0712.0756/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{graphicx} 
3: %\usepackage{figcaps}
4: %\usepackage{balance}
5: %\usepackage{aastex_plus}
6: \newcommand{\x}{\ensuremath{\times}}
7: \newcommand{\arc}{\ensuremath{^{\prime\prime}}}
8: \newcommand{\none}{\ensuremath{^{-1}}}
9: \shorttitle{Steady Heating Model for Active Region Core}
10: \shortauthors{Winebarger, Warren \& Falconer}
11: \renewcommand{\floatpagefraction}{1.0}
12: 
13: %\figcapsoff
14: 
15: 
16: \begin{document}
17: 
18: %% ----------------------------------------------------------------------
19: %% --- title page
20: %% ----------------------------------------------------------------------
21: 
22: \title{Modeling X-ray Loops and EUV "Moss" in an Active Region Core}
23: 
24: \author{Amy R. Winebarger} 
25: \affil{Alabama A\&M University, 4900 Meridian Street, P.O. Box 1268, Normal, AL 35762; winebarger@physics.aamu.edu}
26: \author{Harry P. Warren} 
27: \affil{Naval Research Laboratory, Washington, DC 20375; hwarren@nrl.navy.mil}
28: \author{David A. Falconer}
29: \affil{Marshall Space Flight Center, SD50, Space Science Department, 
30: 	Huntsville Al 35812; david.falconer@msfc.nasa.gov}
31: 
32: %% ----------------------------------------------------------------------
33: %% --- Abstract
34: %% ----------------------------------------------------------------------
35: 
36: \begin{abstract}
37: 
38: The Soft X-ray intensity of loops in active region cores and corresponding footpoint, or moss, intensity observed
39: in the EUV remain steady for several hours of observation.  The steadiness of the emission
40: has prompted many to suggest that the heating in these loops must also be steady, though
41: no direct comparison between the observed X-ray and EUV intensities and the steady heating solutions of the hydrodynamic 
42: equations has yet been made.  In this paper, we perform these simulations and   
43: simultaneously model the X-Ray and EUV moss intensities in one active region core with steady uniform
44: heating.  To perform this task, we introduce a new technique to constrain the model parameters 
45: using the measured EUV footpoint intensity to infer a heating rate.  
46: Using an ensemble of loop structures derived from magnetic field extrapolation of photospheric field, 
47: we associate each field line with a EUV moss intensity, then 
48: determine the steady uniform heating rate on that field line that reproduces the observed EUV intensity within 5\% for a
49: specific cross sectional area, or filling factor.
50: We then calculate the total X-ray filter intensities from all loops in the ensemble and compare this to the
51: observed X-ray intensities.  We complete this task iteratively to determine the filling factor that returns the 
52: best match to the observed X-ray intensities.
53: We find that a filling factor of 8\%  and loops that expand with height provides the best
54: agreement with the intensity in two X-ray filters, though the simulated 
55: SXT Al12 intensity is  147\% the observed intensity and the
56: SXT AlMg intensity is 80\% the observed intensity.
57: From this solution, we determine the required heating rate
58: scales as ${\bar{B}}^{0.29} L^{-0.95}$.  Finally we discuss the future potential of this type of modeling, such as
59: the ability to use density measurements to fully constrain filling factor, and its
60: shortcomings, such as
61: the requirement to use potential field extrapolations to approximate the coronal field. 
62: \end{abstract}
63: 
64: \keywords{Sun: corona}
65: 
66: %% ----------------------------------------------------------------------
67: %% --- Introduction
68: %% ----------------------------------------------------------------------
69: 
70: \section{Introduction}
71: 
72: Determining the timescale of the heating in the solar corona is an important clue to the coronal 
73: heating mechanism.  One strong argument for time independent (steady) heating is the steady emission of observed X-ray 
74: loops in the cores of active regions (e.g., \citealt{yoshida1996}).   The footpoints of these X-ray loops
75: form the reticulated pattern in EUV images typically called ``moss'' (e.g., \citealt{berger1999,fletcher1999}.)
76: Though the intensity of the moss can vary significantly on short time and spatial scales, it
77: is believed that most of the variation comes from spicular material that is embedded
78: within the moss that can obscure and absorb the moss emission (\citealt{depontieu1999}).  The moss
79: intensity  averaged over  small regions is constant for many hours of observations (\citealt{antiochos2003}),
80: further corroborating the hypothesis that the heating in these core regions is steady.
81: 
82: Due to the abundance of loops in the core region, individual loop properties cannot be extracted
83: making it difficult to compare these regions directly with steady heating solutions of the hydrodynamic
84: equations.  Instead, average values for the moss intensities have been compared to the 
85: expected intensity for the footpoints of ``typical'' hot loops.  For instance, \cite{martens2000} found that typical moss
86: intensities matched expected intensities for the footpoints of multi-million degree loops when a 10\% filling
87: factor was included.  Such a filling factor is widely accepted for hot X-ray loops (e.g., \citealt{porter1995}).
88: 
89: Because individual loop properties cannot be extracted, an active region core must be modeled as an
90: ensemble of loops.   
91: There have been several recent attempts to forward model the entire solar corona (\citealt{schrijver2004}),
92: active regions (\citealt{lundquist2006, warren2006, warren2006b}), or bright points
93: (\citealt{brooks2007}) using ensembles of loops that each
94: satisfy the one-dimensional hydrodynamic equations.
95: In these attempts, the volumetric heating rate, $E$, was assumed to be a function of average 
96: magnetic field strength, $\bar{B}$, and loop length, $L$, i.e., $ E = E_0 \bar{B}^\alpha/L^\beta$, where different values of $E_0$, 
97: $\alpha$ and $\beta$ were considered.   
98: The formalism for the heating equations was suggested by
99: \cite{mandrini2000} who determined that different heating mechanisms would release energy 
100: as a function of the loop length and average magnetic field strength along a loop.
101: In the previous ensemble studies, it was found that the X-ray intensity could be well matched
102: by the simulations $\alpha = \beta = 1$ .  The EUV intensity, however, was poorly matched in
103: the case of the whole corona (\citealt{schrijver2004}) and active regions (\citealt{warren2006}).
104: (Note that \cite{schrijver2004} considered the heat flux through the base
105: of the loop, $F$, as a function of the magnetic field at the base, $B_0$, instead of the 
106: volumetric heating rate i.e., $F \sim B_0^\alpha/L^\beta$.  Because $E = F/L$ and
107: $\bar{B} \sim B_0/L$, Schrijver's reults are consistent with the other studies 
108: for $\alpha = 1$.)  
109: 
110: These previous analyses attempted to match the distribution of the total intensity of the entire
111: Sun or in an active region.  They included in their comparisons the long EUV loops
112: that are not in hydrostatic equilibrium (e.g., \citealt{lenz1999}) and are believed to be evolving
113: (see \citealt{winebarger2003, warren2003}).  When \cite{warren2006} compared just the simulated moss intensities to
114: the observed moss intensities, they found approximate agreement for heating rate scaling as $\bar{B}/L$ and 
115: with loops expanding with height.
116: 
117: In this analysis, we compare, for the first time, both the EUV moss and X-ray intensities 
118: to the solutions of one-dimensional hydrodynamic equations for steady uniform heating.
119: We use a new approach to simultaneously match the EUV moss intensity
120: of an active region core and the total X-ray intensity in two filters.  We make no a priori assumption
121: about the relationship between the heating rate and the magnetic field strength and loop length.  
122: Instead we use the moss intensity at the loop footpoint to find the best heating magnitude for each individual 
123: loop.  This approach is based on the determination that the
124: moss intensity observed in a narrow-band EUV filter is linearly proportional to the loop
125: pressure multiplied by a filling factor if the heating in the loop is steady and uniform (\citealt{martens2000}).  
126: The pressure combined with the loop length (determined from magnetic field extrapolation)
127: defines explicitly the heating rate and apex temperature for the loop 
128: (\citealt{rosner1978,serio1981}).
129: 
130: In this paper, we analyze EUV and X-ray emission of Active Region 9107 over 12 hours
131: on 31 May 2000.  This active region has a core region that is bright in X-ray images with
132: large patches of moss that are unobscured by overlying loops in EUV images. We find that the
133: intensity in the region varies little over the 12 hours of observation.
134: We determine the loop geometry using potential field extrapolations of photospheric 
135: field measurements. We select the loop footpoints using the moss intensity as a proxy. 
136: We populate the loops by assuming a filling factor, then using the observed EUV intensities at 
137: one footpoint to constrain the heating rate.  
138: We then compare the total intensity in the X-ray filters to the observed intensities.
139: We complete this process iteratively for different filling factors until the simulated X-ray intensities
140: well match the observed X-ray intensities.  We have completed this process for two different assumptions of the
141: cross sectional area of the loop. First
142: we force the cross sectional area of the loop to remain constant as supported by loop observations (\citealt{klimchuk2000});
143: second, we allow the cross sectional area of the loop to expand with height as $B(s)^{-1}$.
144: We find no satisfactory solution can be obtained for loops with constant cross sectional area.   For loops
145: that expand with height, a filling factor of 8\% was found to produce acceptable agreement between
146: the observed and simulated X-ray intensities in both filters.  From this set of solutions, we
147: determine that the heating scales like $\bar{B}^{0.29}/L^{0.95}$.  Finally we discuss future implications
148: of this method of simultaneous modeling.
149: 
150: 
151: %% ----------------------------------------------------------------------
152: %% --- DATA AND ANALYSIS
153: %% ----------------------------------------------------------------------
154: 
155: \begin{figure}[t!]
156: \centerline{
157: \includegraphics{f1.eps}}
158: \caption{{\it Top row:} TRACE 171\,\AA, and SXT AlMg images of the active region.  {\it Bottom row:}
159: Co-aligned MDI magnetogram and field lines determined from potential field extrapolations.
160: \label{fig:multipanel}}
161: \end{figure}
162: 
163: \begin{figure}[t!]
164: \centerline{
165: \includegraphics{f2.eps}}
166: \caption{Top: The evolution of the TRACE 171 intensity summed over the boxes shown in Figure 1 as a function of time.  Bottom: 
167: The evolution of SXT AlMg and Al12 intensity summed over the entire active region. In all images, the solid line shows the average
168: intensity, the dashed lines show the average $\pm$ one standard deviation.  The SXT data points represented with an asterisk were
169: not considered in the average.
170: \label{fig:time}}
171: \end{figure}
172: 
173: \section{Data Analysis}
174: 
175: The goals of this study were to simultaneously model an Active Region core
176: observed in both the Soft X-ray Telescope (SXT) (\citealt{tsuneta1991}) flown
177: on the {\it Yohkoh} satellite and
178: the {\it Transition Region and Coronal Explorer (TRACE)} 
179: (\citealt{handy1999, schrijver1999}).  To find an adequate active region, the
180: available data bases were browsed for long term, simultaneous observations
181: of a relatively simple, non-flaring active region with an unobscured moss region 
182: observable in the EUV.  We selected Active Region 9017 
183: observed for several hours with both instruments on 2000 May 31.
184: We use the {\it TRACE} 171\,\AA\ filter which is sensitive to the Fe~IX/X lines formed at $\sim 1$\,MK
185: and the  SXT Al, Mg, Mn, C ``sandwich'' filter (AlMg) and the thick Al (Al12) filter  in this study.  
186: The active region is shown in Figure~\ref{fig:multipanel}.  
187: 
188: \subsection{Temporal Evolution}
189: 
190: The evolution of the TRACE
191: intensities averaged over the two regions of moss shown on the
192: TRACE 171 image in Figure~\ref{fig:multipanel} are shown on the top two panels of
193: Figure~\ref{fig:time}.  The average intensity in the left hand box is 14.0$\pm$ 0.6 DN~s$^{-1}$~pix$^{-1}$ and
194: in the right hand box is 9.9 $\pm$ 0.9 DN~s$^{-1}$~pix$^{-1}$.  The errors given are one standard 
195: deviations in the mean values and are less than 10\% of the average values in both cases meaning the
196: intensity in these moss regions remain relatively constant over the 12.5 hours of observation.
197: 
198: The evolutions of the total intensity summed over the active region observed with the SXT AlMg and 
199: Al12 filters are shown in the bottom two panels of Figure~\ref{fig:time}.
200: The active region brightens significantly in both filters between 7:30 and 9:00 UT due to the evolution of two 
201: loops which have previously been studied (\citealt{winebarger2005}).  Excluding this time frame from consideration, 
202: the average of the total intensity was calculated and is shown as a solid line.  In the AlMg filter, the
203: average intensity is 2.2 $\pm  0.2 \times 10^{6}$ DN~s$^{-1}$ and 4.7 $\pm 1.0 \times 10^{4}$ DN~s$^{-1}$ in the
204: AlMg and Al12 filters, respectively.  The standard deviation in the mean value of the AlMg intensity
205: is less than 10\%, while the standard deviation in the Al12 intensity is 20\%. 
206: 
207: Excluding the time frame of the evolving loops, this active region core remains relatively steady for 12.5+ hours
208: of observation.  For the remainder of the paper, we use
209: the {\it TRACE} 171\AA\ image taken at 13:09:33 UT to be representative of the moss intensity
210: and the average SXT AlMg and SXT Al12 intensities for comparison.
211: 
212: %\begin{figure}[t!]
213: %$\centerline{
214: %\resizebox{9cm}{!}{\includegraphics{mossregions.eps}}}
215: %\caption{
216: %\label{fig:mossregions}}
217: %\end{figure}
218: 
219: 
220: \subsection{Determining Loop Geometry}
221: 
222: Because it is difficult to extract information about individual loops from the observations, we use
223: potential field extrapolations of the co-aligned photospheric field measurements  to approximate the loop geometry.
224: We use photospheric field measurements from the Michelson Doppler Imager (MDI, \citealt{scherrer1995}) to estimate
225: the coronal field.   
226: The MDI magnetogram used in this study was taken at 12:48:00 UT and is shown in the second row of Figure~\ref{fig:multipanel}.
227: We also examined a vector magnetograph from the Marshall 
228: Space Flight Center Vector Magnetograph (MSFCVM, \citealt{hagyard1982}) taken
229: at 16:09 UT.  From the vector magnetograph, we determined that the active region was well approximated as potential.
230: 
231: To select the footpoints of the loops, we use the moss region of the TRACE 171 image as a proxy.   We first
232: select all pixels within the moss regions.
233: %, shown as contours in Figure \ref{fig:mossregions}.  
234: These regions
235: were identified visually.  We use the center of the TRACE pixels within these regions as the starting points to
236: trace the field lines.  If we use all the TRACE pixels in these regions as starting points, we would have over
237: 30,000 field lines.  To reduce the number to a more manageable amount, 
238: we bin the TRACE images to a resolution of 1.5\arc; the result being one field line for each 9 TRACE high
239: resolution pixels or a total of about 3,000 field lines.  
240: %We show in the discussion section how this simplification effects the results.
241: We trace the field lines from a height of 2.5\,Mm above the solar surface which is the approximately the measured
242: height of the moss above the limb (\citealt{martens2000}).  Starting at this height also circumvents the problem of having 
243: locally closing field lines that never reach coronal heights (\citealt{warren2006}).
244: After tracing all the field lines, we compare the terminating footpoints and remove any field lines that are duplicates. 
245: A few representative field lines are shown Figure~\ref{fig:multipanel}.  
246: Note that several of the calculated field lines terminate in the negative polarity sunspot where there is 
247: no moss or SXT emission.  This region is highlighted
248: with a circle in Figure~\ref{fig:multipanel}.
249: 
250: \section{Modeling the Loops}
251: 
252: Before describing the details of the modeling portion of this research, it is useful to 
253: consider the implications of the applicable scaling laws.
254: \cite{martens2000} determined that the intensity in the {\it TRACE} 171\AA\  filter was linearly proportional
255: to the base pressure in the loop times a filling factor, i.e., 
256: \begin{equation}
257: p_0 f = 0.050 I_{171} 
258: \end{equation}
259: where $p_0$ is the base pressure in dyne cm$^{-2}$, $f$ is the volumetric filling factor, and
260: $I_{171}$ is the TRACE 171 intensity in DN s$^{-1}$ pixel$^{-1}$.  If we consider only a single TRACE pixel,
261: the filling factor in the above equation would then be related to the cross sectional
262: area of the loop.  A filling factor of unity would imply that the cross sectional area of the loop was at least
263: equal to the TRACE pixel area.  A filling factor of less than 1 would imply that the cross sectional area of the
264: loop was a fraction of the TRACE pixel size.
265: 
266: The above relationship can then be combined with the RTVS scaling laws for uniform heating
267: (\citealt{rosner1978,serio1981}) , i.e.,
268: \begin{eqnarray}
269: T_{max} \sim 1.4 \times 10^3 (p_0 L)^\frac{1}{3} \exp[0.04L/s_p] \\
270: E_H \sim 9.8 \times 10^4 p_0^\frac{7}{6} L^{-\frac{5}{6}} \exp[-0.5L/s_p]
271: \end{eqnarray}
272: where $T_{max}$ is the maximum temperature in the loop in Kelvin,  $L$ is the half length
273: of the loop in cm, $E_H$ is the volumetric heating rate of the loop in ergs cm$^{-3}$ s$^{-1}$,
274: and $s_p$ is the gravitational scale height (generally 47 Mm/MK).
275: These scaling laws were derived using a simplified radiative loss function and assuming
276: the loop semi-circular and perpendicular to the solar surface.  
277: 
278: For a single loop, if the moss intensity and loop length were known, the only free parameter in the
279: above set of equations is the filling factor.  Hence, if other constraining measurements, such as X-ray intensities
280: were also available, it would be possible to calculate the solutions for various filling factors and determine
281: which filling factor satisfactorily reproduced the X-ray intensities.  If no filling factor could be
282: found to return the X-ray intensities, we would conclude that the loop could not be modeled with steady, uniform
283: heating.
284: 
285: This is exactly the test we wish to perform in this study, but 
286: instead of a single loop, we consider an ensemble of such loops.  We follow the same process outlined above 
287: of using the moss intensity to constrain the heating rate on each individual loop for a given filling factor and we further
288: assume that all loops in the bundle have the same filling factor.  Instead of using the scaling 
289: laws given above, however, we solve the steady-state hydrodynamic equations that include the geometry for each loop.  
290: For each filling factor, we solve the equations for all the loops in the bundle, then calculate the total X-ray intensity of
291: all the loops  and compare this value to the observed X-ray intensity in both SXT filters.    
292:  We  complete this process twice, first assuming all loops in the bundle have a constant cross sectional area
293: and then assuming that each loop expands proportional to the inverse of its field strength.
294: 
295: For each field line and filling factor, we use the relationships suggested by \cite{martens2000} and 
296: \cite{rosner1978} to get a first approximation for the heating rate in the loop.  Because these previous works depended
297: on some simplifications and approximations to solve the one dimensional hydrodynamic equations,
298: the heating rate found is only a rough guess to the heating rate 
299: required to match the observed EUV intensities.
300: Using this guess, we then compute the solution to the hydrodynamic equations using a numerical code 
301: created by Aad van Ballegooijen (\citealt{hussain2002, schrijver2005}).   This code allows for the
302: loop geometry and area expansion to be included in the solution.   (Note that we use the radiative loss function calculated
303: by \cite{brooks2006}; all instrument response functions were calculated using the same atomic data and are fully
304: consistent with the radiative loss function and one another.) After computing the solution, we fold the density and temperature
305: through the {\it TRACE} 171 filter response function, compute the simulated moss intensity, and compare
306: the simulated moss intensity to the observed moss intensity associated with that
307: fieldline. If the simulated moss intensity is too low, we increase the heating rate; if the simulated moss intensity
308: is too high, we decrease the heating.  We complete this procedure iteratively until the 
309: simulated moss intensity is within 5\% of the observed moss intensity.  The result is the uniform heating
310: rate and resulting density and temperature on the field line that reproduces the moss intensity associated with the footpoint.
311: After the best hydrodynamic solution for each field line has been computed, we then 
312: calculate the resulting SXT AlMg and Al12  by convolving the density and temperature with the
313: SXT filter response functions.  
314: We sum the SXT filter intensities from all the loops in the ensemble then compare it with the observed AlMg and Al12 
315: intensities.
316: 
317: \section{Results}
318: 
319: Figure~\ref{fig:int_rat} shows ratio of the simulated SXT filter intensities to the observed filter
320: intensities as a function of filling factor for the two different geometry assumptions.  
321: The plot on the left assumes the loops in the ensemble have  constant cross sections, while the plot on the right assumes the 
322: area of each loop expands proportional to $B(s)^{-1}$.  
323: The AlMg filter ratio is shown as the 
324: solid line with crosses and the Al12 ratio is shown as the dashed line with asterisks.  
325: When the curve equals 1, the simulated intensity matches the observed intensity at that filling factor.
326: Horizontal lines show twice the standard deviation implied by the AlMg observations (solid) and Al12 observations
327: (dashed).   For a positive result to be found, the Al12 curve must be within the dashed horizontal lines at the
328: same filling factor as the AlMg curve is within the solid horizontal lines.
329: For the constant cross-section case, no filling factor returns a solution that is within two standard deviations
330: of the observed intensities at the same filling factor.  For the expanding area geometry, a filling factor of 8\% estimated 
331: the AlMg intensity at 2 standard deviations less than the observed intensity (80\% of the observed value) and a Al12 filter 
332: intensity at 2.3 standard deviations above the observed intensity (147\% of the observed value).  This is the best match
333: between the simulations and observations.
334: 
335: \begin{figure*}[t!]
336: \centerline{
337: \resizebox{18cm}{!}{\includegraphics{f3.eps}}}
338: \caption{The ratio between the simulated X-ray intensity and observed X-ray intensity is shown above for the SXT AlMg filter
339: (solid line with crosses) and the SXT Al12 filter (dashed line with asterisks).  The plot on the left shows the results
340: for the constant cross-section case, the plot on the right shows the results for the expanding area case.  The horizontal
341: lines represent two standard deviations in the AlMg intensity (solid) and Al12 intensity (dashed).  The best fit to the
342: X-ray intensities is a filling factor of 8\% in the expanding area case.
343: \label{fig:int_rat}}
344: \end{figure*}
345: 
346: \begin{figure}[t!]
347: \centerline{
348: \includegraphics{f4.eps}}
349: \caption{Left: The observed TRACE 171 and SXT AlMg images.  Right: The simulated TRACE 171 and SXT AlMg images for the
350: best fit case.  The images are displayed linearly and scaled identically.
351: \label{fig:comparison}}
352: \end{figure}
353: 
354: \begin{figure*}[t!]
355: \centerline{
356: \resizebox{18cm}{!}{\includegraphics{f5.eps}}}
357: \caption{A comparison of the simulated and observed TRACE 171 intensities (left) and SXT AlMg intensities (right) 
358: along a single horizontal cut.
359: \label{fig:lc}}
360: \end{figure*}
361: 
362: Figure~\ref{fig:comparison} shows the observed {\it TRACE} 171 and SXT AlMg images as well as the simulated images 
363: for the case that provides the best fit.  
364: The images are displayed linearly with identical scaling.  In all images, a circle is drawn that highlights a region
365: where the simulated and observed morphology differ.  In the observations in this region, there are bright extended EUV loop legs, 
366: but no moss.  There is also little SXT emission in this region. All the simulated images have both moss and SXT emission in the
367: region.   Additionally, there are several bright dots in the simulated EUV image that are not present in the 
368: observation.  These represent footpoint emission from a field line that originates in a moss region, but does
369: not terminate in a moss region region.
370: In Figure~\ref{fig:lc}, we show the intensity along a horizontal cut across both the TRACE 171 images and the SXT AlMg images.  
371: The two arrows in Figure~\ref{fig:comparison} indicate the vertical position of the cut.  
372: The SXT intensities are comparable
373: shapes, though the simulated intensity is less than the observed intensity and shifted to the right.
374: 
375: 
376: 
377: %% ----------------------------------------------------------------------
378: %% --- DISCUSSION
379: %% ----------------------------------------------------------------------
380: 
381: \section{Discussion}
382: 
383: \begin{figure*}[t!]
384: \centerline{
385: \resizebox{18cm}{!}{\includegraphics{f6.eps}}}
386: \caption{The best heating rate for each field line as a function
387: of the base magnetic field strength (top left), average
388: magnetic field strength (top right), and loop length (bottom
389: left).  The heating rate calculated from the fit parameters
390: is also shown as a function of the best heating rate (bottom right).
391: \label{fig:ebl}}
392: \end{figure*}
393: 
394: In this paper, we have used a new technique to infer the steady heating rate from
395: the EUV moss intensities to test whether the loops in active region cores agree with the solutions
396: of the hydrodynamic equations for steady, uniform heating. 
397: We have simulated the core of Active Region 9017 using steady uniform heating
398: along potential field lines to match within 5\% the TRACE 171\,\AA\ moss intensities at
399: the footpoint.  We find the best match between the simulated and observed SXT AlMg and Al12 intensities
400: occurs at a filling factor of 8\% for loops that expand with height proportional to $B(s)^{-1}$.
401: 
402: Our intention with this research was to test whether the steady EUV moss and X-ray core
403: emission could self-consistently agree with ensembles of steadily heated loops.  
404: These results demonstrate that the intensities in these regions can, at best, be matched to steadily uniformly
405: heated loops within approximately two standard deviations in both SXT filters.  This disparity is acceptable
406: considering the systematic errors associated with this study, such as assuming the geometry of the 
407: field was well represented by the potential field and that all loops in the ensemble had the same
408: filling factor or cross sectional area.  Better agreement may have been possible if we had allowed
409: the heating to be non-uniform, but we did not examine this possibility.
410: 
411: Density measurements at the footpoints of the active region core loops would allow for the filling factor to be 
412: measured at each of the loop footpoints.  Knowing the filling factor, moss intensity, and loop length would fully constrain the 
413: steady uniform heating equations and provide a true test of the steady heating model.  This density measurement is now possible
414: with the EIS instrument on Hinode (\citealt{warren2007}). 
415: 
416: 
417: Unlike previous studies of ensembles of loops which heated the loops based on an assumed heating rate proportional to 
418: $\bar{B}^\alpha/L^\beta$, we have calculated the heating rate along each individual field line 
419: based solely on the observed moss intensity and assumed filling factor.
420: We can now characterize the resulting relationships between the
421: calculated heating rate and other characteristics of the region, such as loop length and field strength.
422: Figure~\ref{fig:ebl} shows the correlation between the calculated heating rate and base field strength,
423: average field strength along the loop, and loop length.  Using a regression technique, we find that
424: the volumetric heating rate is best described by
425: $E = 0.051 (\bar{B}/B_0)^{0.29 \pm 0.03} (L/L_0)^{-0.95 \pm 0.01}$.  This calculated heating rate is shown as a function of the
426: real heating rate in the bottom right of Figure~\ref{fig:ebl}.  
427: In the above equations, $\bar{B}$ is the average magnetic field strength along the loop in Gauss, and $L$ is
428: the loop length in Mm and $B_0$ and $L_0$ are chosen to be 76 Gauss and 29 Mm respectively
429: to be comparable to previous work (\citealt{warren2006}). 
430: The correlation coefficient is
431: 0.71 between the log of average field strength and the log of the heating rate and -.94 between the log of the
432: loop length and log of the heating rate.  There is no correlation between the heating rate and footpoint field 
433: strength (correlation coefficient = -0.047).  
434: 
435: The inverse relationship between the heating rate and loop length matches the results of the other forward
436: modeling studies (\citealt{schrijver2004,warren2006}).  However, the previous studies determined the best scaling
437: with the average magnetic field strength was for $1.0$, where we find a scaling of $0.3$.  This discrepancy is most
438: likely due to our strict matching of the TRACE moss intensity, where previous studies focused on matching the SXT
439: intensity and only did a rough comparison with the observed and simulated moss intensity.
440: 
441: In this study, we used potential field extrapolations to approximate the loop geometry and the
442: moss regions themselves to select the footpoints of the heated field lines.  However,
443: the resulting morphological
444: comparison of the observed images with the simulated images show some discrepancy.
445: Specifically, several field lines terminate in the negative polarity sunspot where there is no SXT or TRACE moss
446: emission observed;  this is shown with the circle in Figures~\ref{fig:multipanel} and \ref{fig:comparison}.  
447: Instead, there are several extended loop legs seen in the EUV.  This could be an indication that the 
448: connectivity of the field was not correct; however
449: the connectivity was not improved when using the vector field data or when considering linear force free
450: field extrapolations for different force free parameters.  
451: Another option is that those loops are heated asymmetrically causing moss on one side of the loop
452: and extended EUV emission on the other side of the loop.  Most coronal heating theories rely on photospheric
453: motions as a driving force. Because photospheric motions are suppressed in sunspot
454: regions, asymmetric heating along loops with one footpoint in a sunspot is a strong possibility.  
455: We did not consider asymmetric or non-uniform heating in this study.  In the future, it may be possible
456: to consider the EUV emission at both footpoints to fully limit the asymmetry of the heating function.
457: 
458: \cite{martens2000} derived the relationship between the base pressure, $p_0$, filling
459: factor, $f$, and TRACE 171 moss intensity, $I_{moss}$, i.e., $p_0 f = 0.050 I_{moss}$
460: using a simple radiative loss function and an analytical approximation to the hydrostatic
461: equations.  We derive this equation using the most recent radiative loss function (\citealt{brooks2006}).
462:  We find the relationship $p_0 f = 0.026 I_{moss}$
463: best represents the simulations.  
464: 
465: 
466: 
467: %% ----------------------------------------------------------------------
468: %% --- ACKNOWLEDGMENTS
469: %% ----------------------------------------------------------------------
470: 
471: \acknowledgments
472: ARW was supported by a NASA Sun-Earth Connection Guest Investigator grant
473: and NSF Career grant.
474: 
475: %% ----------------------------------------------------------------------
476: %% --- REFERENCES
477: %% ----------------------------------------------------------------------
478: 
479: \begin{thebibliography}{}
480: 
481: \bibitem[\protect\citeauthoryear{{Antiochos} et~al.}{{Antiochos}
482:   et~al.}{2003}]{antiochos2003}
483: {Antiochos}, S.~K., {Karpen}, J.~T., {DeLuca}, E.~E., {Golub}, L.,  \&
484:   {Hamilton}, P. 2003, \apj, 590, in press
485: 
486: \bibitem[\protect\citeauthoryear{{Berger} et~al.}{{Berger}
487:   et~al.}{1999}]{berger1999}
488: {Berger}, T.~E., {De Pontieu}, B., {Fletcher}, L., {Schrijver}, C.~J.,
489:   {Tarbell}, T.~D.,  \& {Title}, A.~M. 1999, \solphys, 190, 409
490: 
491: \bibitem[\protect\citeauthoryear{{Brooks} \& {Warren}}{{Brooks} \&
492:   {Warren}}{2006}]{brooks2006}
493: {Brooks}, D.~H.,  \& {Warren}, H.~P. 2006, \apjs, 164, 202
494: 
495: \bibitem[\protect\citeauthoryear{{Brooks} \& {Warren}}{{Brooks} \&
496:   {Warren}}{2007}]{brooks2007}
497: {Brooks}, D.~H.,  \& {Warren}, H.~P. 2007, \apj, submitted
498: 
499: \bibitem[\protect\citeauthoryear{{De Pontieu} et~al.}{{De Pontieu}
500:   et~al.}{1999}]{depontieu1999}
501: {De Pontieu}, B., {Berger}, T.~E., {Schrijver}, C.~J.,  \& {Title}, A.~M. 1999,
502:   \solphys, 190, 419
503: 
504: \bibitem[\protect\citeauthoryear{{Fletcher} \& {de Pontieu}}{{Fletcher} \& {de
505:   Pontieu}}{1999}]{fletcher1999}
506: {Fletcher}, L.,  \& {de Pontieu}, B. 1999, \apjl, 520, L135
507: 
508: \bibitem[\protect\citeauthoryear{{Hagyard} et~al.}{{Hagyard}
509:   et~al.}{1982}]{hagyard1982}
510: {Hagyard}, M.~J., {Cumings}, N.~P., {West}, E.~A.,  \& {Smith}, J.~E. 1982,
511:   \solphys, 80, 33
512: 
513: \bibitem[\protect\citeauthoryear{{Handy} et~al.}{{Handy}
514:   et~al.}{1999}]{handy1999}
515: {Handy}, B.~N., et~al. 1999, \solphys, 187, 229
516: 
517: \bibitem[\protect\citeauthoryear{{Hussain} \& {van Ballegooijen}}{{Hussain} \&
518:   {van Ballegooijen}}{2002}]{hussain2002}
519: {Hussain}, G. A.~J.,  \& {van Ballegooijen}, A. 2002, \apj, in preparation
520: 
521: \bibitem[\protect\citeauthoryear{{Klimchuk}}{{Klimchuk}}{2000}]{klimchuk2000}
522: {Klimchuk}, J.~A. 2000, \solphys, 193, 53
523: 
524: \bibitem[\protect\citeauthoryear{{Lenz} et~al.}{{Lenz} et~al.}{1999}]{lenz1999}
525: {Lenz}, D.~D., {Deluca}, E.~E., {Golub}, L., {Rosner}, R.,  \& {Bookbinder},
526:  J.~A. 1999, \apjl, 517, L155
527: 
528: \bibitem[\protect\citeauthoryear{{Lundquist}}{{Lundquist}}{2006}]{lundquist200%
529: 6}
530: {Lundquist}, L.~L. 2006, AAS Solar Physics Division Meeting, 37, 17.03
531: 
532: \bibitem[\protect\citeauthoryear{{Mandrini}, {D{\' e}moulin}, \&
533:   {Klimchuk}}{{Mandrini} et~al.}{2000}]{mandrini2000}
534: {Mandrini}, C.~H., {D{\' e}moulin}, P.,  \& {Klimchuk}, J.~A. 2000, \apj, 530,
535:   999
536: 
537: \bibitem[\protect\citeauthoryear{{Martens}, {Kankelborg}, \&
538:   {Berger}}{{Martens} et~al.}{2000}]{martens2000}
539: {Martens}, P.~C.~H., {Kankelborg}, C.~C.,  \& {Berger}, T.~E. 2000, \apj, 537,
540:   471
541: 
542: \bibitem[\protect\citeauthoryear{{Porter} \& {Klimchuk}}{{Porter} \&
543:   {Klimchuk}}{1995}]{porter1995}
544: {Porter}, L.~J.,  \& {Klimchuk}, J.~A. 1995, \apj, 454, 499
545: 
546: \bibitem[\protect\citeauthoryear{{Rosner}, {Tucker}, \& {Vaiana}}{{Rosner}
547:   et~al.}{1978}]{rosner1978}
548: {Rosner}, R., {Tucker}, W.~H.,  \& {Vaiana}, G.~S. 1978, \apj, 220, 643
549: 
550: \bibitem[\protect\citeauthoryear{{Scherrer} et~al.}{{Scherrer}
551:   et~al.}{1995}]{scherrer1995}
552: {Scherrer}, P.~H., et~al. 1995, \solphys, 162, 129
553: 
554: \bibitem[\protect\citeauthoryear{{Schrijver} et~al.}{{Schrijver}
555:   et~al.}{2004}]{schrijver2004}
556: {Schrijver}, C.~J., {Sandman}, A.~W., {Aschwanden}, M.~J.,  \& {DeRosa}, M.~L.
557:   2004, \apj, 615, 512
558: 
559: \bibitem[\protect\citeauthoryear{{Schrijver} et~al.}{{Schrijver}
560:   et~al.}{1999}]{schrijver1999}
561: {Schrijver}, C.~J., et~al. 1999, \solphys, 187, 261
562: 
563: \bibitem[\protect\citeauthoryear{{Schrijver} \& {van Ballegooijen}}{{Schrijver}
564:   \& {van Ballegooijen}}{2005}]{schrijver2005}
565: {Schrijver}, C.~J.,  \& {van Ballegooijen}, A.~A. 2005, \apj, 630, 552
566: 
567: \bibitem[\protect\citeauthoryear{{Serio} et~al.}{{Serio}
568:   et~al.}{1981}]{serio1981}
569: {Serio}, S., {Peres}, G., {Vaiana}, G.~S., {Golub}, L.,  \& {Rosner}, R. 1981,
570:   \apj, 243, 288
571: 
572: \bibitem[\protect\citeauthoryear{{Tsuneta} et~al.}{{Tsuneta}
573:   et~al.}{1991}]{tsuneta1991}
574: {Tsuneta}, S., et~al. 1991, \solphys, 136, 37
575: 
576: \bibitem[\protect\citeauthoryear{{Warren} \& {Winebarger}}{{Warren} \&
577:   {Winebarger}}{2006a}]{warren2006}
578: {Warren}, H.~P.,  \& {Winebarger}, A.~R. 2006a, \apj, 645, 711
579: 
580: \bibitem[\protect\citeauthoryear{{Warren} \& {Winebarger}}{{Warren} \&
581:   {Winebarger}}{2006b}]{warren2006b}
582: {Warren}, H.~P.,  \& {Winebarger}, A.~R. 2006b, \apj, submitted
583: 
584: \bibitem[\protect\citeauthoryear{{Warren}, {Winebarger}, \& {Mariska}}{{Warren}
585:   et~al.}{2003}]{warren2003}
586: {Warren}, H.~P., {Winebarger}, A.~R.,  \& {Mariska}, J.~T. 2003, \apj, 593,
587:   1174
588: 
589: \bibitem[\protect\citeauthoryear{{Warren}, {Winebarger}, {Mariska}, {Doschek}, \& {Hara}}{{Warren}
590:   et~al.}{2007}]{warren2007}
591: {Warren}, H.~P., {Winebarger}, A.~R.,  {Mariska}, J.~T., {Doschek}, G.~A., \& {Hara}, H. 2007, \apj, in press
592: 
593: \bibitem[\protect\citeauthoryear{{Winebarger} \& {Warren}}{{Winebarger} \&
594:   {Warren}}{2005}]{winebarger2005}
595: {Winebarger}, A.~R.,  \& {Warren}, H.~P. 2005, \apj, 626, 543
596: 
597: \bibitem[\protect\citeauthoryear{{Winebarger}, {Warren}, \&
598:   {Mariska}}{{Winebarger} et~al.}{2003}]{winebarger2003}
599: {Winebarger}, A.~R., {Warren}, H.~P.,  \& {Mariska}, J.~T. 2003, \apj, 587, 439
600: 
601: \bibitem[\protect\citeauthoryear{{Yoshida} \& {Tsuneta}}{{Yoshida} \&
602:   {Tsuneta}}{1996}]{yoshida1996}
603: {Yoshida}, T.,  \& {Tsuneta}, S. 1996, \apj, 459, 342
604: 
605: \end{thebibliography}
606: 
607: 
608: %%%%%%%%%%%%%%%%%%%%%%%%%%
609: %%                      %%
610: %%      Figure Captions %%
611: %%                      %%
612: %%%%%%%%%%%%%%%%%%%%%%%%%%
613: 
614: 
615: \end{document}
616: