1: \documentclass[12pt, preprint]{aastex}
2:
3: %\usepackage{graphicx}
4: %encorporating comments from JTM, 22 Feb. 2008 and 13 March 2008
5: %also encorporating comments from HPW and KM
6: %also incorporating GAD changes on 31 March 2008
7: %also including JTM and HPW changes, 2 April 2008
8: %copied from ms_6.tex, 3 April 2008
9: %revised from ms.tex on 1-2 July 2008
10:
11: \begin{document}
12:
13: \title{Flows and Non-thermal Velocities in Solar Active Regions
14: Observed with the Extreme-ultraviolet Imaging Spectrometer on
15: {\it Hinode}: A Tracer of Active Region Sources of Heliospheric
16: Magnetic Fields?}
17:
18: \author{G. A. Doschek, H. P. Warren, J. T. Mariska}
19: \affil{Space Science Division, Naval Research Laboratory, Washington,
20: DC 20375, USA}
21: \email{george.doschek@nrl.navy.mil}
22:
23: \author{K. Muglach}
24: \affil{ARTEP, Inc., Ellicott City, MD}
25:
26: \author{J. L. Culhane}
27: \affil{Mullard Space Science Laboratory, University College London,
28: Holmbury St. Mary, Dorking, Surrey, RH5 6NT, UK}
29:
30: \author{H. Hara, T. Watanabe}
31: \affil{National Astronomical Observatory, Mitaka, Tokyo 181-8588}
32:
33: \begin{abstract}
34:
35: From Doppler velocity maps of active regions constructed from spectra
36: obtained by the Extreme-ultraviolet Imaging Spectrometer (EIS) on the
37: {\it Hinode} spacecraft we observe large areas of outflow (20-50 km
38: s$^{-1}$) that can persist for at least a day. These outflows occur
39: in areas of active regions that are faint in coronal spectral lines
40: formed at typical quiet Sun and active region temperatures. The
41: outflows are positively correlated with non-thermal velocities in
42: coronal plasmas. The bulk mass motions and non-thermal velocities are
43: derived from spectral line centroids and line widths, mostly from a
44: strong line of \ion{Fe}{12} at 195.12 \AA. The electron temperature
45: of the outflow regions estimated from an \ion{Fe}{13} to \ion{Fe}{12}
46: line intensity ratio is about $1.2-1.4\times10^6$ K. The electron
47: density of the outflow regions derived from a density sensitive
48: intensity ratio of \ion{Fe}{12} lines is rather low for an active
49: region. Most regions average around $7\times10^8$ cm$^{-3}$, but
50: there are variations on pixel spatial scales of about a factor of 4.
51: We discuss results in detail for two active regions observed by EIS.
52: Images of active regions in line intensity, line width, and line
53: centroid are obtained by rastering the regions. We also discuss data
54: from the active regions obtained from other orbiting spacecraft that
55: support the conclusions obtained from analysis of the EIS spectra.
56: The locations of the flows in the active regions with respect to the
57: longitudinal photospheric magnetic fields suggest that these regions
58: might be tracers of long loops and/or open magnetic fields that extend
59: into the heliosphere, and thus the flows could possibly contribute
60: significantly to the solar wind.
61:
62: \end{abstract}
63:
64: \keywords{Sun: solar atmosphere, extreme-ultraviolet}
65:
66: \section{INTRODUCTION}
67:
68: The launch of the Extreme-ultraviolet Imaging Spectrometer (EIS) on
69: the {\it Hinode} spacecraft on September 23, 2006 has given us the
70: opportunity to investigate the dynamics of coronal plasma, i.e, bulk
71: plasma flows and non-thermal motions, in more detail and with higher
72: spatial resolution than previously possible. In addition, the dynamics
73: can be associated with plasma diagnostic measurements of electron
74: temperature and density. EIS can therefore help address fundamental
75: unsolved problems in active region physics, such as heating in active
76: region loops (e.g., Klimchuk 2006; Warren \& Winebarger 2007), and the
77: associations of active regions with the solar wind and heliospheric
78: magnetic flux (e.g., Liewer, Neugebauer, \& Zurbuchen 2004; Schrijver
79: \& DeRosa 2003).
80:
81: Plasma motions might be particularly useful as tracers of either
82: active region heating or sources of solar wind and heliospheric flux.
83: For example, motions that are highly transient, spatially
84: well-localized and comprised of both upflows and downflows, might
85: signify active region loop heating and wave activity. In contrast,
86: motions which are uni-directional (either upflows or downflows),
87: lasting over several days, and that cover an extensive spatial region
88: might be expected to be associated with primarily long loops or open
89: field regions, and are possible indicators of sufficiently open
90: magnetic flux to extend into the heliosphere and thus the flows could
91: contribute to the solar wind. The association of flows with closed or
92: open field can be facilitated by the creation of flow maps of active
93: regions contructed by rastering the EIS spectrometer slit over an
94: active region, and then correlating the maps with the underlying
95: photospheric magnetic field.
96:
97: Obtaining rasters of active regions and the creation of Doppler
98: velocity flow maps was an early science goal of the EIS data analysis
99: program. Doschek et al. (2007) (and more recently Harra et al. 2008)
100: showed that in active regions there are extensive areas of low
101: intensity that are characterized by net bulk plasma outflows and
102: enhanced non-thermal motions. The appearance of these outflow regions
103: in areas of very low intensity is striking when compared with maps of
104: spectral line intensity, also constructed from the EIS rasters. In
105: this paper we expand the results of Doschek et al. (2007) and discuss
106: in detail EIS rasters of Doppler shifts and non-thermal motions for
107: two active regions. We estimate electron temperatures and densities
108: of the outflowing regions, discuss data from other {\it Hinode}
109: instruments and other satellites, and speculate on the types of
110: magnetic structures that contain the flows and turbulent motions. We
111: also estimate the mass flux of the flows into the corona. We will
112: show that the locations of the flows in the active regions and the
113: underlying morphology of the photospheric longitudinal magnetic flux
114: suggest a link with heliospheric open fields that originate in active
115: regions. The flows seem more associated with the solar wind and
116: heliospheric magnetic fields than with heating of active region loops.
117:
118: Experimentally, flows are determined from measurements of a spectral
119: line centroid, coupled with information on the rest wavelength of the
120: line. There are numerous observations of Doppler shifts in the
121: literature, the most recent being obtained from the Solar Ultraviolet
122: Measurements of Emitted Radiation (SUMER) spectrometer on the {\it
123: Solar and Heliospheric Observatory} (e.g., Teriaca et al. 1999a;
124: Warren, Mariska, \& Wilhelm 1997). Mass motions can be related to
125: heating in coronal flux tubes and the acceleration of the solar wind
126: (e.g., Hassler et al. 1999; Tu et al. 2005; McIntosh et al. 2006).
127: The solar wind work has centered mostly on Doppler shift measurements
128: in coronal holes. There are also numerous coronal flux tube dynamical
129: numerical simulation models in the literature (e.g., Hansteen 1993;
130: Warren \& Winebarger 2007; and Patsourakos \& Klimchuk 2006). These
131: models attempt to explain observed density and temperature
132: measurements in coronal loops and relate the results to coronal
133: heating mechanisms.
134:
135: The origin of excess optically thin spectral line widths beyond their
136: thermal Doppler widths is unclear. Since a rocket flight by Boland et
137: al. (1975) it has been known that spectral lines from ions present in
138: the solar transition region and corona have wider profiles than
139: expected from pure thermal Doppler broadening alone, based on the
140: ionization equilibrium temperatures at which they are formed. The
141: profiles of the majority of the lines are Gaussian or close to
142: Gaussian, and the excess broadening does not appear to be a strong
143: function of position on the solar disk. The bulk of our current
144: knowledge of these motions in non-flaring plasma has been obtained
145: from the S082-B slit spectrograph on {\it Skylab} (e.g., Mariska
146: 1992), the High Resolution Telescope Spectrograph (HRTS) rocket
147: stigmatic spectrograph (e.g., Bartoe 1982), and SUMER (e.g., Chae,
148: Sch\"uhle, \& Lemaire 1998a).
149:
150: The excess spectral line broadening is usually expressed as a
151: non-thermal random mass motion component of unknown origin. It has
152: been cited as an expected signature of waves in the corona (e.g.,
153: Mariska, Feldman, \& Doschek 1978) and magnetic reconnection (e.g.,
154: Parker 1988). It could also be due to different multiple plasma flow
155: speeds in structures smaller than the spatial resolution of the
156: spectrometer because the convolved Gaussian-like line profile would be
157: wider than the widths of the individual Gaussians composing the
158: overall profile. Most recently, McIntosh et al. (2008) argue from
159: {\it Hinode} high spatial resolution Solar Optical Telescope (SOT)
160: observations that Alfv\,{e}nic waves observed in small-scale
161: spicular-type structures are responsible for the excess broadening.
162:
163: Although there are numerous published observations of Doppler shifts
164: and non-thermal motions, there is little information, particularly for
165: the corona, on the detailed relationship of the motions to actual
166: structures, e.g., coronal loops. There is far less information on the
167: relationship of these dynamical parameters to other physical
168: parameters such as electron temperature and density. The apparent
169: close relationship between large line widths and bulk plasma flows
170: seen in the EIS spectra is one of the most interesting aspects of the
171: new observations.
172:
173: In Section 2 we present pertinent details of the EIS on {\it Hinode}.
174: In Section 3 we discuss the observations and data reduction. Results
175: are given in Section 4 and the implications of the results are
176: discussed in section 5.
177:
178: \section{THE EIS ON {\it HINODE}}
179:
180: The EIS is described in detail by Culhane et al. (2007) and Korendyke
181: et al. (2006). The instrument consists of a multi-layer telescope and
182: spectrometer. The telescope mirror and spectrometer grating are
183: divided into two halves, each of which is coated with different Mo/Si
184: multi-layers. This results in two observed extreme-ultraviolet narrow
185: wavebands: 170-210 \AA\ and 250-290 \AA. Light from the telescope is
186: focused onto the entrance aperture of the spectrometer. The grating
187: then diffracts and focuses the light onto two CCD detectors.
188:
189: The telescope mirror is articulated and different regions of the Sun
190: can be focused onto the spectrometer aperture by fine mirror motions.
191: The entrance aperture of the spectrometer can be one of four options:
192: a 1\arcsec\ slit, a 2\arcsec\ slit, a 40\arcsec\ slot, or a
193: 266\arcsec\ slot. The slits/slots are oriented in the solar
194: north/south direction. Their heights can be variable, with a maximum
195: height for most observations of 512\arcsec.
196:
197: There are several modes in which EIS can be operated. Images of solar
198: regions can be constructed by rastering a slit or slot west to east
199: across a given solar area with a set exposure time at each step. At
200: each raster position it is possible to read-out the entire CCD and
201: obtain a complete spectrum for each wavelength band. It is also
202: possible to select a small set of lines, falling in narrower spectral
203: windows, the choices of which depend on the objectives of the
204: observation.
205:
206: The spatial resolution of EIS along the slit is approximately
207: 2\arcsec\ (1\arcsec\ per pixel) and the spectral dispersion is quite
208: high, 0.0223 \AA\ per pixel. The instrumental full width at half
209: maximum measured in the laboratory prior to launch is 1.956 pixels.
210: However, we adopt an in-orbit instrumental width of 2.5 pixels, or
211: 0.056 \AA. This width was obtained by comparing the width of the EIS
212: \ion{Fe}{12} 193.51 \AA\ line observed above the limb with
213: \ion{Fe}{12} forbidden line observations made from the {\it Skylab}
214: S082-B spectrograph (Cheng, Doschek, \& Feldman 1979) and SUMER
215: (Doschek \& Feldman 2000). The in-orbit instrument width is still
216: being investigated, and may undergo small revisions in the future.
217:
218: \section{OBSERVATIONS AND DATA REDUCTION}
219:
220: Data from two active regions were examined for this work. These
221: regions are listed in Table 1, along with the start times of the EIS
222: observations, the solar locations, and the EIS exposure times. The
223: locations are given in arcseconds relative to Sun center. Positive
224: locations are west/north; negative ones are east/south. The
225: observations consist of rasters using the 1\arcsec\ slit stepped from
226: west to east in 1\arcsec\ increments for a total of 255 pointings.
227: The slit height is 256 pixels. Each spectrum at each location in the
228: raster contains 20 spectral lines, in 16 pixel wide spectral windows.
229: The locations in Table 1 refer to the centers of the rasters. Data
230: from other active regions have been qualitatively examined and support
231: the general conclusions from the two regions discussed herein.
232:
233: Most of the line width and position results discussed in this paper
234: were obtained from the 195.12 \AA\ \ion{Fe}{12} line. This is the
235: most intense line in the EIS spectrum in part because its wavelength
236: is near the maximum of the multi-layer efficiency. It is therefore
237: ideal for line profile measurements because good count rates are
238: available for statistically meaningful results. In addition, the
239: ratio of \ion{Fe}{12} lines (186.89+186.85)/195.12) is used to obtain
240: electron densities, and the ratio of an \ion{Fe}{13} 202.04 \AA\ line
241: to the \ion{Fe}{12} 195.12 \AA\ line is used to detect changes in
242: electron temperature. The atomic data were obtained from the CHIANTI
243: (version 5.2) atomic database (e.g., Landi et al. 2006). All lines
244: discussed in this paper are listed in Table 2 with their temperatures
245: of peak formation in ionization equilibrium (Mazzotta et al. 1998).
246:
247: Since EIS scans from west to east, the western edges of the active
248: regions were observed earlier than the eastern edges. However, all
249: spectral quantities for a given spectral line (e.g., the parameters of
250: a Gaussian fit) compared at any single position in a raster were
251: recorded simultaneously on a single CCD. Therefore, all spectral
252: quantities for a single line are perfectly co-aligned in position as
253: well as being recorded at the same time.
254:
255: Observationally we have found that the width of the 195 \AA\ line is
256: slightly larger than the widths of the other \ion{Fe}{12} lines for
257: reasons not yet completely understood. Part of the reason is a blend
258: with another relatively weak \ion{Fe}{12} line. We have quantified
259: the difference by carrying out a detailed comparison between the
260: \ion{Fe}{12} 193.51 \AA\ and 195 \AA\ line widths using data from
261: another active region, from which a width correction factor for the
262: 195 \AA\ line was determined. We have linearly subtracted this
263: correction, 2.605 m\AA, from the 195 \AA\ width in order to derive
264: non-thermal velocities from the 195 \AA\ line. Even without this
265: correction, we also found that we obtain the same qualitative
266: non-thermal line widths with the \ion{Fe}{13} line at 202 \AA\ as with
267: the 195 \AA\ line, as well as with the \ion{Fe}{12} line near 193 \AA.
268: We prefer the stronger 195 \AA\ line rather than the \ion{Fe}{12} 193
269: \AA\ line for line profile measurements because of its better counting
270: statistics. Subtracting the 2.605 m\AA\ correction, compared with not
271: subtracting the correction, reduces the \ion{Fe}{12} non-thermal
272: velocity by about 5 km s$^{-1}$ at around 30 km s$^{-1}$ and by about
273: 3 km s$^{-1}$ at about 70 km s$^{-1}$.
274:
275: We have reduced the data using the standard EIS software data
276: reduction package. The data were corrected for cosmic ray hits, hot
277: and warm pixels, detector bias, and dark current. Data numbers were
278: converted to intensities in ergs cm$^{-2}$ s$^{-1}$ sr$^{-1}$
279: \AA$^{-1}$. An additional effect that was corrected for is a
280: variation of line position over the {\it Hinode} orbit due to
281: temperature variations in the spectrometer caused mainly by a varying
282: exposure to infrared heating from Earth. This variation produces an
283: approximately sinusoidal variation of spectral line centroids. It is
284: removed by averaging centroid positions over a length of slit for
285: which the underlying solar region is mostly a quiet Sun region without
286: any obvious transient activity. Another correction that was applied
287: is a correction for a slight tilt of the slit on the CCDs, which
288: affects wavelengths along a north/south direction. These corrections
289: allow us to make Doppler maps of active regions in spectral lines for
290: comparisons with the non-thermal velocities, provided rest wavelengths
291: can be determined.
292:
293: The parameters discussed in this paper are total spectral line
294: intensities, spectral line widths, and spectral line wavelengths.
295: These parameters are determined from the reduced data using a standard
296: Gaussian fitting procedure for each spectral line.
297:
298: EIS does not have an absolute wavelength calibration source. The
299: wavelength scale for EIS was calibrated as described in Brown et
300: al. (2007). However, in order to make active region Doppler maps, a
301: choice must be made for each active region of a spectral line
302: wavelength that we decide represents zero Doppler velocity. For each
303: active region discussed herein, we have determined this wavelength by
304: averaging low intensity regions over large areas that are outside of
305: the active region. These defined rest wavelengths are within about
306: 0.002 \AA\ or 3 km s$^{-1}$ of each other.
307:
308: Previous work with SUMER (e.g., Brekke, Hassler, \& Wilhelm 1997;
309: Chae, Yun, \& Poland 1998b; Teriaca, Banerjee, \& Doyle 1999b)
310: indicates that the average quiet Sun Doppler velocity at the
311: temperature of \ion{Fe}{12} is close to zero, or at most a few
312: kilometers per second. There is some indication that a small velocity
313: for \ion{Fe}{12} would be a blueshift (Teriaca et al. 1999b), which
314: would make the flows we observe slightly larger than we report here.
315: But this needs confirmation.
316:
317: The non-thermal speed is obtained from the full width at half maximum
318: (FWHM, in m\AA) of the line profile from the expression,
319:
320: \begin{equation}
321: FWHM = 1.665\times10^3~\frac{\lambda}{c}~\sqrt{\frac{2kT}{M} +
322: V^2}~~~,
323: \end{equation}
324: where $\lambda$ is the wavelength (in \AA\ in this paper), $c$ is the
325: speed of light, $k$ is the Boltzmann constant, $T$ is the electron
326: temperature, and $M$ is the ion mass. Equation (1) assumes that the
327: instrumental width (56 m\AA) has been removed from the line profile
328: and that all broadening mechanisms are Gaussian. Furthermore, if
329: ionization equilibrium is assumed, there is the tacit assumption that
330: the electron and ion temperatures are equal. These assumptions are
331: highly likely valid in the low corona where the EIS observations are
332: made, because electron densities at these altitudes in the corona are
333: on the order of 10$^8$ to several times 10$^9$ cm$^{-3}$. At these
334: densities equilibration times between electrons and ions are very
335: short, and ionization and recombination processes are very rapid for
336: ions such as \ion{Fe}{12}.
337:
338: \section{RESULTS}
339:
340: \subsection{The August 23, 2007 Active Region}
341:
342: Figure 1 shows a summary of the results for the August 23, 2007 active
343: region. The upper left panel is an intensity plot for the
344: \ion{Fe}{12} 195.12 \AA\ line. The middle and lower left panels show
345: the line centroid shift (colorbar units are m\AA) and FWHM of the line
346: (colorbar units in km s$^{-1}$), respectively. The top right panel
347: shows a co-aligned {\it Hinode} SOT magnetogram of the area within the
348: boxed region.The middle and bottom right hand panels show the centroid
349: and FWHM respectively within the boxed region shown in the left hand
350: panels.
351:
352: The striking aspect of Figure 1 is that the Doppler shifts and
353: non-thermal line widths are not only well-correlated, but that the
354: largest values for these quantities occur in regions where the
355: \ion{Fe}{12} line intensity is very low. The \ion{Fe}{10} (see Table
356: 2) image does show activity in the area of the flows and non-thermal
357: motions, but it is not spatially correlated well with the \ion{Fe}{12}
358: features. For the region in Figure 1, the largest Doppler shift
359: speeds (outflows) are on the order of 15-20 km s$^{-1}$ and the
360: non-thermal speeds become as large as 55-60 km s$^{-1}$. The SOT
361: magnetogram in Figure 1 indicates that the outflow regions occur over
362: fields of one dominant magnetic polarity, which may indicate either
363: open field lines or long closed loops.
364:
365: The temperature and density distributions in the flow regions are
366: shown in Figure 2. The upper two panels show the electron temperature
367: and density as a function of the FWHM of the \ion{Fe}{12} line for the
368: region within the box in the upper right panel of Figure 1. The
369: temperature was determined from the \ion{Fe}{13} to \ion{Fe}{12} line
370: ratio assuming ionization equilibrium and an isothermal plasma. The
371: density was determined from the ratio of the \ion{Fe}{12} lines in
372: Table 2. Figure 1 shows that the FWHM and Doppler shifts are
373: well-correlated, so the data in the upper panels of Figure 2 that
374: correspond to the largest FWHMs also correspond to the largest outflow
375: speeds. There is a clump of data near a FWHM of 70 m\AA\ and a
376: scattering of points towards larger FWHM that does not either increase
377: or decrease with temperature. These data arise from the outflow
378: regions. The large outflow regions have a characteristic temperature
379: of about $1.2\times10^6$ K and a density of about $7\times10^8$
380: cm$^{-3}$, just above quiet Sun values. Thus the large outflow
381: regions are more representative of quiet Sun regions than active
382: regions.
383:
384: The bottom two panels of Figure 2 show histograms of the Doppler shift
385: and the FWHM. Both exhibit tails that show the correlation of FWHM
386: with Doppler speed. The data in Figure 2 that clump near 70 m\AA\ and
387: near zero Doppler speed are from the pixels surrounding the outflow
388: regions. The clear correlation of Doppler speed and non-thermal
389: velocity is shown in Figure 3. These results were obtained by
390: removing the instrumental width from the line profiles and assuming an
391: electron temperature as defined for Figure 2.
392:
393: The August 23 active region was not bright enough to compare with good
394: statistics the \ion{Fe}{12} outflow speeds and FWHMs with other lines.
395: The strength of the activity was weak enough such that there is no
396: visible sunspot at the location of the active region. However, there
397: is a general positive correlation with similar data from the
398: \ion{Fe}{13} line in Table 2, and some, but reduced, correlation with
399: the \ion{Fe}{10} line. {\it Transition Region and Coronal Explorer}
400: ({\it TRACE}) images show considerable activity in the 171 \AA\ band
401: near the EIS \ion{Fe}{12} outflows, but the spatial correlation of
402: {\it TRACE} and EIS features is not particularly good.
403:
404: The spectral line profiles in the wide-line outflow regions look
405: fairly symmetric, with no obvious blueshifted separate component to an
406: otherwise stationary component, as seen in X-ray line profiles at the
407: onset of many flares. This result is similar to that found by Doschek
408: et al. (2007) for an active region observed in December, 2006.
409:
410: We note that there was considerable activity in the bright part of the
411: August 23 region throughout the time of the EIS raster. Observations
412: with {\it TRACE} and the Extreme UltraViolet Imagers (EUVI) on the two
413: {\it Solar Terrestrial Relations Observatory} ({\it STEREO})
414: spacecraft show that the obvious large loop system in the shape of a
415: $u$ in the \ion{Fe}{12} intensity image near the center of the image
416: did not exist at the start time of the EIS raster. It is the result
417: of a filament eruption that occurred with the ends of the filament
418: apparently anchored east and west of the outflow regions. The outflow
419: regions do not appear to be involved in any of the activity to the
420: east in the bright part of the active region.
421:
422: The August 23 region was also observed on August 22, twice during
423: August 23, and once on August 24. The August 22/23 observations cover
424: a 24 hour period and the August 24 observation extends the coverage by
425: about 12 more hours. Outflows in the same general area are seen as in
426: the observation under discussion. Thus the outflows appear to be
427: persistent with lifetimes on the order of at least 1.5 days. This
428: result was also found by Del Zanna (2008), who in addition confirmed
429: the conclusions in Doschek et al. (2007).
430:
431: \subsection{The December 11, 2007 Active Region}
432:
433: Figure 4 shows results for the December 11 active region presented in
434: the same format as in Figure 1. Again, there is a clear separation of
435: the largest outflow regions from the most intense regions in the
436: \ion{Fe}{12} line, and there is a clear correlation between outflows
437: and spectral line widths. And again, the large outflow regions are
438: located in an area of primarily a single polarity. However, the
439: December 11 active region is stronger than the August 23 region and
440: contains a prominent sunspot group.
441:
442: The SOT G band, Ca H, and magnetogram images of the active region are
443: shown in Figure 5, co-aligned with the EIS FWHM image. It is clear
444: that the large sunspot in the G band image lies south of the large
445: FWHM region. The circular region surrounded by large FWHM regions in
446: the upper left Figure 5 panel is not delineating the sunspot, but
447: rather is close to a region of weak magnetic field, as seen by
448: comparing the magnetogram with the FWHM image. This circular region
449: is also a region of weak Ca H emission. The alignment of the region
450: in the SOT data with the EIS data is not precise, but is within about
451: 2\arcsec.
452:
453: Similar results as shown for the August 23 region in Figure 2 are
454: shown for the December 11 region in Figure 6. These results
455: correspond to the data within the boxed area shown in the upper right
456: panel of Figure 4. As found for the August 23 region, pixels for
457: which large FWHM values are derived are characterized by quiet Sun
458: temperatures and relatively low densities. In contrast, the regions
459: immediately surrounding the large width areas have densities in places
460: that are on the order of $10^{10}$ cm$^{-3}$ and somewhat higher
461: electron temperatures.
462:
463: The lower two panels of Figure 6 show histograms of the Doppler shift
464: and FWHM for the \ion{Fe}{12} line. In this case there is clearly a
465: strong outflow component in the Doppler shift histogram, much stronger
466: than for the August 23 region. Although the color scheme in Figure 4
467: shows where the strongest outflows are coming from, this can be seen
468: more clearly from Figure 7. The upper left panel of Figure 7 shows
469: the Doppler shift, and the black and white images show where Doppler
470: shifts of different magnitudes come from. The white pixels in each
471: image represent those pixels for which the Doppler shift shown in
472: Figure 6 is equal to or less than the value of $Ds$ m\AA\ specified
473: for each image. Thus the data in the upper right panel of Figure 7
474: illustrate that all the outflows pixels with Doppler shifts less than
475: -8 m\AA\ occupy the white area shown. The position of this value of
476: $Ds$ can be seen in the Doppler shift histogram of Figure 6 and it
477: represents essentially all of the outflow regions. The part of the
478: Doppler shift histogram near zero shift in Figure 6 arises from all
479: the regions surrounding the dark region in the upper left panel of
480: Figure 7. As $Ds$ is progressively decreased, the white area is
481: reduced and reveals clearly the regions of strongest Doppler shift.
482: The shift $Ds$ can be converted to a speed $V$ in km s$^{-1}$ by, $V$
483: = $3\times10^{5}$$\times$$Ds$/195.120.
484:
485: The correlation between non-thermal speed and Doppler shift is given
486: in Figure 8. The non-thermal speeds reach values of about 90 km
487: s$^{-1}$ and the maximum Doppler outflow speeds are on the order of 45
488: km s$^{-1}$. Note that Doppler shifts as large as 140 km s$^{-1}$, as
489: reported by Sakao et al. (2007) for a different active region, are not
490: observed (However, see Harra et al. 2008).
491:
492: The December 11 active region is bright enough to examine the outflow
493: images in a number of different spectral lines (see Table 2), formed
494: at different electron temepratures. Results for the December 11
495: outflow region are shown in Figure 9. In Figure 9 there is a general
496: resemblance of all the outflow images, implying a correlation.
497: However, a close inspection shows that the easternmost (leftmost)
498: outflowing region has approximately the same location in all the
499: images, but the westernmost outflowing region shifts gradually more
500: westward as the temperature of line formation increases. This is
501: clearly seen by comparing the rightmost upper and lower panels in the
502: figure, showing the \ion{Fe}{12} and \ion{Fe}{15} images,
503: respectively. Note also that the fine structure varies among all the
504: images, implying small-scale multi-temperature outflowing regions.
505:
506: \ion{Fe}{12} line profiles in the outflow regions are symmetric in
507: some areas, particularly the westernmost regions, but in the
508: easternmost regions an asymmetric blue wing appears frequently on
509: profiles. This further supports the concept of multiple flow sites
510: with flows at different Doppler speeds.
511:
512: Inspection of EIS images for spectral lines other than the
513: \ion{Fe}{12} line formed at lower and higher temperatures, as well as
514: {\it STEREO}/EUVI movies, shows many long and not too bright loops in
515: the vicinity of the outflows. These loops generally have a westerly
516: component; some are even mostly parallel to the east/west direction.
517: Other loops have substantial southern and northern components.
518:
519: \section{DISCUSSION}
520:
521: From the above analysis of active region dynamics with EIS we may
522: conclude:
523:
524: 1. Extensive areas of active regions where the \ion{Fe}{12} emission
525: line intensity is weak exhibit higher Doppler shift outflows and
526: higher non-thermal velocities than found in other areas of the active
527: regions where the line intensity is much stronger. This does not rule
528: out large flows and line widths in high intensity areas of active
529: regions; we are simply implying that there will be regions of low
530: intensity in active regions that have significant Doppler shift and
531: line width signals. The outflows range from a few km s$^{-1}$ to as
532: much as 50 km s$^{-1}$. The non-thermal motions range from about 20
533: to 90 km s$^{-1}$.
534:
535: 2. The temperatures of the outflow regions obtained from the intensity
536: ratio of an \ion{Fe}{13} line to an \ion{Fe}{12} line are about
537: $1.4\times10^6$ K, and electron densities vary from about
538: $5\times10^8$ to about $10^{10}$ cm$^{-3}$, depending on the
539: particular region.
540:
541: 3. The outflow regions are concentrated primarily over or near
542: magnetic regions of a single polarity, and can last for at least
543: periods of a day and a half although there are variations within the
544: flow region.
545:
546: 4. There is some evidence for variations of temperature within flow
547: structures and multiplicity of flows, i.e., the excess widths of the
548: lines need not be due to turbulence even in cases where the line
549: profiles appear well-fit by single Gaussian profiles. It is possible
550: that a wide profile is composed of the line profiles of many
551: outflowing regions, each slightly Doppler shifted with respect to each
552: other due to slightly different Doppler shifts. This scenario might
553: produce a summed wide profile that could be fit with a single
554: Gaussian. A multi-component scenario definitely holds for the 11
555: December event. Some of the outflow profiles show an obvious
556: secondary outflowing component such that the overall profile departs
557: significantly from a Gaussian.
558:
559: 5. The \ion{Fe}{12} intensity in the outflow regions is much less than
560: in the bright active region loops. However, it is brighter than in a
561: disk coronal hole that was observed by EIS on September 28, 2007 near
562: 01:45 UT. In this coronal hole the \ion{Fe}{12} line is significantly
563: fainter than in the active region outflow regions.
564:
565: The faintness of the structures could be due to intermittency in the
566: outflows. That is, if the outflows are events with time scales
567: significantly less that the exposure times, they will appear faint
568: simply because they are bright for only part of the exposure time. For
569: example, two hypothetical regions, equally bright per second, would
570: appear different by a factor of, say, 5 in brightness if one region
571: only lasted 5 seconds while the other lasted for a longer exposure
572: time of 25 seconds. Intermittent events of this nature would have to
573: persist over time intervals of at least a day to explain the
574: persistence of the outflows.
575:
576: From inspection of context data from a number of spacecraft it appears
577: that the outflow regions are associated with long magnetic flux tubes,
578: and/or open field lines that extend into the heliosphere. It is
579: interesting to calculate the mass flux in the outflowing regions. For
580: the December 11 event the outflow mass flux for all outflows greater
581: than about 12 km s$^{-1}$ in the boxed region of Figure 4 is about
582: $1\times10^{-5}$ gm cm$^{-2}$ s$^{-1}$ and is about $1.6\times10^{-6}$
583: g cm$^{-2}$ s$^{-1}$ for speeds exceeding 31 km s$^{-1}$. For the
584: August 23 event the mass outflow flux in the boxed region of Figure 1
585: is about $5.7\times10^{-7}$ g cm$^{-2}$ s$^{-1}$ for speeds exceeding
586: 12 km s$^{-1}$. These quantities are obtained by multiplying the flow
587: speed by the mass density for each pixel in the boxed regions with a
588: outflow speed greater than the specified speed, and then summing the
589: result over all pixels.
590:
591: More investigation is needed to determine whether or not this mass
592: flux, or a fraction of it, contributes to the solar wind.
593: Alternatively, some of the mass flux might be confined within long
594: closed flux tubes. Liewer et al. (2004), Schrijver \& DeRosa (2003),
595: Luhmann et al. (2002), Neugebauer et al. (2002), and Wang \& Sheeley
596: (2002) have found from data and modeling that active regions can
597: contribute substantially to the heliospheric magnetic field and solar
598: wind. Liewer et al. (2004) find that the open field lines are from
599: the edges of active regions (see Figures 11 and 12 in their paper), in
600: locations similar to where we find the flows discussed in this paper.
601: Schrijver \& DeRosa (2003) assert that connections with the
602: heliospheric field to active regions are common, long-lived, and that
603: during solar maximum the contributions of active regions to the
604: interplanetary magnetic field can be as much as 50\%. As seen in
605: Liewer et al. (2004), Figure 14 in the Schrijver \& DeRosa paper shows
606: open field lines connected to the edge of a plage region. Also, a
607: comparison of their simulation results with the {\it TRACE} image in
608: their figure shows that the open field regions emanate from a region
609: that is largely dark in the {\it TRACE} image, qualitatively similar
610: to the locations of the EIS flows in the active regions we have
611: discussed. It is therefore quite tempting to associate at least some
612: of the EIS flows as sources of the solar wind confined to open field
613: lines that extend into the heliosphere.
614:
615: Some of the Doppler results of this paper are similar to those found
616: by Harra et al. (2008), who also analyzed flows in an active region
617: discussed previously by Sakao et al. (2007), and these authors suggest
618: that the flows might contribute to the solar wind. These authors
619: found the same types of flows as reported earlier by Doschek et al
620: (2007), but Doschek et al. (2007) did not make a connection with the
621: solar wind.
622:
623: In addition to the two active regions analyzed in detail in this
624: paper, we have qualitatively examined five other active regions that
625: exhibit similar behavior to the two regions we discussed in detail.
626: However, more studies with EIS will be needed along with simulations
627: of the solar magnetic field in order to test quantitatively the solar
628: wind connection. We are continuing to obtain high resolution rasters
629: of active regions and are refining our knowledge of the zero velocity
630: centroid wavelengths of lines in the EIS spectrum. In particular,
631: deeper exposures that reveal the flows in spectral lines formed over a
632: broad temperature range should be useful in connecting the flows back
633: to their loop footpoint origins. More work also needs to be done on
634: the line profiles in some of the flow regions, which exhibit obvious
635: indications of multiple loop flows.
636:
637: {\it Hinode} is a Japanese mission developed and launched by
638: ISAS/JAXA, collaborating with NAOJ as domestic partner, and NASA (USA)
639: and STFC (UK) as international partners. Scientific operation of the
640: {\it Hinode} mission is conducted by the {\it Hinode} science team
641: organized at ISAS/JAXA. This team mainly consists of scientists from
642: institutes in the partner countries. Support for the post-launch
643: operation is provided by JAXA and NAOJ, STFC, NASA, ESA (European
644: Space Agency), and NSC (Norway). We are grateful to the {\it Hinode}
645: team for all their efforts in the design, build, and operation of the
646: mission.
647:
648: The authors acknowledge support from the NASA {\it Hinode} program and
649: from ONR/NRL 6.1 basic research funds.
650:
651: \begin{references}
652:
653: \reference{bar1982} Bartoe, J.-D., Advances in Space Research, Vol. 2,
654: No. 4, 185
655:
656: \reference{bol1975} Boland, B. C. et al. 1975, \mnras, 171, 697
657:
658: \reference{bre1997} Brekke, P., Hassler, D. M., \& Wilhelm, K. 1997,
659: \solphys, 175, 349
660:
661: \reference{bro2007} Brown, C. M., Feldman, U., Korendyke, C. M., \&
662: Hara, H. 2007, \apjs, in press
663:
664: \reference{cha1998b} Chae, J., Sch\"uhle, U., \& Lemaire, P. 1998a,
665: \apj, 505, 957
666:
667: \reference{cha1998b} Chae, J., Yun, H. S., \& Poland, A. I. 1998b,
668: \apjs, 114, 151
669:
670: \reference{che1979} Cheng, C.-C., Doschek, G. A., \& Feldman, U. 1979, \apj, 227, 1037
671:
672: \reference{cul2007} Culhane, J. L. et al. 2007, \solphys, 243, 19
673:
674: \reference{del2008} Del Zanna, G. 2008, \aap, 481, L49
675:
676: \reference{dos2000} Doschek, G. A., \& Feldman, U. 2000, \apj, 529, 599
677:
678: \reference{dos2007} Doschek, G. A., et al. 2007, \apj, 667, L109
679:
680: \reference{han1993} Hansteen, V. 1993, \apj, 402, 741
681:
682: \reference{har2008} Harra, L. K., Sakao, T., Mandrini, C. H., Hara,
683: H., Imada, S., Young, P. R., van Driel-Gesztelyi, L., \& Baker,
684: D. 2008, \apj, 676, L147
685:
686: \reference{has1999} Hassler, D. M., Dammasch, I. E., Lemaire, P.,
687: Brekke, P., Curdt, W., Mason, H. E., Vial, J.-C., \& Wilhelm, K. 1999,
688: Science, 283, 810
689:
690: \reference{kli2006} Klimchuk, J. A. 2006, \solphys, 234, 41
691:
692: \reference{kor2006} Korendyke, C. M., et al. 2006, Applied Optics,
693: vol. 45, Issue 34, 8674
694:
695: \reference{lan2006} Landi, E., Del Zanna, G., Young, P. R., Dere,
696: K. P., Mason, H. E., \& Landini, M. 2006, \apjs, 162, 261
697:
698: \reference{lie2004} Liewer, P. C., Neugebauer, M., \& Zurbuchen,
699: T. 2004, \solphys, 223, 209
700:
701: \reference{luh2002} Luhmann, J. G., Li, Y., Arge, C. N., Gazis, P. R.,
702: \& Ulrich, R. 2002, \jgr, 107, No. A8, 1154
703:
704: \reference{mar1978} Mariska, J. T., Feldman, U., \& Doschek,
705: G. A. 1978, \apj, 226, 698
706:
707: \reference{mar1992} Mariska, J. T. 1992, in {\it The Solar Transition
708: Region}, Cambridge University Press, Cambridge, United Kingdom
709:
710: \reference{maz1998} Mazzotta, P., Mazzitelli, G., Colafrancesco, S., \&
711: Vittorio, N. 1998, \aaps, 133, 403
712:
713: \reference{mci2006} McIntosh, S. W., Davey, A. R., \& Hassler,
714: D. M. 2006, \apj, 644, L87
715:
716: \reference{mci2008} McIntosh, S. W., De Pontieu, B., \& Tarbell,
717: T. D. 2008, \apj, 673, L219
718:
719: \reference{neu2002} Neugebauer, M., Liewer, P. C., Smith, E. J.,
720: Skoug, R. M., \& Zurbuchen, T. H. 2002, \jgr, 107, No. A12, 1488
721:
722: \reference{par1988} Parker, E. N. 1988, \apj, 330, 474
723:
724: \reference{pat2006} Patsourakos, S., \& Klimchuk, J. A. 2006, \apj,
725: 647, 1452
726:
727: \reference{sak2007} Sakao, et al. 2007, Science, Vol. 318, 1585
728:
729: \reference{sch2003} Schrijver, C. J., \& DeRosa, M. L. 2003, \solphys,
730: 212, 165
731:
732: \reference{ter1999a} Teriaca, L., Doyle, J. G., Erdelyi, R., Sarro,
733: L. M. 1999a, \aap, 352, L99
734:
735: \reference{ter1999b} Teriaca, L., Banerjee, D., \& Doyle, J. G. 1999b,
736: \aap, 349, 636
737:
738: \reference{tu12005} Tu, C.-Y., Zhou, C., Marsch, E., Xia, L.-D., Zhao,
739: L., Wang, J.-X., \& Wilhelm, K. 2005, Science, 308, 519
740:
741: \reference{wan2002} Wang, Y.-M., \& Sheeley, N. R., Jr. 2002, \jgr,
742: 107, No. A10, 1302
743:
744: \reference{war1997} Warren, H. P., Mariska, J. T., \& Wilhelm,
745: K. 1997, \apj, 490, L187
746:
747: \reference{war2007} Warren, H. P., \& Winebarger, A. R. 2007, \apj, 666, 1245
748:
749: \end{references}
750:
751: \clearpage
752:
753: \begin{table}\label{table1}
754: \caption{Observed Active Regions in 2007}
755: \begin{center}
756: \begin{tabular}{llll}
757: \hline
758: \hline
759: Date & Time (UT) & Location & EIS Exposure Time (s)\\
760: \hline
761: 23~August & $01:55:43$ & $-518\arcsec,-211\arcsec$ & $60$\\
762: 11~December & $00:24:16$ & $-178\arcsec,-144\arcsec$ & $60$\\
763: \hline
764: \end{tabular}
765: \end{center}
766: \end{table}
767:
768: \begin{table}\label{table1}
769: \caption{Spectral Lines and Temperatures of Formation}
770: \begin{center}
771: \begin{tabular}{lll}
772: \hline
773: \hline
774: Ion & Wavelength (\AA) & Temperature (K)\\
775: \hline
776: \ion{Fe}{8} & $185.21$ & $6.3\times10^5$\\
777: \ion{Fe}{10} & $184.54$ & $1.0\times10^6$\\
778: \ion{Fe}{12} & $186.89+186.85$ & $1.6\times10^6$\\
779: \ion{Fe}{12} & $195.12$ & $1.6\times10^6$\\
780: \ion{Fe}{13} & $202.04$ & $1.6\times10^6$\\
781: \ion{Fe}{14} & $274.20$ & $1.8\times10^6$\\
782: \ion{Fe}{15} & $284.16$ & $2.0\times10^6$\\
783: \hline
784: \end{tabular}
785: \end{center}
786: \end{table}
787:
788: \clearpage
789:
790: \begin{figure*}[t!]
791: \centerline{
792: \includegraphics[clip,scale=0.7]{f1a.ps}
793: \includegraphics[clip,scale=0.7]{f1b.ps}
794: }
795: \caption{Left panels: Top - images of \ion{Fe}{12} 195.12 \AA\
796: intensity for the August 23 active region; middle - centroid shift
797: (blue is towards the observer); bottom - line width; right panels: top
798: - SOT magnetogram within the boxed region; middle - centroid shift of
799: region in the boxed area in the left panels; bottom - line width
800: within the boxed region. The ordinates are in the north-south
801: direction; the abscissae are in the east-west direction.}
802: \end{figure*}
803:
804: \clearpage
805:
806: \begin{figure}[t!]
807: \plotone{f2.ps}
808: \caption{Top panels: Electron density and temperature distributions
809: within the boxed area of Figure 1 as a function of FWHM of the Fe XII
810: 195 \AA\ line. The temperature is derived from the
811: \ion{Fe}{13}/\ion{Fe}{12} ratio. Bottom panels: histograms of
812: centroid shift (left) and FWHM (right) within the boxed area of Figure
813: 1. The centroid shift in m\AA\ is converted to Doppler speed (upper
814: axis). In the bottom right panel, 65 m\AA\ FWHM corresponds to 16 km
815: s$^{-1}$ non-thermal velocity and 92 m\AA\ corresponds to 61 km
816: s$^{-1}$.}
817: \end{figure}
818:
819: \clearpage
820:
821: \begin{figure}[t!]
822: \plotone{f3.ps}
823: \caption{Correlation of Doppler speed with non-thermal speed for the
824: data within the boxed area of Figure 1.}
825: \end{figure}
826:
827: \clearpage
828:
829: \begin{figure*}[t!]
830: \centerline{
831: \includegraphics[clip,scale=0.7]{f4a.ps}
832: \includegraphics[clip,scale=0.7]{f4b.ps}
833: }
834: \caption{Left panels: Top - images of \ion{Fe}{12} 195.12 \AA\
835: intensity for the December 11 active region; middle - centroid shift
836: (blue is towards the observer); bottom - line width; right panels: top
837: - SOT magnetogram within the boxed region; middle - centroid shift of
838: region in the boxed area in the left panels; bottom - line width
839: within the boxed region. The ordinates are in the north-south
840: direction; the abscissae are in the east-west direction.}
841: \end{figure*}
842:
843: \clearpage
844:
845: \begin{figure}[t!]
846: \plotone{f5.ps}
847: \caption{Upper left and lower right panels: the \ion{Fe}{12} FWHM and
848: SOT magnetogram shown in Figure 4, respectively. Upper right panel:
849: the SOT G-band image; lower left panel: the SOT Ca H image.}
850: \end{figure}
851:
852: \clearpage
853:
854: \begin{figure}[t!]
855: \plotone{f6.ps}
856: \caption{Top panels: Electron density and temperature distributions
857: within the boxed area of Figure 4 as a function of FWHM of the Fe XII
858: 195 \AA\ line. The temperature is derived from the
859: \ion{Fe}{13}/\ion{Fe}{12} ratio. Bottom panels: histograms of
860: centroid shift (left) and FWHM (right) within the boxed area of Figure
861: 4. The centroid shift in m\AA\ is converted to Doppler speed (upper
862: axis). In the bottom right panel, 65 m\AA\ FWHM corresponds to 16 km
863: s$^{-1}$ non-thermal velocity and 110 m\AA\ corresponds to 82 km
864: s$^{-1}$.}
865: \end{figure}
866:
867: \clearpage
868:
869: \begin{figure}[t!]
870: \plotone{f7.ps}
871: \caption{Locations of outflows for the December 11 event (see Figure
872: 4) according to magnitude of outflow. The white areas in the black
873: and white figures represent pixels where the Doppler shift is equal
874: to or less than $Ds$ from zero Doppler shift in Angstrom units.
875: Negative $Ds$ is defined as an outflow.}
876: \end{figure}
877:
878: \clearpage
879:
880: \begin{figure}[t!]
881: \plotone{f8.ps}
882: \caption{Correlation of Doppler speed with non-thermal speed for the
883: data within the boxed area of Figure 4}
884: \end{figure}
885:
886: \clearpage
887:
888: \begin{figure}[t!]
889: \plotone{f9.ps}
890: \caption{Comparison of Doppler shifts interpreted as outflows for the
891: December 11 event (see Figure 4 boxed area) in different spectral
892: lines (see Table 2) formed over a range of temperatures.}
893: \end{figure}
894:
895: \clearpage
896:
897: \end{document}
898: