evolution of the underlying magnetic field is closely linked to the evolution
of plasma composition observed in the corona.
For all regions and subregions where FIP bias
decreased, we identified episodes of flux emergence/cancellation and moving magnetic features
(MMFs). We quantified the amount of flux added to the
region containing part of the well-formed supergranular
cell in the ARs negative field. What appeared to be
small-scale, intermittent flux emergence episodes in fact
added ∼3% of new flux to region R1 and ∼0.5% to the
entire AR over two days. This is small compared to the
large-scale flux emergence of 13%, which was added over
two days to decaying AR 11112 (Tarr et al. 2014).
It is widely accepted that active regions begin decaying
almost as they emerge into the photosphere. The decay
process spreads the concentrated magnetic flux over an
ever-increasing area due to supergranular buffeting. Supergranular flows advect the magnetic field to cell edges
creating ‘holes’ of relatively free-field regions. Most of the
flux disappears through small-scale emergence and subsequent cancellation (van Driel-Gesztelyi & Green 2014).
MMFs also contribute to the removal of magnetic field as
opposite polarities cancel (Harvey & Harvey 1973; Hagenaar & Shine 2005).
We observed the effect on plasma composition of smallscale flux emergence within supergranular cells and flux
cancellation along their cell boundaries in AR 11389.
High-FIP bias is conserved/amplified only in localized
areas of high magnetic flux density i.e. in the AR core.
In all other areas within the AR, FIP bias decreased.
During the normal course of the AR decay process, photospheric composition plasma is supplied to coronal loops
containing high-FIP bias plasma via reconnection of new
emerging loops and the pre-existing field on time scales
(hours) that are shorter than those of enrichment processes (days). This is in contrast to repetitive heating
episodes in loops where high-FIP bias plasma concentrated at loop footpoints (Baker et al. 2013) is transported into the corona by chromospheric evaporation.
Such a process is occurring in young ARs where strong
magnetic field provides substantial magnetic energy released in many episodes. Heating, which is a function of
magnetic flux density, is strongly decreasing with the dispersing fields in decaying ARs (van Driel-Gesztelyi et al.
2003; D´emoulin et al. 2003).
Hinode/EIS has provided a new opportunity to reexamine the evolution of plasma composition in ARs at
the highest temporal and spatial resolutions. We have
extended the work of Widing & Feldman (2001) based
on Skylab observations of very young ARs to the later,
decaying phase of a much older AR. In an aging AR, FIP
bias is modulated by small-scale flux emergence. This result has possible implications for the FIP bias levels observed in the solar wind. Slow wind FIP bias measured
in situ is a steady 3–4 (e.g. Geiss et al. 1995; Zurbuchen
& von Steiger 2006). Values in excess of or below 3–4 are
rare, suggesting that very high-FIP bias measured in ARs
(>5) does not find a pathway into the slow SW. Then,
only mature and not too dispersed ARs are expected to
contribute to the slow SW.
Baker et al.
Hinode is a Japanese mission developed and launched
by ISAS/JAXA, collaborating with NAOJ as a domestic
partner, NASA and STFC (UK) as international partners. Scientific operation of Hinode is by the Hinode science team organized at ISAS/JAXA. This team mainly
consists of scientists from institutes in the partner countries. Support for the post-launch operation is provided
by JAXA and NAOJ (Japan), STFC (U.K.), NASA,
ESA, and NSC (Norway). LvDG and DML acknowledge
the European Community FP7/2007-2013 programme
through the eHEROES Network (EU FP7 Space Science
Project No.284461). LvDG acknowledges the Hungarian government for grant OTKA K 109276. The work
of DHB was performed under contract with the Naval
Research Laboratory and was funded by the NASA Hinode program. DB thanks STFC for support under Consolidated Grant ST/H00260/1. We thank the referee for
his/her constructive comments, which have improved the
Antiochos, S. K., Miki´
c, Z., Titov, V. S., Lionello, R., & Linker,
J. A. 2011, ApJ, 731, 112
Baker, D., Brooks, D. H., D´
emoulin, P., et al. 2013, ApJ, 778, 69
Brooks, D. H., Ugarte-Urra, I., & Warren, H. P. 2015, Nature
Communications, 6, 5947
Brooks, D. H., & Warren, H. P. 2011, ApJ, 727, L13
Brooks, D. H., Warren, H. P., Williams, D. R., & Watanabe, T.
2009, ApJ, 705, 1522
Choudhary, D. P., Gosain, S., Gopalswamy, N., et al. 2013,
Advances in Space Research, 52, 1561
Culhane, J. L., Harra, L. K., James, A. M., et al. 2007,
Sol. Phys., 243, 19
DeForest, C. E., Hagenaar, H. J., Lamb, D. A., Parnell, C. E., &
Welsch, B. T. 2007, ApJ, 666, 576
emoulin, P., van Driel-Gesztelyi, L., Mandrini, C. H., Klimchuk,
J. A., & Harra, L. 2003, ApJ, 586, 592
Dwivedi, B. N., Curdt, W., & Wilhelm, K. 1999, ApJ, 517, 516
Feldman, U., & Widing, K. G. 2003, Space Sci. Rev., 107, 665
Geiss, J., Gloeckler, G., & von Steiger, R. 1995, Space Sci. Rev.,
72, 49
Guo, Y., D´
emoulin, P., Schmieder, B., et al. 2013, A&A, 555, A19
Hagenaar, H. J., & Shine, R. A. 2005, ApJ, 635, 659
Harvey, K., & Harvey, J. 1973, Sol. Phys., 28, 61
Laming, J. M. 2004, ApJ, 614, 1063
—. 2009, ApJ, 695, 954
—. 2012, ApJ, 744, 115
—. 2014, Living Reviews in Solar Physics, Submitted
Lanzafame, A. C., Brooks, D. H., & Lang, J. 2005, A&A, 432,
Lanzafame, A. C., Brooks, D. H., Lang, J., et al. 2002, A&A, 384,
Lemen, J. R., Title, A. M., Akin, D. J., et al. 2012, Sol. Phys.,
275, 17
McKenzie, D. L., & Feldman, U. 1992, ApJ, 389, 764
Morgan, H., & Druckm¨
uller, M. 2014, Sol. Phys., 289, 2945
Saba, J. L. R., & Strong, K. T. 1993, Advances in Space
Research, 13, 391
Scherrer, P. H., Schou, J., Bush, R. I., et al. 2012, Sol. Phys., 275,
Schou, J., Scherrer, P. H., Bush, R. I., et al. 2012, Sol. Phys., 275,
Schrijver, C. J., & Zwaan, C. 2000, Solar and Stellar Magnetic
Activity (Cambridge University Press 2000)
Sheeley, Jr., N. R. 1995, ApJ, 440, 884
—. 1996, ApJ, 469, 423
Tarr, L. A., Longcope, D. W., McKenzie, D. E., & Yoshimura, K.
2014, Sol. Phys., 289, 3331
Thornton, L. M., & Parnell, C. E. 2011, Sol. Phys., 269, 13
van Driel-Gesztelyi, L., D´
emoulin, P., Mandrini, C. H., Harra, L.,
& Klimchuk, J. A. 2003, ApJ, 586, 579
van Driel-Gesztelyi, L., & Green, L. M. 2014, Living Reviews in
Solar Physics, 11, 5
von Steiger, R. 1998, Space Sci. Rev., 85, 407
von Steiger, R., Zurbuchen, T. H., Geiss, J., et al. 2001,
Space Sci. Rev., 97, 123
Waldmeier, M. 1955, Ergebnisse und Probleme der
Sonnenforschung. (Academische Verlag Gesellschaft; 2.,
Erweiterte Auflage edition)
Widing, K. G. 1997, ApJ, 480, 400
Widing, K. G., & Feldman, U. 1995, ApJ, 442, 446
—. 2001, ApJ, 555, 426
Young, P. R., & Mason, H. E. 1997, Sol. Phys., 175, 523
Zurbuchen, T. H., Fisk, L. A., Gloeckler, G., & Schwadron, N. A.
1998, Space Sci. Rev., 85, 397
Zurbuchen, T. H., Fisk, L. A., Gloeckler, G., & von Steiger, R.
2002, Geophys. Res. Lett., 29, 1352
Zurbuchen, T. H., & von Steiger, R. 2006, in ESA Special
Publication, Vol. 617, SOHO-17. 10 Years of SOHO and Beyond
Active region FIP bias evolution
Figure 1. Context images of AR-CH complex on 2012 January 4 at 08:00 UT. (a) Full disk SDO/HMI line-of-sight magnetogram saturated
at ± 200 G and (b) SDO/AIA 193 ˚
A image. (c) Zoomed SDO/HMI, (d) SDO/AIA 171 ˚
A and (e) 193 ˚
A high-resolution images (processed
using the Multiscale Gaussian Normalization (MGN) technique of Morgan & Druckm¨
uller 2014). Black box indicates the Hinode/EIS field
of view which data are shown in Figure 2.
Baker et al.
Figure 2. (Top to bottom) Hinode/EIS Fe xii 195.12 ˚
A high-resolution intensity (processed using MGN technique of Morgan &
uller 2014), S x 264.223 ˚
A – Si x 258.375 ˚
A composition maps, without/with SDO/HMI magnetogram contours of ±200 G
(blue/black contours) for (left to right) 2012 January 4 at 09:40 UT, 2012 January 5 at 04:15 UT, 2012 January 6 at 13:40 UT.
Active region FIP bias evolution
Figure 3. Composition maps overplotted with contours designating major regions R1–R5, subregions 3a–3d, and the main polarity
inversion line (PIL; solid black lines) on 2012 Jan 4 at 09:40 UT and 2012 Jan 6 at 13:40 UT.
Figure 4. Probability distribution functions (PDFs) of FIP bias pixel−1 within major regions R1–R5 (top panel) and subregions 3a–3d
(bottom panel) for 2012 Jan 4 at 09:40 UT (black) and 2012 Jan 6 at 13:40 UT (red). (Contours of the regions are shown in Figure 3).
Baker et al.
Figure 5. SDO/HMI magnetograms of R1 contour showing the locations of flux emergence events. In panel (a) the yellow arrows indicate
the branches of the inverse-Y shaped negative field discussed in Section 4.1.
Figure 6. SDO/HMI magnetograms corresponding to the EIS FOV for rasters of AR 11389 on January 4 and 6 overplotted with regions
of SDO/AIA 171 ˚
A (yellow) and 193 ˚
A (bronze) intensity. On the 4th, some of the long loops on both sides of the inverted Y-shaped
negative field (171 ˚
A see yellow arrows in Figure 5) are connected to dispersed positive field of AR 11388 (located outside the field of
view to the west as seen in Figure 1a) while other loops connect with the positive polarity of AR 11389. Two days later, the right branch
of the inverted-Y shaped field no longer exists suggesting that loops originally rooted there were forced to reconnect elsewhere within the
AR-CH complex. Bipoles BP1 and BP2 are located on the northeast of AR 11389 and their coronal loops are shown with overplotted 193
A intensity. See both movies.
Active region FIP bias evolution
Figure 7. SDO/HMI zoomed images showing magnetic field evolution described in Section 4.2. Yellow arrows indicate locations of flux
cancellation along the fragmented PIL in R4. The HMI FOV and the sizes of the red/blue boxes do not correspond to the EIS FOV.