Table Of ContentDraftversion February1,2017
PreprinttypesetusingLATEXstyleemulateapjv.5/2/11
FIRST DETECTION OF SIGN-REVERSED LINEAR POLARIZATION FROM THE FORBIDDEN [O I] 630.03
NM LINE
A.G. de Wijn
HighAltitudeObservatory,NationalCenter forAtmosphericResearch,P.O.Box3000, Boulder,CO80307, USA
H. Socas-Navarro N. Vitas
7 InstitutodeAstrof´ısicadeCanarias,AvdaV´ıaL´acteaS/N,LaLagunaE-38205,Tenerife,Spain
1 Draft versionFebruary 1, 2017
0
2 ABSTRACT
We report on the detection of linear polarization of the forbidden [Oi] 630.03 nm spectral line.
n
The observations were carried out in the broader context of the determination of the solar oxygen
a
J abundance,animportantprobleminastrophysicsthatstillremainsunresolved. Weobtainedspectro-
polarimetric data of the forbidden [Oi] line at 630.03 nm as well as other neighboring permitted
0
lines with the Solar Optical Telescope of the Hinode satellite. A novel averagingtechnique was used,
3
yielding very high signal-to-noise ratios in excess of 105. We confirm that the linear polarization
is sign-reversed compared to permitted lines as a result of the line being dominated by a magnetic
]
R dipole transition. Our observations open a new window for solar oxygen abundance studies, offering
S an alternative method to disentangle the Nii blend from the [Oi] line at 630.03 nm that has the
. advantage of simple LTE formation physics.
h
Subject headings: line: profiles — Sun: abundances — techniques: polarimetric
p
-
o
r 1. INTRODUCTION and contaminated by a Nii blend.
t Theoldparadigmhadanabundanceofoxygenaround
s Forbiddenlineshavebeenextensivelystudiedforsolar
a coronalmagnetometry,andsuccessfullyusedtoinferthe 800 ppm determined first by Lambert (1978) as an
[ update to the result of Lambert (1968), and sub-
orientation of the coronal magnetic field in the plane of
1 the sky aswell asthe longitudinalstrength(Judge et al. sequently refined by others (e.g., Anders & Grevesse
1989). The abundance of oxygen was determined by
v 2013). However, forbidden lines are exceedingly weak
fitting one-dimensional semi-empirical models to spec-
3 when observed on the disk of the sun and are not com-
tral features produced by oxygen atoms and molecules
9 monlyobserved,withonlyafewexceptions. Oneofthese
that contain oxygen atoms. The abundance of oxy-
7 is the [Oi] line at 630.03 nm. We present here the first
gen has been revised downward several times since
8 observations of linear polarization in this forbidden line
(e.g., Grevesse & Sauval 1998). Abundances under
0 and confirm that it exhibits the sign reversal compared
500 ppm were proposed by Allende Prieto et al. (2001)
. to permitted lines, which is expected by theory.
1 and Asplund et al. (2004) who accounted for the Ni
Oxygenisthethirdmostabundantelementintheuni-
0 blend, used updated atomic parameters, and a new gen-
verse after hydrogen and helium. It plays an important
7
erationof 3Dmodels intheir analyses. The implications
1 role in defining the structure of stellar interiors, both
of adopting the lower value are so far-reachingacrossall
: as an opacity source and as a donor of free electrons.
v The abundance of other elements that also play an im- fields of astrophysics that the ensuing controversy has
i beenreferredtointheliteratureasthesolaroxygencrisis
X portant role but do not have abundance indicators in
(Ayres et al. 2006). More than a decade later, the solar
the photospheric spectrum, such as Ne, is often mea-
r oxygenabundanceproblemstillawaitsasatisfactoryres-
a sured in solar coronal lines relative to oxygen. A pre-
olution. Several papers from various groups have been
cise knowledge of the oxygen abundance is crucial in,
published with disparate results (Asplund et al. 2009).
e.g., the construction of solar and stellar interior mod-
Recent work suggests that the systematic uncertainties
els that, in turn, are used for the development of stellar
in the traditional abundance determination analyses are
evolution theories and models and, ultimately, the dat-
much larger than previously thought, to the point that
ing of globular clusters and other astrophysical objects.
this much-debated discrepancy might be within the ex-
Unfortunately, our current understanding of the oxygen
pected empirical error (Socas-Navarro 2015). It is nec-
abundance in the solar system is neither complete nor
essary to search for alternative diagnostics that could
accurate. It is not possible to determine the oxygen
provide complementary information and break the cur-
abundance in meteorites because it is a highly volatile
rent impasse. Spectro-polarimetry is a novel observa-
element. The few spectral indicators oxygen produces
tionaltechnique in the context of abundance determina-
in the solar photosphere are all either affected by com-
tionsthatshowsmuchpromise(Socas-Navarro& Norton
plex formation physics, extremely weak forbidden tran-
2007; Centeno & Socas-Navarro2008).
sitions, and/or contaminated by blends. The 630.03 nm
line studied in this paper is both a forbidden transition 2. OBSERVATIONSANDREDUCTION
The SpectroPolarimeter (SP, Lites et al. 2013)
[email protected] on the Solar Optical Telescope (Tsuneta et al.
2 De Wijn, Socas-Navarro,and Vitas
Figure 1. Left to right: Continuum intensity, area masks, and line-of-sightflux density scaled between 0and 2000 Mxcm−2. The gray
shadingoftheareamaskscorrespondsfromlighttodarktoweaklymagnetizedpixels,pixelsinthemagneticnetwork,plage,orpores,the
sunspotpenumbra,andsunspotumbra,respectively. Whitepixelsdonotbelongtoanymask.
2008; Suematsu et al. 2008; Ichimoto et al. 2008; are filtered to first remove structures smaller than a di-
Shimizu et al. 2008) of the Hinode spacecraft amond shape of 5×5 pixels, then filtered to fill in holes
(Kosugi et al. 2007) was used in an innovative mode to smaller than that diamond shape. The penumbral area
observe the spectral region from 629.97 to 630.35 nm. maskis thencreatedwiththosepixels inthe strong-field
The instrument normally observes a 0.24-nm wide areamaskthatarepartofthefeaturesthatoverlapwith
region at 630.21 nm, but was programmed to observe the umbral area mask, but that are not in the umbral
thewiderregionsplitintworeadoutsinrapidsuccession areamask. Finally,thedatawasfilteredtoremovenoise
with 10 pixels of overlap in the spectral direction. The using a low-pass filter in the spectral domain.
observations lasted for 3 hours and 52 minutes starting The [Oi] line that is the subject of our study is so
just after 12:07 UTC on November 22, 2010. The slit weak that its linear polarization signal is only similar in
was kept stationary and initially positioned at approx. magnitude to the photon noise in the strongly magne-
′′ ′′
(5 ,350 ), slightly ahead of a sunspot that then moved tized environment of a sunspot. In principle, it is pos-
through the slit aperture due to solar rotation. An area sible to improve the signal-to-noise (SNR) ratio of the
of approx. 36′′ × 82′′ was rastered in this way. It is observations at the cost of sacrificing spatial resolution
shown in Fig. 1. The image shows substantial distortion by averaging over the field of view. However, this is not
that results from pointing drift that is corrected by the straightforwardinthe caseofpolarizationprofileswhose
correlation tracker under normal operating conditions. shape and sign depend on the geometry of the magnetic
In this case, however, the correlation tracker must be field. Linearpolarizationsignalsemergingfrommagnetic
reset after each pair of exposures in order to raster the regions that have a 90-degree azimuth separation in the
FOV because it tracks solar rotation as well as pointing frameoftheobserverareoppositeandcanceleachother.
drift. The magnetic field inferred from the Fei lines is likely
Thetworeadoutswerefirstmergedandthenprocessed to be slightly biased toward higher layers than those
with the standard calibration code that was modified to probed by the [Oi] line, but they are both formed in
acceptthewiderspectralregion. Thecalibrateddatawas the photosphere and the variation with height of the
then processed with the NCAR/HAO MERLIN Milne- field orientation is expected to be small. We thus use
Eddington inversion code1 to yield the flux density and the magnetic field geometry derived from the inversion
the orientation of photospheric magnetic field from the of the magnetically sensitive Fei lines, observed simul-
two Fei lines at 630.15 and 630.25 nm. taneously and cospatially, to rotate the Stokes Q and U
Maskswerecreatedtoseparateweakly-magnetizedar- profilesateachpixeltoacommonreferenceframebefore
eas from the sunspot umbra and penumbra, and from averaging. Differences in formation height would merely
areas with strong magnetic field outside of the sunspot. introduce a small amount of signal cancellation in our
Theweakly-magnetizedareamaskisdefinedasthosepix- procedure, making the profile detection harder but not
els that have a flux density below 400 Mx/cm2, while invalidating the results presented below.
the strong-field area mask contains those pixels that We can estimate the resulting SNR by the propaga-
have flux density above 600 Mx/cm2. The umbra of the tion of random error. The masks for the weakly mag-
sunspot is defined as the pixels in the strong-field area netized areas, strong fields, penumbra, and umbra con-
mask that have a continuum intensity less than half the tain 150140,36833,26376,and 5612 pixels, respectively.
meancontinuumintensityofthefieldofview. Themasks The SNR of a standard observation using both beams
ofthe polarizationanalysisanda 4.8 s exposure was ap-
prox.103 towardtheendof2010(Lites et al.2013). Our
1http://www2.hao.ucar.edu/csac/csac-spectral-line-inversions
Linear polarization from the [O I] 630.03 nm line 3
Figure 2. Average intensity and polarization profiles in the areas defined in Fig. 1. The gray shade background shows the spread of
profiles. Prominent lines are identified in the intensity plots. Some parts of the profiles have been enlarged to show details. Residual
polarizationreferstowhatisleftintheStokes-Uprofileaftertheprofileshavebeenrotatedtoareferenceframeinwhichallofthesignal
should be in the Stokes-Q profile. The arrowheads at the top and bottom of each panel indicate the area of overlap between the two
detector readouts.
observationsuse both beams but havea longerexposure is present in all areas defined in Fig. 1. The signal is
time of12.8s. We thusfind SNRs of6.3×105, 3.1×105, predictablyweakestinthe weakly-magnetizedareas,but
2.6×105,and1.2×105,forthe spectraaveragedoverall still visisble. While some of the lines are much stronger
pixels in the above masks. Polarimetry relies on a mea- (e.g., the Fei 630.15 and 630.25 nm lines), others are
surementofdifferences,andisconsequentlysubstantially of similar strength (e.g., the Scii 630.07nm line), which
lesssensitivetosystematicerrorsthanameasurementof rulesoutreasonsforthedifferencesuchasaspecificmag-
intensity. With such high SNR it is however reasonable netic field configuration.
to assume that systematic errors will dominate even if The pertinent difference between the [Oi] line and all
they are not readily apparent. other transitions in the spectra is that its electric dipole
componentisforbiddenbyquantumselectionrules. The
3. DISCUSSION
absorption is dominated by the much weaker magnetic
The observed spectral range shown in Fig. 2 contains dipole component. The observed polarization profile is
six prominent spectral lines, with a seventh just at the in agreement with the theory of generation of polar-
blue extreme of the spectral FOV. All six lines exhibit ized light that predicts that the linear polarization orig-
the same sign pattern in circular polarization (negative- inated by a magnetic dipole term has the opposite sign
positive from blue to red). In linear polarization, how- structure to that produced by the electric dipole term
ever, we observe a reversal of the polarization signal. (Landi Degl’Innocenti & Landolfi 2004, Sect. 6.8). The
All but the [Oi] line show the same pattern (positive- circular polarization, on the other hand, has the same
negative-positive), while that line exhibits just the op- sign in the magnetic and the electric dipole terms.
posite behavior (negative-positive-negative). In other We observe that the linear polarization signal of the
words, the linear polarization signal in this line has the forbidden [Oi] line dominates that of the permitted Nii
oppositesigncomparedtothefiveothersintheobserved line. This is in agreement with expectations as the [Oi]
spectral range. All of these spectral features have been line is stronger than the Nii line even when assum-
observedsimultaneouslyandhavegonethroughthesame ing a very low oxygen abundance (Allende Prieto et al.
calibration and data reduction procedures. The effect
4 De Wijn, Socas-Navarro,and Vitas
2001), at least outside the sunspot. Furthermore, oxygen line.
the effective Land´e factor for the [Oi] line is higher
than for the Nii line (1.25 and 0.51, respectively,
The National Center for Atmospheric Research is
Centeno & Socas-Navarro2008).
sponsored by the National Science Foundation. HSN
A precise understanding of the [Oi] line polarization
and NV gratefully acknowledge support from the Span-
will open new windows for the challenging diagnostics
ish Ministry of Economy and Competitivity through
of the solar oxygen abundance. The blended Nii line
project AYA2014-60476-P (Solar Magnetometry in the
is an electric dipole transition. Therefore, the [Oi] and
Era of Large Solar Telescopes). AdW is grate-
Nii lines exhibit similar behavior in Stokes I and V, but
ful to the Spanish MINECO for supporting a visit
opposite in Stokes Q and U. The intensity spectrum of
to the IAC through the 2011 Severo Ochoa Pro-
the blend will remainsimilar to a single line because the
gram SEV-2011-0187. The data reported in this
blend is unresolved,i.e., the line widths aregreaterthan
paper are available from the Hinode Data Archive
the separation of the components. However, the small
(http://hinode.nao.ac.jp/hsc_e/darts_e.shtml).
wavelengthshift betweenthe lines willgiveriseto anin-
tricate pattern of features in the polarized spectra that
can help resolve ambiguities because the polarized line REFERENCES
profiles have structure on smaller scales than the inten-
sityspectrum(see Centeno & Socas-Navarro2008). The AllendePrieto,C.,Lambert,D.L.,&Asplund,M.2001, ApJ,
differentsolarfeaturesintheFOVhavedifferentthermal 556,L63
andmagneticpropertiesthatwillaffectthe[Oi]andNii Anders,E.,&Grevesse,N.1989,Geochim.Cosmochim.Acta,53,
197
linesdifferently,andthusresultindifferent,distinctpat-
Asplund,M.,Grevesse,N.,Sauval,A.J.,AllendePrieto,C.,&
terns. Measurements of the polarization properties of
Kiselman,D.2004,A&A,417,751
thelinecanyieldapowerfulhandleonrelativestrengths Asplund,M.,Grevesse,N.,Sauval,A.J.,&Scott, P.2009,
that should allow us to determine accurate constraints ARA&A,47,481
on the relative abundances of O and Ni. However, the Ayres,T.R.,Plymate,C.,&Keller,C.U.2006,ApJS,165,618
Centeno, R.,&Socas-Navarro,H.2008, ApJ,682,L61
determinationoftheoxygenabundancewillrequirecare-
Grevesse,N.,&Sauval,A.J.1998,SpaceSci.Rev.,85,161
fulconsiderationofthemanyintricaciesoftheformation
Ichimoto,K.,Lites,B.,Elmore,D.,etal.2008, Sol.Phys.,249,
of this closely blended line, as well as attention to oxide 233
formation particularly in the umbra, even if the relative Judge,P.G.,Habbal,S.,&Landi,E.2013,Sol.Phys.,288,467
strengths of the components are well determined. Kosugi,T.,Matsuzaki,K.,Sakao, T.,etal.2007,Sol.Phys.,243,
3
4. CONCLUSION Lambert,D.L.1968,MNRAS,138,143
—.1978, MNRAS,182,249
Wehavepresentedthefirstdirectcomparisonofsimul- LandiDegl’Innocenti, E.,&Landolfi,M.,eds.2004, Astrophysics
taneously observed linear polarization in forbidden and andSpaceScienceLibrary,Vol.307,PolarizationinSpectral
Lines
permittedlinesformedinthesolarphotosphere. Ourob-
Lites,B.W.,Akin,D.L.,Card,G.,etal.2013,Sol.Phys.,283,
servationofthelinearpolarizationprofileofthe[Oi]line
579
hasbeenmade possibleduetoanovelaveragingmethod Shimizu,T.,Nagata, S.,Tsuneta,S.,etal.2008,Sol.Phys.,249,
of Stokes profiles that takes the signal polarity into ac- 221
count. The observed behavior is in agreement with the Socas-Navarro,H.2015, A&A,577,A25
Socas-Navarro,H.,&Norton,A.A.2007,ApJ,660,L153
theory of generation of polarized light, and, in the con-
Suematsu,Y.,Tsuneta,S.,Ichimoto, K.,etal.2008, Sol.Phys.,
text of the solar oxygen abundance, carries the promise
249,197
of a robust way to differentiate the [Oi] line at 630 nm Tsuneta,S.,Ichimoto, K.,Katsukawa, Y.,etal.2008,Sol.Phys.,
fromtheNiiblendthathistoricallyhascompromisedthe 249,167
desirable utility of the LTE formation of the forbidden