0805.2013/MS.TEX
1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2005 December 5
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8: 
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12: 
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: %%\documentclass[12pt,preprint]{aastex}
19: 
20: %% manuscript produces a one-column, double-spaced document:
21: 
22: \documentclass[preprint]{aastex}
23: %%\documentclass[manuscript]{aastex}
24: 
25: \def\lsim{\mathrel{\rlap{\lower4pt\hbox{\hskip1pt$\sim$}}
26:     \raise1pt\hbox{$<$}}}         %less than or approx. symbol
27: \def\gsim{\mathrel{\rlap{\lower4pt\hbox{\hskip1pt$\sim$}}
28:     \raise1pt\hbox{$>$}}}         %greater than or approx. symbol
29: 
30: \usepackage{natbib}
31: \usepackage{graphicx}
32: \pdfoutput=1
33: 
34: 
35: %% preprint2 produces a double-column, single-spaced document:
36: 
37: 
38: %% Sometimes a paper's abstract is too long to fit on the
39: %% title page in preprint2 mode. When that is the case,
40: %% use the longabstract style option.
41: 
42: %% \documentclass[preprint2,longabstract]{aastex}
43: 
44: %% If you want to create your own macros, you can do so
45: %% using \newcommand. Your macros should appear before
46: %% the \begin{document} command.
47: %%
48: %% If you are submitting to a journal that translates manuscripts
49: %% into SGML, you need to follow certain guidelines when preparing
50: %% your macros. See the AASTeX v5.x Author Guide
51: %% for information.
52: 
53: 
54: %% You can insert a short comment on the title page using the command below.
55: 
56: 
57: %% If you wish, you may supply running head information, although
58: %% this information may be modified by the editorial offices.
59: %% The left head contains a list of authors,
60: %% usually a maximum of three (otherwise use et al.).  The right
61: %% head is a modified title of up to roughly 44 characters.
62: %% Running heads will not print in the manuscript style.
63: 
64: 
65: %% This is the end of the preamble.  Indicate the beginning of the
66: %% paper itself with \begin{document}.
67: 
68: \begin{document}
69: 
70: %% LaTeX will automatically break titles if they run longer than
71: %% one line. However, you may use \\ to force a line break if
72: %% you desire.
73: 
74: \title{CN-Cycle Solar Neutrinos and Sun's Primordial Core Metalicity}
75: 
76: %% Use \author, \affil, and the \and command to format
77: %% author and affiliation information.
78: %% Note that \email has replaced the old \authoremail command
79: %% from AASTeX v4.0. You can use \email to mark an email address
80: %% anywhere in the paper, not just in the front matter.
81: %% As in the title, use \\ to force line breaks.
82: 
83: \author{W. C. Haxton}
84: \affil{Institute  for   Nuclear  Theory  and  Dept.   Physics,  University  of
85:   Washington, Seattle, WA 98195, USA}
86: \email{haxton@u.washington.edu}
87: \and
88: \author{A. M. Serenelli}
89: \affil{Institute for Advanced Study, Princeton, NJ 08540, USA}
90: \email{aldos@ias.edu}
91: 
92: %% Notice that each of these authors has alternate affiliations, which
93: %% are identified by the \altaffilmark after each name.  Specify alternate
94: %% affiliation information with \altaffiltext, with one command per each
95: %% affiliation.
96: 
97: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
98: %% style, abstract will output a Received/Accepted line after the
99: %% title and affiliation information. No date will appear since the author
100: %% does not have this information. The dates will be filled in by the
101: %% editorial office after submission.
102: 
103: \begin{abstract}
104: We argue that it may be possible to exploit neutrinos from the CN cycle and pp
105: chain  to determine the  primordial solar  core abundances  of C  and N  at an
106: interesting level of  precision.  Such a measurement would  allow a comparison
107: of  the  Sun's  deep interior  composition  with  it  surface, testing  a  key
108: assumption of the standard solar  model (SSM), a homogeneous zero-age Sun.  It
109: would   also  provide   a   cross-check  on   recent  photospheric   abundance
110: determinations that have altered the  once excellent agreement between the SSM
111: and  helioseismology.   As  further   motivation,  we  discuss  a  speculative
112: possibility   in  which   photospheric  abundance/helioseismology   puzzle  is
113: connected  with  the   solar-system  metal  differentiation  that  accompanied
114: formation of the gaseous giant planets.
115: 
116: The  theoretical  relationship between  core  C and  N  and  the $^{13}$N  and
117: $^{15}$O solar neutrino fluxes can be  made more precise (and more general) by
118: making use of the Super-Kamiokande and SNO $^8$B neutrino capture rates, which
119: calibrate  the  temperature  of  the  solar  core.  The  primordial  C  and  N
120: abundances can then be obtained from  these neutrino fluxes and from a product
121: of nuclear  rates, with little  residual solar model dependence.   We describe
122: some of the  recent experimental advances that could  allow this comparison to
123: be  made  (theoretically)  at  the  $\sim$  9\%  level,  and  note  that  this
124: uncertainty may  be reduced further  due to ongoing  work on the  S-factor for
125: $^{14}$N(p,$\gamma$).  The  envisioned measurement might be  possible in deep,
126: large-volume detectors using organic scintillator, e.g., Borexino or SNO+.
127: \end{abstract}
128: 
129: 
130: %% Keywords should appear after the \end{abstract} command. The uncommented
131: %% example has been keyed in ApJ style. See the instructions to authors
132: %% for the journal to which you are submitting your paper to determine
133: %% what keyword punctuation is appropriate.
134: 
135: \keywords{}
136: 
137: \section{Introduction}
138: 
139: Over  three  decades  one of  the  most  intriguing  problems in  physics  and
140: astrophysics has  been that  of the missing  solar neutrinos,  the discrepancy
141: between        Ray        Davis's        chlorine-detector        measurements
142: (\citealt{davis,davis_bahc}) and the 
143: predictions  of  the standard  solar  model  (SSM)  developed by  Bahcall  and
144: collaborators \citep{bpb00,bp04,bsb_mc}, by Turck-Chieze and collaborators
145: \citep{brun1,brun2}, and  others.  Part  of the problem's  fascination has
146: been the tension between stellar theory and particle physics: arguments for new
147: neutrino physics required one to believe that {\it ab initio} models correctly
148: predicted the solar core temperature to an accuracy of about 1\%.
149: 
150: Gradually  the   combination  of  quantitative  tests  of   the  solar  model,
151: particularly determinations  of the interior sound  speed via helioseismology,
152: and new solar neutrino experiments, Kamioka \citep{kamioka}, SAGE \citep{sage} 
153: and GALLEX/GNO \citep{gallex}, and Super-Kamiokande \citep{sk1,sk2}, made the 
154: arguments for  new physics compelling.   The pattern of solar  neutrino fluxes
155: proved difficult to attribute to any plausible variation in the SSM.  With the
156: direct detection  of both  electron- and heavy-flavor  solar neutrinos  in the
157: Sudbury  Neutrino  Observatory,  solar   neutrino  experiments  had  not  only
158: demonstrated that electron neutrinos oscillate  on their way to the earth, but
159: also     determined    the     parameters    governing     that    oscillation
160: \citep{sno1,sno2,sno3}.  
161: Progress has continued with  the KamLAND \citep{kamland} reactor experiment and
162: the  Borexino collaboration's  efforts \citep{borexino}  to  measure low-energy
163: solar neutrinos in real time.   Borexino's recent results for the $^7$Be solar
164: neutrino flux are consistent with Large Mixing Angle (LMA) solution.
165: 
166: Because the incorporation of neutrino  mass and mixing into the standard model
167: requires  new physics,  the  field's  attention has  naturally  turned to  the
168: unresolved particle physics questions, such as the mass hierarchy, mass scale,
169: third mixing  angle, and CP-violating phases. \citet{APS}  summarizes the open
170: questions and the envisioned future experimental program.  
171: 
172: Here we return  to one of the initial motivations  for solar neutrino physics,
173: using the neutrino flux  as a probe of the SSM.  We  will argue that important
174: tests  of the Sun  and its  initial conditions  can be  made by  measuring the
175: CN-cycle neutrinos.  Such a measurement would test a key assumption of the SSM
176: -- that  convective mixing  during the  early pre-main-sequence  Hayashi phase
177: produced a homogeneous Sun, and  that subsequent evolution has not appreciably
178: altered the  distribution of metals -- an  assumption that may now  be in some
179: degree of  conflict with  helioseismology.  This assumption  is the  basis for
180: taking the SSM's  primordial core metal abundances from  today's surface metal
181: abundances.   We  argue   that  a  series  of  recent   advances  --  SNO  and
182: Super-Kamiokande  measurements  of  the  $^8$B neutrinos,  new  cross  section
183: measurements for $^{14}$N(p,$\gamma$) and  for certain pp-chain reactions, and
184: detector developments such as Borexino  and SNO+, might allow one to determine
185: the primordial  abundances of  C and  N, with little  dependence on  the solar
186: model: to  good accuracy,  these abundances can  be derived  from experimental
187: quantities, namely the CN-cycle neutrino  fluxes, the $^8$B neutrino flux, and
188: nuclear cross sections and oscillation parameters measured in the laboratory.
189: 
190: Such a  check of the SSM  is of added  interest because recent 3D  modeling of
191: photospheric absorption  lines has  led to a  downward revision in  the metal
192: content  of  the solar  convective  zone  \citep{ags05}.  This  significantly
193: alters the 
194: once  spectacular  agreement  between  the  SSM  and  helioseismology  in  the
195: temperature region below the solar convective zone, $\sim$ 2-5 $\times$ 10$^6$
196: K \citep{bbps05,bsb05,antia,montalban}. In this 
197: region C, N, O, Ne, and Ar are partially 
198: ionized  and  particularly  O and  Ne  have  a  significant influence  on  the
199: radiative opacity. 
200: % Wick, I've removed the following paragraph because, as it turns out, C and N
201: % have  very little  influence on the  radiative opacity below  the convective
202: % zone. 
203: % That is, the specific metals that can be probed via the CNO neutrino fluxes,
204: % principally C and N, are among those that govern the opacity in the region
205: % where helioseismology discrepancies are the largest. 
206: 
207: A  quantitative comparison between  the Sun's  surface and  core metalicities
208: could prove  useful in understanding  the chemical evolution of  other gaseous
209: bodies in  our solar system,  whose interiors are  not as easily  probed.  The
210: Galileo and Cassini  missions found significant metal enrichments  in the H/He
211: atmospheres of Jupiter and Saturn, e.g.,  abundances of C and N of $\sim$ four
212: times solar for Jupiter and $\sim$ 4-8 for Saturn \citep{guillot}.  Planetary 
213: models that  take account of these  data indicate that the  gaseous giants are
214: very significant  solar-system metal reservoirs.   We discuss, because  of the
215: size of these reservoirs and the  time they were created, the possibility that
216: they  might have  some  connection  with the  current  conflict between  solar
217: interior   (helioseismology)  and   surface  (photospheric   absorption  line)
218: abundance determinations.
219: 
220: Finally, just  as the solar  neutrino program to  date has provided  our first
221: quantitative  test of  the nuclear  astrophysics governing  proton  burning in
222: low-mass main-sequence stars, a solar  CN-cycle neutrino program would give us
223: our  first   experimental  constraints  on   the  process  by   which  massive
224: main-sequence stars burn hydrogen.  The CN  cycle is thought to have driven an
225: early convective stage  in our Sun, and is also important  to the evolution of
226: the first generation of massive metal-poor stars, where it turns on only after
227: carbon has been synthesized by the triple alpha process.
228: 
229: \section{The CNO Bi-Cycle and its Neutrinos}
230: 
231: The need for two mechanisms to  burn hydrogen was recognized in the pioneering
232: work  of  Bethe  and  collaborators.   The pp-chain,  which  dominates  energy
233: production  in  our  Sun  and  other  low-mass  main-sequence  stars,  can  be
234: considered a primary process in  which the chain's ``catalysts'' -- deuterium,
235: $^3$He, and  $^7$Be/$^7$Li, the  elements participating in  intermediate steps
236: shown  in  Fig.~\ref{fig:one} --  are  synthesized  as  the chain  burns  to
237: equilibrium.
238: 
239: But the  sharper $T$-dependence of the  CNO cycle is necessary  to account for
240: the structure  of massive main-sequence  stars.  Unlike the pp-chain,  the CNO
241: bi-cycle  (Fig.~\ref{fig:two})  is a  secondary process:  the catalysts  for H
242: burning  are the  pre-existing metals.   Thus the  CNO contribution  to energy
243: generation is  directly proportional to  the stellar-core number  abundance of
244: the primordial metals.   The CN-cycle, denoted by I  in Fig.~\ref{fig:two}, is
245: an important SSM  neutrino source.  The cycle conserves  the number abundance,
246: but alters the distribution of metals as it burns into equilibrium, eventually
247: achieving equilibrium abundances proportional to the inverse of the respective
248: rates.
249: 
250: The reactions controlling early conversion of metals in the solar core and the
251: approach  to equilibrium  are  $^{12}$C(p,$\gamma$) and  $^{14}$N(p,$\gamma$):
252: these are  the next-to-slowest and  slowest rates in the  lower-temperature CN
253: cycle, respectively.  The  central temperature of the solar  core at the onset
254: of nuclear  burning, $T_7  \sim 1.34$, corresponds  to a $^{12}$C  lifetime of
255: about  2  $\cdot$ 10$^7$  y.   Thus  the  initial out-of-equilibrium  CN-cycle
256: conversion of $^{12}$C  to $^{14}$N in the central region of  the early Sun is
257: complete and  rapid.  The associated energy  release is thought  to render the
258: central portion of the solar core  convectively unstable for a period of about
259: $10^8$ y.   That is, the steep temperature  dependence of $^{12}$C(p,$\gamma$)
260: produces composition, opacity, and  thus thermal gradients sufficient to drive
261: convection.  The temperature  at which the $^{12}$C lifetime  is comparable to
262: the Sun's 4.57 b.y.  lifetime is $T_7 \sim 1.0$.  In the SSM this includes
263: essentially the entire  energy-producing core, $ R \lsim  0.18 R_\odot$ and $M
264: \lsim 0.29 M_\odot$, so that nearly  all of the core's primordial $^{12}$C has
265: been converted  to $^{14}$N.  This  change in the chemical  composition alters
266: the opacity  and, at the 3\%  level, the heavy-element mass  fraction $Z$, SSM
267: effects first explored by \citet{bu88}. 
268: 
269: The  $^{14}$N(p,$\gamma$)  reaction determines  whether  equilibrium has  been
270: achieved.  The $^{14}$N  lifetime is shorter than the age of  the Sun for $T_7
271: \gsim 1.33$.  Therefore equilibrium for the CN cycle has been reached only for
272: $R \lsim  0.1 R_\odot$, corresponding to the  central 7\% of the  Sun by mass.
273: Consequently, over a significant portion  of the outer core, $^{12}$C has been
274: converted   to  $^{14}$N,  but   further  reactions   are  inhibited   by  the
275: $^{14}$N(p,$\gamma$) bottleneck.
276: 
277: The BSP08(GS) SSM \citep{bps08} -- which employs
278: values for $Z$ and the $^{14}$N(p,$\gamma$) S-factor given below -- predicts a
279: modest  CN-cycle  contribution  to   solar  energy  generation  of  0.8\%  but
280: substantial fluxes of neutrinos 
281: %
282: \begin{displaymath}
283: ^{13}\mathrm{N} (\beta^+)^{13}\mathrm{C}~~E_\nu \lsim 1.199 
284: \mathrm{~MeV}~~\phi = (2.93^{+0.91}_{-0.82}) \times 10^8/\mathrm{cm^2s}
285: \end{displaymath} 
286: %
287: \begin{displaymath}
288: ^{15}\mathrm{O} (\beta^+)^{15}\mathrm{N} ~~E_\nu \lsim 1.732
289: \mathrm{~MeV}~~\phi          =          (2.20^{+0.73}_{-0.63})          \times
290: 10^8/\mathrm{cm^2s}. 
291: \end{displaymath}
292: %
293: Here uncertainties reflect conservative abundance  uncertainties  as defined
294: empirically in \citet{bs05}.
295: The first  reaction is part of the  path from $^{12}$C to  $^{14}$N, while the
296: latter  follows $^{14}$N(p,$\gamma$).   Thus neutrinos  from  $^{15}$O $\beta$
297: decay  are produced  in the  central core:  95\% of  the flux  comes  from the
298: CN-equilibrium region, described above.   About 30\% of the $^{13}$N neutrinos
299: come from outside  this region, primarily because of  the continued burning of
300: primordial  $^{12}$C: this  accounts for  the  somewhat higher  flux of  these
301: neutrinos.  There is  also a small but fascinating  contribution from $^{17}$F
302: $\beta$ decay,
303: %
304: \begin{displaymath}
305: ^{17}\mathrm{F}(\beta^+)^{17}\mathrm{O}~~        E_\nu       \lsim       1.740
306: \mathrm{~MeV}~~\phi=(5.82\pm 3.04) \times 10^6/\mathrm{cm^2s} 
307: \end{displaymath}
308: %
309: a reaction  fed by (p,$\gamma$) on  primordial $^{16}$O: the  cycling time for
310: the second  branch of  the CNO  bi-cycle, for solar  core conditions,  is much
311: longer than the  solar age.  The flux of these neutrinos  appears too small to
312: allow  a  test   of  the  Sun's  primordial  oxygen   content  by  this  means
313: \citep{bahcallbook}.
314: 
315: The  SSM makes  several  reasonable assumptions,  including local  hydrostatic
316: equilibrium (the balancing of the gravitational force against the gas pressure
317: gradient), energy  generation by proton  burning, a homogeneous  zero-age Sun,
318: and boundary conditions  imposed by the known mass,  radius, and luminosity of
319: the  present Sun.   It assumes  no significant  mass loss  or  accretion.  The
320: homogeneity assumption  allows the primordial  core metalicity to be  fixed to
321: today's surface  abundances.  Corrections for  the effects of diffusion  of He
322: and the  heavy elements  over 4.57  b.y. of solar  evolution are  included, and
323: generally been helpful in improving  the agreement between SSM predictions and
324: parameters probed in helioseismology.
325: 
326: The assumption  of a homogeneous zero-age  Sun is based on  arguments that the
327: early pre-main-sequence Sun passed through a fully convective, highly luminous
328: Hayashi  phase,   homogenizing  the  Sun.   Yet,  as   recently  discussed  in
329: \citet{winnick}, whether this homogeneity persists
330: until the main sequence depends on  the Sun's metal accretion history.  In the
331: subsequent late pre-main-sequence phase (the Henyey phase), the Sun approaches
332: the  main sequence  by  establishing  and growing  a  radiative core.   Metals
333: accreted onto the Sun in or after this phase would not be mixed into the core.
334: Thus in  principle, if the accreted material  had a metal content  that is not
335: uniform in time,  differences between the surface and core  could arise in the
336: Henyey phase. \citet{winnick} have  discussed scenarios in  which such
337: accretion might produce  a convective zone enriched in  metals relative to the
338: radiative zone.  For many years one  motivation for models with such ``low Z''
339: cores was to  lower the $^8$B neutrino flux,  reducing the discrepancy between
340: the SSM and the results of the Davis experiment.
341: 
342: The SSM  assumes no such differentiation  occurs.  While this  assumption of a
343: homogeneous zero-age Sun may be correct, there are few observational checks on
344: proto-solar evolution.   But one possibility might be  the CN-cycle neutrinos.
345: The flux of  these neutrino should depend nearly linearly  on the initial core
346: abundance of C  and N.  If other uncertainties  affecting predictions of these
347: fluxes  could be brought  under control,  and if  these fluxes  were measured,
348: constraints on the core's primordial C and N might be obtained.
349: 
350: Solar surface  abundances are known, determined from  analyses of photospheric
351: atomic  and  molecular spectral  lines.   Traditionally  the associated  solar
352: atmosphere  modeling has  been done  in one  dimension, in  a time-independent
353: hydrostatic  analysis that incorporates  convection via  mixing-length theory.
354: But  much improved  3D  models of  the  solar atmosphere  have been  developed
355: recently  to treat  the radiation-hydrodynamics  and time  dependence  of this
356: problem.  This  approach is essentially  parameter-free and has been  shown to
357: accurately reproduce  average line  profiles, improve the  consistency between
358: different  line measurements  (e.g.,  among the  various  sources of  C and  O
359: lines), and bring the solar abundances  into better accord with other stars in
360: the solar neighborhood.  The  improved analysis, however, substantially lowers
361: the solar  metalicity from the  previous standard, Z=0.0169  \citep{gs98}, to
362: Z=0.0122  \citep{ags05},  and thus  alters  SSM  predictions. Hereafter  we
363: denote these as the GS and and AGS abundances, respectively. 
364: 
365: Solar models that use  the GS solar composition, the most up  to date of which
366: is the BPS08(GS) \citep{bps08} but including also the BP00 \citep{bpb00}, BP04
367: \citep{bp04}  and BS05(OP)  \citep{bsb05} models,  are in  excellent agreement
368: with those deduced from helioseismology.   But those computed with the revised
369: abundance are in 
370: much poorer  agreement, with  discrepancies exceeding 1\%  in the  region just
371: below  the   convective  zone  (R  $\sim  0.65-0.70   {\rm  R}_\odot$). 
372: Associated  properties of the SSM,  such as
373: the depth of the convective zone and the surface He abundance, are also now in
374: conflict  with helioseismology.   As extensively  discussed  in \citet{bsb_mc}
375: discrepancies   are   significantly   above   measurement  and   solar   model
376: uncertainties. 
377: 
378: The   reduced  core   opacity  also   lowers   the  SSM   prediction  of   the
379: temperature-dependent $^8$B  neutrino flux by about 20\%:  the predicted $^8$B
380: flux using  the GS abundances  and Opacity Project  \citep{opacity} opacities,
381: model BPS08(GS) is 5.95 $\times$ 10$^6$/cm$^2$s, 
382:  which drops  to 4.72  $\times$ 10$^6$/cm$^2$s when  AGS abundances  are used,
383:  model BPS08(AGS).  These results can be compared to the $^8$B neutrino flux
384: deduced from  the 391-day salt-phase  SNO data set  of 4.94 $\pm$  0.21 (stat)
385: $^{+0.38}_{-0.34}$    $\times$     10$^6$/cm$^2$s    \citep{sno_salt}.     The
386: Super-Kamiokande 
387: combined  unbinned (binned)  analysis using  the salt-phase  SNO data  finds a
388: best-fit flux  of 4.91  (4.86)$\times$ 10$^6$/cm$^2$ \citep{hosaka}.   Thus the
389: BPS08(GS) and BPS08(AGS) predictions are 1.2 and 0.95 times the experimental
390: central values  of the  combined analysis.  Both  results are  consistent with
391: experiment, given  current experimental (9.5\%) and  theoretical ($\sim$ 16\%)
392: uncertainties.
393: 
394: Finally,  we  describe a  speculative  scenario  to  illustrate why  a  direct
395: measurement of solar  core metalicity might be important  to our understanding
396: of solar system formation.  If one  were to attempt to construct a solar model
397: that reproduces both a sound-speed profile consistent with helioseismology and
398: the \citet{ags05}  photospheric abundances,  that model  would likely
399: have  a convective  zone that  is  {\it depleted}  in metals  relative to  the
400: radiative core,  not elevated as  in the low-Z  model familiar from  the solar
401: neutrino puzzle.   It is possible  to envision a  a scenario where  this could
402: happen --  one that connects the  chemistry of the Sun's  convective zone with
403: that of the  planets.  First, there is clear  evidence that solar-system metal
404: differentiation  occurred, associated  with  the formation  of the  metal-rich
405: gaseous giants Jupiter, Saturn,  Neptune, and Uranus.  The gaseous atmospheres
406: of  Jupiter and  Saturn are  believed to  have been  formed by  accretion onto
407: relatively small ($\sim$ 10 M$_\oplus$)  rocky/icy cores, at a time when solar
408: formation is nearly complete  and the bulk of the gas in  the nebular disk has
409: dissipated.  These atmospheres are  established over $\sim$ 1-10 million years
410: \citep{bod}.  It is thus plausible that the gaseous planets were formed after
411: the  sun had  developed  a radiative  core  and an  isolated convective  zone.
412: Second,  the amount  of  chemical  differentiation in  the  gaseous giants  is
413: suggestive: modelers estimate that Jupiter's  total metal content is between 8
414: and 39 M$_\oplus$, or Z $\sim$  2.5-12.3\%, while that of Saturn is between 13
415: and 28  M$_\oplus$, or Z  $\sim$ 13.7-29.4\% \citep{saumon}.  The  excess metal
416: contained in  all four  gaseous giants, $\sim$  40-90 M$_\oplus$  depending on
417: modeling uncertainties  \citep{guillot}, is comparable to  the apparent deficit
418: of metal in the convective zone  ($\sim$ 50 M$_\oplus$), were one to associate
419: the GS abundances with the radiative zone (formed from primordial gas) and the
420: AGS abundances  with the convective zone.  The  late-forming gaseous envelopes
421: of Jupiter and Saturn could account for up to 40 M$_\oplus$ of this excess.
422: 
423: 
424: Thus it is possible that some mechanism operating in a chemically altered disk
425: -- perhaps proto-planets scouring out metal-rich dust grains that have settled
426: to the disk  midplane -- might result both in metal  enrichment of the gaseous
427: giants and  a reservoir  of metal-depleted gas  in the  circum-planetary disk.
428: Could some  portion of that  gas later be  deposited on the Sun,  reducing the
429: effective  Z of  the  convective zone?   This  question has  been raised  once
430: before, by \citet{castro}, who then explored helioseismology in
431: a two-zone  model motivated by  this possibility.  While  we are
432: not advocating for  such a scenario, simple estimates,  including those above,
433: do seem to support its plausibility.  Numerical calculations indicate that the
434: time  scale  for the  Sun  to accrete  gas,  influenced  gravitationally by  a
435: Jupiter-mass body orbiting at a distance  $\sim$ 5 AU, is relatively short, on
436: the order of $  \sim 5 \times 10^5$ yr \citep{strom}.  It  has also been noted,
437: based on the needed planetesimal  deposition rate and the tidal radius ($\sim$
438: 0.36 AU)  of a fully grown Jupiter,  that the planet would  have perturbed the
439: orbits of  about 2500  M$_\oplus$ of  gas, or 35\%  of the  mass of  the Sun's
440: present  convective  envelope  \citep{podolak}.   Thus  at least  some  of  the
441: conditions necessary for convective-zone dilution appear to be satisfied.
442: 
443: Though it is beyond the scope of  the present paper, it might be worthwhile to
444: pursue this  question further, assuming late-time  accretion of metal-depleted
445: gas (motivated by the gaseous giant metal reservoirs and their assumed time of
446: formation).   While  the  work by  \citet{castro}  is  a  first step  in  this
447: direction, if 
448: one takes the  accretion scenario seriously, then (depending  on the timing of
449: accretion  with respect to  the development  of the  radiative/convective zone
450: boundary) there should be a memory  of the accretion in the modern sun's upper
451: radiative zone --  a transition region between GS  interior abundances and AGS
452: surface abundances which depends on the volume and composition of the accreted
453: material.  Helioseismology would  thus become  a probe  of the  solar system's
454: late-stage  accretion  history.   This   history  would  be  linked  to  solar
455: development: in  the standard Hayashi-track description of  the proto-Sun, the
456: scenario would  only make sense if  the planet atmospheres were  formed in the
457: Henyey  phase  or  later.   However,  recent numerical  simulations  of  cloud
458: collapse and early solar evolution found that the convective envelope develops
459: earlier,  spans the  outer third  of the  proto-Sun by  radius,  and resembles
460: closely that of the modern  Sun \citep{wuchterl}.  This would extend the window
461: for dilution of  the convective zone to earlier  times.  Finally, the scenario
462: would predict chemical correlations between the planets and convective zone in
463: our sun and, perhaps, in other solar-like planetary systems.
464: 
465: The  various  threads  summarized  above, plus  additional  considerations  we
466: discuss  in this paper,  provide strong  motivation for  experiments measuring
467: CN-cycle neutrinos:
468: %
469: \begin{itemize}
470: 
471: \item A measurement of the CN  neutrino flux would provide an independent test
472: of solar metalicity, complementing photospheric determinations.
473: 
474: \item This measurement would test  the SSM postulate of a homogeneous zero-age
475: Sun, one of the assumptions important to helioseismology, the $^8$B neutrino flux,
476: and other SSM predictions that depend  on the metalicity of the Sun's interior
477: radiative zones.
478: 
479: \item It would  place constraints on metal accretion  that might have occurred
480: subsequent  to the Hayashi  phase, as  the pre-main-sequence  convective solar
481: core is established.
482: 
483: \item  The   current  solar  neutrino   program  has  helped   to  demonstrate
484: experimentally  that  the nuclear  astrophysics  foundations  of our  standard
485: theory of low-mass main-sequence  stellar evolution are valid.  Solar CN-cycle
486: neutrinos  provide our one  opportunity to  extend such  tests to  the nuclear
487: physics governing heavier main-sequence stars.
488: 
489: \item It is conceivable the a quantitative comparison of the Sun's surface and
490: interior metalicities might be important to more general problems of chemical
491: differentiation during solar-system formation.
492: 
493: \end{itemize}
494: %
495: 
496: 
497: \section{The Sun as a Calibrated Laboratory}
498: 
499: Independent  of questions  about  the Sun's  pre-main-sequence evolution,  one
500: recognizes that the Sun's inner core would have been mixed at the onset of the
501: main sequence due  to the initial out-of-equilibrium burning  of $^{12}$C.  It
502: has been  recognized for many years  that a measurement of  the CN-cycle solar
503: neutrino flux would, in principle, determine the metalicity of this core zone,
504: allowing  a comparison with  abundance determined  from the  solar atmosphere.
505: But in the past several years  new developments have occurred that now seem to
506: suggest such a measurement could be practical.  These include:
507: %
508: \begin{itemize}
509: 
510: \item  Accurate  calibrations  of  the  solar  core  temperature  by  SNO  and
511: Super-Kamiokande;
512: 
513: \item Tight constraints on the  oscillation parameters and matter effects that
514: will determine the flavor content of the CN and $^8$B neutrino fluxes;
515: 
516: \item  Recent  measurements of  the  controlling  reaction  of the  CN  cycle,
517: $^{14}$N(p,$\gamma$),  that  have significantly  reduced  the nuclear  physics
518: uncertainties affecting SSM predictions of CN-cycle fluxes; and
519: 
520: \item New ideas for high-counting  rate experiments that would be sensitive to
521: CN-cycle  neutrinos,  and from  which  reliable  terrestrial  fluxes could  be
522: extracted.
523: 
524: \end{itemize}
525: %
526: Our  analysis  uses  previous  SSM  work  in  which  the  logarithmic  partial
527: derivatives $\alpha(i,j)$  for each neutrino  flux $\phi_i$ are  evaluated for
528: the SSM input parameters $\beta_j$,
529: \begin{equation}
530: \alpha(i,j)  \equiv  {\partial   \ln{\left[  \phi_i/\phi_i(0)  \right]}  \over
531: \partial \ln{\left[ \beta_j / \beta_j(0)\right]}}
532: \end{equation}
533: where  $\phi_i(0)$  and  $\beta_j(0)$   denote  the  SSM  best  values.   This
534: information, in combination with  the assigned uncertainties in the $\beta_j$,
535: then  provides  an  estimate of  the  uncertainty  in  the SSM  prediction  of
536: $\phi_i$.   In particular,  crucial to  the current  analysis is  the  work of
537: \citet{bs05}, who evaluated the dependence on
538: the  mass  fractions  (measured  relative  to  hydrogen)  of  different  heavy
539: elements,
540: \begin{equation}
541: \beta_j =  \mathrm{mass~fraction~of~element~j \over mass~fraction~of~hydrogen}
542: \equiv X_j
543: \end{equation}
544: Having this information  not as a function of the  overall metalicity $Z$, but
545: as  a  function  of the  individual  abundances,  allows  us to  separate  the
546: ``environmental'' effects  of the  metals in the  solar core from  the special
547: role of  primordial C and N as  catalysts for the CN  cycle.  By environmental
548: effects  we mean  the influence  of the  metals on  the opacity  and  thus the
549: ambient  core  temperature, which  controls  the  rates of  neutrino-producing
550: reactions of both the pp-chain and CN cycle.  Simply put, our strategy here is
551: to  use  the  temperature-dependent  $^8$B  neutrino  flux  to  calibrate  the
552: environmental  effects  of  the  metals  and of  other  SSM  parameters,  thus
553: isolating the  special CN-cycle  dependence on primordial  C+N.  We  find this
554: primordial abundance can  be expressed, with very little  residual solar model
555: uncertainty, in  terms of the measured  $^8$B neutrino flux  and nuclear cross
556: sections that have been determined in  the laboratory.  In fact, we argue that
557: the  resulting expression is  likely more  general than  the SSM  context from
558: which it is derived.
559: 
560: The  partial derivatives  allow one  to define  the power-law  dependencies of
561: neutrino fluxes, relative to the SSM best-value prediction $\phi_i(0)$
562: \begin{equation}
563: \phi_i   =   \phi_i(0)  \prod_{j=1}^N   \left[   {\beta_j  \over   \beta_j(0)}
564: \right]^{\alpha(i,j)}
565: \label{eq:prod}
566: \end{equation}
567: where the product extends over  $N$ SSM input parameters.  This expression can
568: be  used  to  evaluate  how  SSM  flux  predictions  will  vary,  relative  to
569: $\phi_i(0)$, as the  $\beta_j$ are varied.  Alternatively, the  process can be
570: inverted:  a flux  measurement  could in  principle  be used  to constrain  an
571: uncertain input parameter.
572: 
573: The baseline SSM calculation for our calculations is BPS08(AGS) \citep{bps08},
574: which uses the recently determined AGS
575: abundances for  the volatile  elements C, N,  O, Ne,  and Ar, rather  than the
576: previous  GS standard  composition. It  should be  noted that  AGS  includes a
577: downward revision by 0.05 dex of  the Si photospheric abundance compared to GS
578: and,  accordingly, a  similar  reduction in  the  meteoritic abundances.   The
579: partial derivatives needed in the present calculation are summarized in
580: Tables \ref{table:one} (solar model parameters and nuclear cross sections) and
581: \ref{table:two} (abundances).
582: 
583: The SSM  estimate  of  uncertainties  in the  various  solar
584: neutrino fluxes 
585: $\phi_i$  can  be  obtained  by  folding  the  partial  derivatives  with  the
586: uncertainties in the underlying $\beta_j$.  In particular, it is convenient to
587: decompose Eq.  \ref{eq:prod} into its  dependence on solar  parameters, non-CN
588: metals, nuclear S-factors, and the primordial C and N abundances,
589: %
590: \begin{eqnarray}
591: \phi_i =  \phi_i^{SSM} \times~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ \nonumber
592: \\ \left(  \prod_{j \in \mathrm{\{Solar\}}} \left[  {\beta_j \over \beta_j(0)}
593: \right]^{\alpha(i,j)}  \prod_{j  \in   \mathrm{\{Metals  \neq  C,N\}}}  \left[
594: {\beta_j \over  \beta_j(0)} \right]^{\alpha(i,j)} \right)  \nonumber \\ \times
595: \prod_{j   \in  \mathrm{\{Nuclear\}}}   \left[   {\beta_j  \over   \beta_j(0)}
596: \right]^{\alpha(i,j)}  \prod_{j \in  \mathrm{\{C,N\}}}  \left[ {\beta_j  \over
597: \beta_j(0)} \right]^{\alpha(i,j)} .~~~~~~~
598: \label{eq:prod2}
599: \end{eqnarray}
600: 
601: The  two  terms  within  the  brackets will  be  designated  ``environmental''
602: uncertainties --  SSM solar and abundance parameters  that primarily influence
603: neutrino flux predictions through changes they induce in the core temperature.
604: These are, respectively, the uncertainties in the photon luminosity $L_\odot$,
605: the  mean  radiative opacity,  the  solar age,  and  calculated  He and  metal
606: duffusion; and the fractional abundances of O, Ne, Mg, Si, S, Ar, and Fe.  The
607: estimated  1$\sigma$ fractional  uncertainties for  the solar  parameters have
608: been previously evaluated and are listed in Table \ref{table:three}.
609: 
610: The heavy elements abundances in
611: BPS08(AGS) are taken from  the meteoritic abundances where available (Mg,
612: Si, S,  and Fe) and otherwise  from photospheric abundances  (for the volatile
613: elements   C,   N,  O,   Ne,   Ar).   As   mentioned  before,   the   assigned
614: historical/conservative 1$\sigma$ fractional 
615: uncertainties  shown in  Table  \ref{table:four} were  defined empirically  in
616: \citet{bs05} by
617: %
618: \begin{equation}
619: {\Delta \beta_i \over  \beta_i} = \left| {\mathrm{Abundance}_i^{\mathrm{GS}} -
620: \mathrm{Abundance}_i^{\mathrm{AGS}}  \over (\mathrm{Abundance}_i^{\mathrm{GS}}
621: + \mathrm{Abundance}_i^{\mathrm{AGS}})/2)} \right|.
622: \end{equation}
623: This definition generates the uncertainties shown in Table \ref{table:four}.
624: 
625: The next term  contains the effects of nuclear  cross section uncertainties on
626: flux predictions.  The $\beta_j$ are  the S-factors for p+p ($S_{11}$), $^3$He
627: +  $^3$He ($S_{33}$),  $^3$He+$^4$He ($S_{34}$),  p +  $^7$Be ($S_{17}$),  e +
628: $^7$Be ($S_{e7}$),  and p +  $^{14}$N ($S_{114}$).  Their  estimated 1$\sigma$
629: fractional  uncertainties, which  we discuss  below, are  also shown  in Table
630: \ref{table:three}.
631: 
632: The last  term is the contribution of  the primordial C and  N abundances.  As
633: Table   \ref{table:two}  shows,  pp-chain   neutrino  fluxes   are  relatively
634: insensitive to variations  in these abundances, as the  heavier nuclei like Fe
635: have a more important influence on the core opacity.  But the expected, nearly
636: linear  response  of  the  $^{13}$N  and $^{15}$O  neutrino  fluxes  to  these
637: abundances is apparent.   These are the abundances we  would like to constrain
638: by a  future measurement of the  $^{13}$N and $^{15}$O  solar neutrino fluxes.
639: Such  a measurement  begins to  be of  interest if  these abundances  could be
640: determined with an accuracy significantly better than 30\%.  Note that the C/N
641: abundance term in Eq. (\ref{eq:prod2}) for $\phi(^{13}\mathrm{N})$
642: \begin{eqnarray}
643:  \prod_{j   \in   \mathrm{\{C,N\}}}    \left[   {\beta_j   \over   \beta_j(0)}
644:  \right]^{\alpha(i,j)}=  ~~~~~~~~~~~~~\nonumber \\  \left( {X(^{12}\mathrm{C})
645:  \over  X(^{12}\mathrm{C})_{SSM}}  \right)^{0.874} \left(  {X(^{14}\mathrm{N})
646:  \over X(^{14}\mathrm{N})_{SSM}} \right)^{0.142}
647: \end{eqnarray}
648: is not quite linear in the overall C+N abundance.  An overall scaling of
649: primordial C  and N, $X/X_{SSM} \rightarrow$ 1+$\delta$  yields the dependence
650: $(1+\delta)^{1.016}$, so 2\%  steeper than linear.  In addition  to the direct
651: dependence  of the  CN-cycle on  C and  N, increasing  the C+N  abundance also
652: increases the opacity and core temperature, to which the CN neutrino flux also
653: responds.  
654: 
655: Were  one  to vary  the  11  SSM  parameters designated  as  ``environmental''
656: according to their assigned uncertainties (taking them to be uncorrelated), an
657: 7.5\% SSM  net uncertainty in $\phi(^{13}\mathrm{N})$ would  be obtained.  But
658: we can do better than this by exploiting SNO and Super-Kamiokande measurements
659: of the $^8$B  neutrino flux, a ``thermometer'' that is  even more sensitive to
660: the solar core environment than the CN neutrinos.  Below we discuss the use of
661: SNO in  this way --  the arguments are  simpler for this detector,  because it
662: provides a measurement  of the $^8$B neutrino flux  independent of flavor.  In
663: the next section we use Super-Kamiokande, the case of most interest because it
664: exploits the  same reaction, elastic  scattering, as the  proposed CN-neutrino
665: detectors, allowing some common errors  to cancel: the SNO results then become
666: crucial input, helping to constrain the effects of neutrino oscillations.
667: 
668: 
669: One can express $\phi(^{13}\mathrm{N})$
670: uncertainties in terms of $\phi(^8\mathrm{B})$, while minimizing the residual
671: solar environmental error,  i.e. by minimizing the contribution  of the factor
672: in parenthesis in the expression
673: %
674: \begin{eqnarray}
675: &&{\phi     (^{13}\mathrm{N})     \over     \phi^{SSM}(^{13}\mathrm{N})}     =
676: \left[{\phi(^{8}\mathrm{B})                                               \over
677:     \phi^{SSM}(^{8}\mathrm{B})}\right]^{K_{(13,8)}} 
678: \nonumber \\ &\times& \left(  \prod_{j \in \mathrm{\{Solar\}}} \left[ {\beta_j
679: \over  \beta_j(0)}  \right]^{\gamma_{j}}  \prod_{j \in  \mathrm{\{Metals  \neq
680: C,N\}}}  \left[   {\beta_j  \over  \beta_j(0)}   \right]^{\gamma_{j}}  \right)
681: \nonumber \\ &\times& \prod_{j \in \mathrm{\{Nuclear\}}} \left[ {\beta_j \over
682: \beta_j(0)}   \right]^{\gamma_{j}}  \prod_{j   \in   \mathrm{\{C,N\}}}  \left[
683: {\beta_j \over \beta_j(0)} \right]^{\gamma_{j}}
684: \label{eq:prod3}
685: \end{eqnarray}
686: where
687: \begin{equation}
688: \gamma_{j}       \equiv      \alpha(^{13}\mathrm{N},j)       -      K_{(13,8)}
689: \alpha(^8\mathrm{B},j)
690: \end{equation}
691: by a suitable choice of the constant $K_{(13,8)}$. 
692: Using   the   SSM   logarithmic   derivatives  $\alpha(i,j)$   and   parameter
693: uncertainties         $\Delta        \beta_j/\beta_j$         of        Tables
694: \ref{table:one}-\ref{table:four},  we find $K_{(13,8)}=  0.608$. To  check the
695: consistency of this  procedure, we have performed a  Monte Carlo simulation of
696: solar  models   where  the  11   environmental  input  quantities   have  been
697: varied   simultaneously.    We   find   a  tight   correlation   between   the
698: $\phi(^{13}{\rm N})$
699: and  $\phi(^{8}{\rm  B})$  fluxes.  In Figure~\ref{fig:correl}  we  show  this
700: correlation on the  top-left panel and the linear fit to  the data, from which
701: we find $K_{(13,8)}= 0.599$, the
702: value we adopt  for this paper, very close to that  derived from the power-law
703: exponents. The top-right panel shows the residuals of the fit and its standard
704: deviation $\sigma=2.8\%$.  We note here that the bulk of the dispersion in the
705: $\phi(^{8}{\rm B})-\phi(^{13}{\rm  N})$ correlation is due  to the uncertainty
706: in the 
707: diffusion  rate.   This  can  be  understood  as  follows.  All  environmental
708: quantities 
709: affect  these  neutrino fluxes  by  modifying  the  temperature in  the  solar
710: core. Diffusion has,  however, the additional effect of  increasing the number
711: of CN nuclei  in the core, leading to a  directly proportional increase in
712: $\phi(^{13}{\rm 
713: N})$   but   not   in   $\phi(^{8}{\rm    B})$.    This   can   be   seen   in
714: Tables~\ref{table:one} and 
715: \ref{table:two},   where  all  the   power-law  exponents   for  environmental
716: quantities are larger for $\phi(^{8}{\rm B})$ than for $\phi(^{13}{\rm N})$ (a
717: natural    consequence    of   the    larger    temperature   dependence    of
718: $\phi(^{8}{\rm   B})$).   The   only   exception  is   diffusion,   on   which
719: $\phi(^{13}{\rm N})$ 
720: shows  a stronger  dependence than  $\phi(^{8}{\rm  B})$. Were  we to  exclude
721: diffusion as a source of 
722: uncertainty, the dispersion in the $\phi(^{8}{\rm B})-\phi(^{13}{\rm N})$
723: correlation would only be $\sim 0.5\%$.
724: 
725: 
726: Expressing $\phi(^{13}{\rm N})$ in the form of Eq.~(\ref{eq:prod3})
727: has two advantages.  First, as we detail below, it 
728: reduces the  overall theoretical uncertainty  in the relationship  between the
729: primordial C and N abundances and the $^{13}$N neutrino flux.  Second, this 
730: relationship should  be more  general than  the SSM context  from which  it is
731: derived:  the correlation  between the  various neutrino  fluxes  $\phi_i$ and
732: $T_c$ has been demonstrated to hold  even when SSM parameters have been varied
733: far outside their accepted SSM uncertainties.  
734: % Figure \ref{fig:castellani},
735: % taken from \citet{castellani}, is one well-known 
736: % illustration of this. 
737: 
738: 
739: A   simple  way   to  fix   the  first   term  on   the  right-hand   side  of
740: Eq.  (\ref{eq:prod3}) is  by  using the  SNO  measurement of  the total  $^8$B
741: neutrino  flux, thereby  eliminating many  SSM and  LMA  oscillation parameter
742: uncertainties in  terms of  a measured quantity.   As the SNO  statistical and
743: systematic errors  combined in quadrature  give an uncertainty 9.4\%,  the net
744: uncertainty  in  this  term   is  then  5.6\%.   The  remaining  environmental
745: uncertainty is  encoded in  the bracketed terms  in Eq.  (\ref{eq:prod3}), the
746: deviations in the dependence of the  $^{13}$N and $^8$B neutrino fluxes from a
747: naive $T_c$ power law.  From our  Monte Carlo simulation we find this residual
748: uncertainty is 2.8\%, and thus
749: quite small in comparison to the SNO uncertainty: the total environmental
750: uncertainty  is $\sim$  6.3\%.  The  use of  the SNO  result to  constrain the
751: environmental  uncertainty  thus reduces  this  uncertainty,  relative to  the
752: result obtain  previously by direct  variation of SSM input  parameters within
753: their assigned SSM uncertainties.
754: 
755: The remaining uncertainty arising in the evaluation of Eq. (\ref{eq:prod3}) is
756: the nuclear factor which, given that the expression involves both the $^{14}$N
757: and $^8$B fluxes, depends on a combination of pp-chain and CN-cycle S-factors.
758: One finds from Tables \ref{table:one}- \ref{table:four} that the uncertainties
759: are dominated by $S_{17}$, $S_{34}$,  which controls the pp-chain branching to
760: the ppII and ppIII cycles, and $S_{114}$.  One of the reasons that the CN
761: neutrino fluxes are potentially a quantitative probe of the Sun's primordial C
762: and  N  are  recent  improvements  particularly  in  determinations  of  the
763: last two $S$-factors.  
764: 
765: The traditional SSM value for $S_{34}$ is based on the 1998 evaluation of 
766: \citet{adelberger}, 0.53 $\pm$ 0.05 keV b.  The relatively large error
767: bar  on  the recommended  value  reflected  apparent systematic  disagreements
768: between  experiments detecting prompt  $\gamma$ rays  and counting  the $^7$Be
769: activity.  Since this evaluation new, high-statics measurements have been made
770: by  a Weizmann  Institute  group (\citealt{singh},  an
771: activity  measurement),  by the  LUNA  collaboration (\citealt{confortola},  a
772: combination of prompt $\gamma$ and activity measurements, 
773: as well as \citealt{gyurky}, an activity measurement), and
774: by  the  Seattle   group  (\citealt{brown},  an  activity
775: measurement).   If these  data are  extrapolated  to threshold  with the  same
776: theoretical fitting function,  one finds values of S(0)  of 0.546 $\pm$ 0.020,
777: 0.560  $\pm$  0.017,  0.545  $\pm$   0.017,  and  0.595  $\pm$  0.018  keV  b,
778: respectively.  The spread  of these  results is  somewhat larger
779: than is expected  on the basis of the uncertainties.   They have been combined
780: by Snover (private communication) in a way that takes into account this spread
781: by 
782: inflating  the errors  according  to the  associated  chi-squares, while  also
783: accounting  for possible  correlations  between the  LUNA  activity and  total
784: results  \citep{gyurky,confortola}.  The  result is  0.564 $\pm$  0.020  keV b.
785: There is an additional theoretical uncertainty associated with the theoretical
786: fitting function that is used to extrapolate these data to threshold to obtain
787: S(0).  As  discussed in \citet{brown}, generally  this variation is  at the few
788: percent  level, though  it can  reach higher  values if  fits are  required to
789: reproduce higher-energy data.  The procedure  we followed is based on the LUNA
790: group's  work on its  activation data  \citep{gyurky}, where  three theoretical
791: models  (\citealt{kajino,csoto}  and  \citealt{descouvemount})  were  used  to
792: derived  an extrapolated zero-energy  result.  The  resulting best  values for
793: $S(0)$ have a range of $\pm$ .019  keV b, or $\pm$ 3.4\%.  Thus we adopt 3.4\%
794: as  a theoretical extrapolation  uncertainty, which  we combine  in quadrature
795: with the Snover recommendation to obtain  a final result of 0.564 $\pm$ 0.028.
796: This corresponds to a 4.9\% uncertainty, which can be compared to the 9.4\%
797: uncertainty recommended in the \citet{adelberger} evaluation.
798: 
799: The most important  nuclear physics uncertainty in the  analysis is $S(0)$ for
800: $^{14}$N(p,$\gamma$), a reaction  that has been the subject  of recent intense
801: study.    The  $S_{114}$   value  and   the  8.4\%   uncertainty   adopted  in
802: \citet{bsb_mc}, 1.69 $\pm$ 0.14 keV b, was 
803: obtained from combining  the LUNA results of \citet{formicola},  1.7 $\pm$ 0.2
804: keV b (determined from data taken at or above 
805: center-of-mass energies of 140 keV),  with the TUNL results of \citet{runkle},
806: 1.68 $\pm$ 0.09 (stat) $\pm$ 0.16 (sys) keV b.  
807: Subsequently a  series of  measurements have  been done at  LUNA in  which the
808: cross  section  was measured  to  center-of-mass energies  as  low  as 70  keV
809: \citep{imbriani,lemut,bemmerer,traut}.   In particular,  \citet{imbriani} give
810: $S(0)$ = 1.61 $\pm$ 0.08 keV b, corresponding to a 
811: 5\% error, based on data  obtained for center-of-mass energies between 119 and
812: 367 keV.  We  use the Imbriani value for $S(0)$ in  this paper, a conservative
813: choice given that this fit was  made prior to the extension of measurements to
814: 70 keV.   Furthermore, work is underway  on the energy range  above the lowest
815: resonance, $\sim$ 300-400 keV, a  region which limits the interference pattern
816: analysis, and on improved  r-matrix analyses (Wiescher, private communication)
817: which take into account all reaction channels.  Thus we expect a more definite
818: analysis of $S(0)$ and of  its experimental and theoretical uncertainties will
819: be available soon. 
820: 
821: All of this can then be summarized in the theoretical relationship:
822: \begin{eqnarray}
823: &&{\phi     (^{13}\mathrm{N})     \over     \phi^{SSM}(^{13}\mathrm{N})}     =
824: \left[{\phi^{SNO}(^{8}\mathrm{B})                                         \over
825: \phi^{SSM}(^{8}\mathrm{B})}\right]^{0.599} \nonumber \\  &\times& \left[ 1 \pm
826: 2.8\%   (\mathrm{resid.~environ.})  \pm   5.0\%)   (\mathrm{nuclear})  \right]
827: \nonumber      \\      &\times&      \left(     {X(^{12}\mathrm{C})      \over
828: X(^{12}\mathrm{C})_{SSM}}  \right)^{0.858}  \left(  {X(^{14}\mathrm{N})  \over
829: X(^{14}\mathrm{N})_{SSM}} \right)^{0.141}.
830: \label{eq:prod4}
831: \end{eqnarray}
832: The first term,  given the SNO total flux combined error  of $\sim$ 9.4\%, has
833: an uncertainty of 6.4\%.  Thus one  consequence of the recent work on pp-chain
834: and CN-cycle S-factors  is the reduction of the  nuclear physics uncertainties
835: to the  level of  the SNO measurements.   The overall uncertainty,  adding the
836: SNO,  residual  environmental, and  nuclear  uncertainties  in quadrature,  is
837: 8.6\%,  small compared  to the  conservative uncertainty  assigned to  the AGS
838: abundances of $\sim$ 30\%.
839: 
840: Under  an overall  scaling  of  primordial C  and  N, $X/X_{SSM}  \rightarrow$
841: 1+$\delta$, the quantity being constrained responds as
842: \begin{equation}
843: (1+\delta)^{0.999} \sim 1 + \delta
844: \label{eq:delta2}
845: \end{equation}
846: Because  we  have removed  the  ``environmental''  effects  of all  metals  in
847: Eq.   (\ref{eq:prod4}),   we   find   the  expected,   nearly   exact   linear
848: proportionality  between the primordial  metals C  and N  and the  CN neutrino
849: flux.
850: 
851: The arguments can be repeated for the $^{15}$O flux, the more interesting case
852: experimentally because of its higher endpoint energy.  We find
853: \begin{eqnarray}
854: &&{\phi     (^{15}\mathrm{O})     \over     \phi^{SSM}(^{15}\mathrm{O})}     =
855: \left[{\phi(^{8}\mathrm{B})  \over  \phi^{SSM}(^{8}\mathrm{B})}\right]^{0.828}
856: \nonumber \\ &\times& \left(  \prod_{j \in \mathrm{\{Solar\}}} \left[ {\beta_j
857: \over  \beta_j(0)}  \right]^{\gamma_{j}}  \prod_{j \in  \mathrm{\{Metals  \neq
858: C,N\}}}  \left[   {\beta_j  \over  \beta_j(0)}   \right]^{\gamma_{j}}  \right)
859: \nonumber \\ &\times& \prod_{j \in \mathrm{\{Nuclear\}}} \left[ {\beta_j \over
860: \beta_j(0)}   \right]^{\gamma_{j}}  \prod_{j   \in   \mathrm{\{C,N\}}}  \left[
861: {\beta_j \over \beta_j(0)} \right]^{\gamma_{j}}
862: \label{eq:prod3a}
863: \end{eqnarray}
864: where
865: \begin{equation}
866: \gamma_{j} \equiv \alpha(^{15}\mathrm{O},j) - 0.828 \alpha(^8\mathrm{B},j).
867: \end{equation}
868: The  lower two panels  in Figure~\ref{fig:correl}  show the  tight correlation
869: between the  $\phi (^{15}\mathrm{O})$  and $\phi (^{8}\mathrm{B})$  fluxes and
870: the residuals of the linear fit. 
871: 
872: Evaluating associated parameter uncertainties as before, one finds
873: \begin{eqnarray}
874: &&{\phi     (^{15}\mathrm{O})     \over     \phi^{SSM}(^{15}\mathrm{O})}     =
875: \left[{\phi^{SNO}(^{8}\mathrm{B})                                         \over
876: \phi^{SSM}(^{8}\mathrm{B})}\right]^{0.828} \nonumber \\  &\times& \left[ 1 \pm
877: 2.6\%   (\mathrm{resid.~environ.})   \pm   7.1\%  (\mathrm{nuclear})   \right]
878: \nonumber      \\      &\times&      \left(     {X(^{12}\mathrm{C})      \over
879: X(^{12}\mathrm{C})_{SSM}}  \right)^{0.805}  \left(  {X(^{14}\mathrm{N})  \over
880: X(^{14}\mathrm{N})_{SSM}} \right)^{0.199}
881: \label{eq:prod4b}
882: \end{eqnarray}
883: The larger  environmental parameter, 0.828,  is expected because  the $^{15}$O
884: neutrino  flux  has  a  somewhat   steeper  dependence  on  the  average  core
885: temperature than the $^{13}$N flux.  This increases the errors associated with
886: SNO  uncertainties (7.8\%)  and  nuclear cross  sections  (the uncertainty  in
887: $S_{34}$ and other pp-chain cross sections determines the quality of our $^8$B
888: neutrino thermometer).   Thus the overall uncertainty  in this ``theoretical''
889: relation  between the $^{15}$O  neutrino flux  and the  core C/N  abundance is
890: 10.8\%.   As  before, if  one  considers  a scaling  of  primordial  C and  N,
891: $X/X_{SSM} \rightarrow$ 1+$\delta$, the last term becomes
892: \begin{equation}
893: (1+\delta)^{1.004} \sim 1 + \delta
894: \label{eq:delta3}
895: \end{equation}
896: showing the  linear relationship  between the $^{15}$O  neutrino flux  and the
897: primordial C+N abundance.
898: 
899: Now we discuss a somewhat  more detailed (and improved) analysis that exploits
900: similarities between  Super-Kamiokande and  future CN neutrino  detectors, and
901: allows us  to fold in  what we have  learned about neutrino  oscillations from
902: terrestrial experiments like KamLAND.
903: 
904: 
905: \section{The Analysis for Elastic Scattering and Neutrino Oscillations}
906: 
907: Equations (\ref{eq:prod3}) and  (\ref{eq:prod3a}) give relationships among the
908: primordial  core  C  and  N  abundances, other  SSM  uncertainities,  and  the
909: instantaneously produced CN-cycle and $^8$B neutrino fluxes.  These equations,
910: attractive because  of their simplicity, are somewhat  idealized, because they
911: do not address  how the left-hand sides of these  equations will be determined
912: experimentally.  For example, neutrino  oscillations during the transit to the
913: earth will  alter the  flavors of these  neutrino in an  energy-dependent way,
914: influencing detector  responses.  Fluxes  determined in most  experiments will
915: have  to be corrected  for such  effects, including  the uncertainties  in the
916: neutrino  mass  difference  $\delta   m_{12}^2$  and  mixing  angle  $\sin^2{2
917: \theta_{12}}$.  In  the discussion  of the previous  section, we  avoided this
918: issue  in  evaluating  the   right-hand  sides  of  Eq.  (\ref{eq:prod3})  and
919: (\ref{eq:prod3a}) by  employing the  SNO result for  the total  $^8$B neutrino
920: flux.  But the  first results on the CN-cycle fluxes,  needed on the left-hand
921: sides of Eqs. (\ref{eq:prod3}) and  (\ref{eq:prod3a}), are most likely to come
922: from  $\nu$-e   inelastic  scattering  experiments,   where  $\sigma(\nu_\mu)/
923: \sigma(\nu_e) \sim$ 0.15.  Thus to  derive the instantaneous (solar) values of
924: these fluxes, one would have to  correct the detector response for the effects
925: of flavor  mixing.  Oscillations  are one of  several uncertainties  that will
926: produce correlated responses in both $^8$B and CN neutrino detectors.  Thus we
927: need an  analysis that accounts for  such correlations: it  is advantageous to
928: develop  this analysis,  as one  can then  make use  of  another statistically
929: powerful experiment (Super-Kamiokande) while supplementing SNO data with other
930: constraints on flavor mixing (e.g., KamLAND).
931: 
932: Below  we compare Super-Kamiokande  and Borexino/SNO+  rates, which  exploit a
933: common  detection mechanism, $\nu_x$-e  elastic scattering,  making correlated
934: errors  easier to identify.   Super-Kamiokande is  potentially the  best solar
935: thermometer because of its statistical precision.
936: 
937: In this approach we express the total $^8$B flux (the instantaneous solar flux
938: needed in Eqs. (\ref{eq:prod3}) and (\ref{eq:prod3a})) as
939: \begin{eqnarray}
940: {\phi(^8\mathrm{B})   \over  \phi^{SSM}(^8\mathrm{B})}&=&  {\phi(^8\mathrm{B})
941: \langle  \sigma^{SK}(^8\mathrm{B}, \delta m_{12}^2,\theta_{12})  \rangle \over
942: \phi^{SSM}(^8\mathrm{B})      \langle     \sigma^{SK}(^8\mathrm{B},     \delta
943: m_{12}^2,\theta_{12})      \rangle      }      \nonumber      \\      &\equiv&
944: {R^{SK}_{exp}(^8\mathrm{B})     \over     R^{SK}_{cal}(^8\mathrm{B},    \delta
945: m_{12}^2,\theta_{12})}
946: \label{eq:fluxratio}
947: \end{eqnarray}
948: Here $\langle  \sigma^{SK} \rangle$ is  an effective cross section  that takes
949: into  account  all of  the  neutrino  flavor  and detector  response  (trigger
950: efficiencies,  resolution,  cross  section  uncertainties, etc.)  issues  that
951: determine  the   relationship  between  a  measured  detector   rate  and  the
952: instantaneous  solar flux.   The numerator  of  the ratio  on the  right is  a
953: directly   measured  experimental   quantity:  the   Super-Kamiokande  elastic
954: scattering rate for producing  recoil electrons with apparent energies between
955: 5.0  and  20 MeV,  per  target  electron per  second.   The  denominator is  a
956: conversion  factor  that  relates   the  instantaneous  $\nu_e$  flux  to  the
957: experimental  rate:  the  cross  section  for  $\nu_x-e$  elastic  scattering,
958: averaged  over   a  normalized  $^8$B  spectrum,  defined   for  the  specific
959: experimental  conditions of  Super-Kamiokande,  and including  the effects  of
960: flavor mixing.   This conversion factor is essentially  a laboratory quantity:
961: it can be calculated from  laboratory measurements of detector properties, the
962: $\beta$ decay  spectrum, the underlying neutrino-electron  cross sections, and
963: most  critically, the  parameters governing  oscillations.  We  describe these
964: factors below.
965: 
966: The  experimental   rate  comes  from   the  1496  days  of   measurements  of
967: Super-Kamiokande I \citep{hosaka}.  From the SK I rate/kiloton/year
968: \begin{equation}
969: 520.1        \pm         5.3        \mathrm{(stat)}        ~{}^{+18.2}_{-16.6}
970:     \mathrm{(sys)}~\mathrm{kton^{-1}~y^{-1}}.
971: \label{eq:rate1}
972: \end{equation}
973: we find $R^{SK}_{exp}(^8\mathrm{B})$,
974: \begin{equation}
975: 4.935    \pm    0.05   \mathrm{(stat)}~{}^{+0.17}_{-0.16}\mathrm{(sys)}~\times
976:   10^{-38}~\mathrm{{electron}^{-1} s^{-1}}
977: \end{equation}
978: (or  $\sim$ 0.049  Solar Neutrino  Units, or  SNUs).  The  dominant systematic
979: error  includes  estimates  for  the  energy  scale  and  resolution,  trigger
980: efficiency, reduction, spallation dead time,  the gamma ray cut, vertex shift,
981: background  shape  for  signal  reduction, angular  resolution,  and  lifetime
982: uncertainties.  The  combined statistical and  systematic error is  $\sim \pm$
983: 3.6\%.
984: 
985: To evaluate the denominator in  Eq. (\ref{eq:fluxratio}) we need the suitably
986: averaged cross section, defined for the window used by the SK I collaboration,
987: \begin{eqnarray}
988:  \langle \sigma^{SK}(^8\mathrm{B},  \delta m_{12}^2,\theta_{12}) \rangle= \int
989: d  E_\nu  \phi_{norm}^{^8\mathrm{B}}(E_\nu)~~~~~~~~~~~  \nonumber &&\\  \times
990: \left[       P_{\nu_e}(E_\nu,\delta       m^2_{12},\sin^2{2      \theta_{12}})
991: \int_{T=0}^{T^{max}(E_\nu)}   dT    ~\sigma_{\nu_e}^{es}(T)   \right.   ~~~~&&
992: \nonumber    \\   \left.    +    ~P_{\nu_\mu}(E_\nu,\delta   m^2_{12},\sin^2{2
993: \theta_{12}}) \int_{T=0}^{T^{max}(E_\nu)} dT ~\sigma_{\nu_\mu}^{es}(T) \right]
994: ~~~&&   \nonumber  \\  \times   \int_{5.0~\mathrm{MeV}}^{20.0~\mathrm{MeV}}  d
995: \epsilon_a                                     f_{\mathrm{trigger}}(\epsilon_a)
996: \rho(\epsilon_a,\epsilon_t=T+m_e)~~~~~~
997: \label{eq:theorypart}
998: \end{eqnarray}
999: where  $\phi_{norm}^{^8\mathrm{B}}(E_\nu)$ is  the  normalized $^8$B  neutrino
1000: spectrum.   Equation  (\ref{eq:theorypart})  involves  an  integral  over  the
1001: product    of   this   spectrum    and   the    energy-dependent   oscillation
1002: probabilities.  ($P_{\nu_e}+P_{\nu_\mu}$=1, assuming oscillations  into active
1003: flavors.  $P_{\nu_\mu}$ can be defined as the oscillation probability to heavy
1004: flavors, if  the effects  of three flavors  are considered.)  A  given $E_\nu$
1005: fixes the range of kinetic energies  $T$ of the scattered electron, over which
1006: an integration  is done; in the  laboratory frame $T^{max}  = 2 E_\nu^2/(m_e+2
1007: E_\nu)$.   The  integrand  includes  the  elastic  scattering  cross  sections
1008: $\sigma^{es}(T)$   for   electron   and   heavy-flavor   neutrinos   and   the
1009: Super-Kamiokande  resolution   function  $\rho(\epsilon_a,\epsilon_t)$,  where
1010: $\epsilon_t=T+m_e$  is the true  total electron  energy while  $\epsilon_a$ is
1011: apparent energy, as deduced from the number of phototube hits in the detector.
1012: Finally,   an  integral   must  be   done  over   the  window   used   by  the
1013: experimentalists, apparent electron energies $\epsilon_a$ between 5.0 and 20.0
1014: MeV.  The  deduced counting  rate includes the  triggering probability  that a
1015: event of apparent energy $\epsilon_a$ will be recorded in the detector.
1016:         
1017: As the  uncertainties associated  with triggering efficiencies,  energy scale,
1018: and resolution are already  incorporated in the deduced Super-Kamiokande event
1019: rate (Eq.~\ref{eq:rate1}),  we are free to use  best-value functions in such
1020: an  analysis.   Here  we employ  simple  fits  to  the measurements  given  in
1021: \citet{hosaka}, a Gaussian resolution function
1022: \begin{equation}
1023: \rho(\epsilon_a,\epsilon_t)={1  \over  \sqrt{2  \pi} \sigma(\epsilon_t)}  \exp
1024: \left[ -{(\epsilon_t-\epsilon_a)^2 \over 2 \sigma(\epsilon_t)^2} \right]
1025: \label{eq:resolution}
1026: \end{equation}
1027: where $\sigma(\epsilon_t) \sim 0.326  \epsilon_t^{0.642}$, or about 14\% at 10
1028: Mev; and a relatively sharp trigger efficiency
1029: \begin{equation}
1030: f_{\mathrm{trigger}}(\epsilon_a)   =    {1   \over   2}   +    {1   \over   2}
1031: \tanh^5{\left[{\epsilon_a-\bar{\epsilon}_a \over \sigma'}\right]}
1032: \end{equation}
1033: where $\bar{\epsilon}_a \sim$ 3.6 MeV and $\sigma' \sim$ 0.172 MeV.
1034: 
1035: The   expression  for   the   CN-cycle  neutrino   response   is  similar   to
1036: Eq. (\ref{eq:fluxratio})
1037: \begin{eqnarray}
1038: {\phi(^{15}\mathrm{O})       \over       \phi^{SSM}(^{15}\mathrm{O})}      &=&
1039: {R_{exp}^{B/S}(^{15}\mathrm{O})   \over   \phi^{SSM}(^{15}\mathrm{O})  \langle
1040: \sigma^{B/S}(^{15}\mathrm{O}, \delta  m_{12}^2,\theta_{12}) \rangle} \nonumber
1041: \\      &\equiv&     {      ~R^{B/S}_{exp}(\mathrm{^{15}\mathrm{O}})     \over
1042: R^{B/S}_{cal}(^{15}\mathrm{O}, \delta  m_{12}^2,\theta_{12})} \nonumber \\ &=&
1043: {R_{exp}^{B/S}(\mathrm{CN})/(1             +\alpha(0.8,1.3))             \over
1044: R^{B/S}_{cal}(^{15}\mathrm{O}, \delta m_{12}^2,\theta_{12})) }
1045: \label{eq:fluxratio2}
1046: \end{eqnarray}
1047: Here the experimental rate for $^{15}$O neutrinos has been written in terms of
1048: the  total  CN-neutrino  rate  $R^{B/S}_{exp}(\mathrm{CN})$ by  introducing  a
1049: correction   factor    $\alpha$   discussed   below.     No   measurement   of
1050: $R^{B/S}_{exp}(\mathrm{CN})$   currently   exists,   of  course.    But   such
1051: measurements could be made in  Borexino or SNO+, existing or planned detectors
1052: that  will  use large  volumes  of  organic  scintillator, placed  quite  deep
1053: underground (we discuss these detectors in the concluding section).  A
1054: window  for the  apparent  kinetic energy  $T$  of the  scattered electron  of
1055: 0.8-1.3 MeV has been discussed by the Borexino group.  As the $^7$Be 0.866 MeV
1056: line  corresponds  to  $T^{max}  \sim$  0.668 MeV,  this  window  would  limit
1057: contamination from $^7$Be neutrino recoil electrons.  As Borexino has achieved
1058: a resolution of $\sim$  8.7\% at $T$ = 751 MeV and  $\sim$ 9.1\% at 0.825 MeV,
1059: we   take   (for  simulation   purposes)   a   Gaussian  resolution   function
1060: (Eq.~\ref{eq:resolution}) with
1061: \begin{equation}
1062: \sigma(T) \sim 0.08 \mathrm{MeV} \sqrt{T/\mathrm{MeV}}
1063: \end{equation}
1064: We  also  adopt  a  nominal  step-function  trigger,  $f_{\mathrm{trigger}}  =
1065: \theta(T-0.25 {\mathrm{MeV}})$, though the trigger does not influence rates in
1066: the high-energy window of interest for CN neutrinos.
1067: 
1068: The factor $\alpha(0.8,1.3) \sim$ 0.120  is the ratio of the measured $^{13}$N
1069: to  $^{15}$O  neutrino  rates  in  the observation  window,  a  correction  we
1070: introduce to  convert the total rate  to the rate for  $^{15}$O neutrinos.  In
1071: principle this  is a measurable quantity:  because of the  lower endpoint, the
1072: relative importance of $^{13}$N neutrinos  drops quickly with energy.  For the
1073: $^{15}$O neutrinos,  60\% of the events  would reside in  bins between 0.8-1.0
1074: MeV,  with the  remaining 40\%  between 1.0-1.3  MeV.  If  one looks  at total
1075: events ($^{13}$N and $^{15}$O),  $^{13}$N neutrinos are responsible for $\sim$
1076: 19\%  of the  events between  0.8-1.0  MeV, but  only 1.0\%  of those  between
1077: 1.0-1.3 MeV, taking BPS08(AGS) SSM best-value fluxes.
1078: Such  a bin analysis will have to  contend with the contribution
1079: from  the line-source pep  neutrinos as  well, however,  so the  accuracy with
1080: which $\alpha$ can be measured is certainly not clear.
1081: 
1082: However, an experimental determination is probably not required.  We can write
1083: $\alpha$ as
1084: \begin{equation}
1085: \alpha  =   {\phi(^{13}\mathrm{N})  \over  \phi(^{15}\mathrm{O})}   {  \langle
1086: \sigma^{B/S}(^{13}\mathrm{N},   \delta  m_{12}^2,\theta_{12})   \rangle  \over
1087: \langle \sigma^{B/S}(^{15}\mathrm{O}, \delta m_{12}^2,\theta_{12}) \rangle}
1088: \end{equation}
1089: The cross section ratio can  be evaluated, yielding 0.086(1 $\pm$ 0.0036), when
1090: the LMA oscillation  parameters are varied over the full  range allowed by the
1091: KamLAND  combined analysis.   The  reason for  the  very small  error is  that
1092: variations  in  these  parameters  tend  to  affect  the  two  cross  sections
1093: identically: the 0.8-1.3  MeV event window is narrow,  and clearly differences
1094: must  vanish in  the limit  of a  zero-width window.   The flux  ratio  at the
1095: parameter point defined by the BPS08(AGS) SSM
1096: best values is 1.40.  The changes in 
1097: this ratio obtained  by varying SSM parameters can  be evaluated by procedures
1098: similar to those leading to Eqs. (\ref{eq:prod4}) and (\ref{eq:prod4b}),
1099: \begin{eqnarray}
1100: \alpha  &=&  0.120   (1  \pm  .0036)  \left[{\phi^{SNO}(^{8}\mathrm{B})  \over
1101: \phi^{SSM}(^{8}\mathrm{B})}\right]^{-0.229} \nonumber \\ &\times& \left[ 1 \pm
1102: 0.24\% (\mathrm{resid.~envir.}) \pm 2.0\% (\mathrm{nuclear}) \right] \nonumber
1103: \\   &\times&  \left(   {X(^{12}\mathrm{C})   \over  X(^{12}\mathrm{C})_{SSM}}
1104: \right)^{0.053}  \left(  {X(^{14}\mathrm{N})  \over  X(^{14}\mathrm{N})_{SSM}}
1105: \right)^{-0.058} ~~
1106: \label{eq:prod4c}
1107: \end{eqnarray}
1108: One  finds that  $\alpha$  is stable  under  reasonable parameter  variations:
1109: changes  induced by the  nuclear physics  or core  temperature tend  to affect
1110: these fluxes  in similar  ways.  This includes variations  in core
1111: metals, the  quantity we hope  to constrain: adjustments in the C  or N
1112: primordial abundance by 30\% produce changes at or below 1.5\%, that is, 0.120
1113: (1 $\pm$ 0.015).   If the overall metalicity is changed,  keeping the C/N ratio
1114: fixed, this becomes 0.1\%.  One  concludes from these exercises that the ratio
1115: of  events in  the  0.8-1.3 MeV  window  from $^{13}$N  and $^{15}$O  neutrino
1116: interactions to that from $^{15}$O neutrinos alone, is 1.120 $\pm$ 0.003, when
1117: all sources of uncertainty are considered.
1118: 
1119: The final step is  to plug Eqs.~\ref{eq:fluxratio}~and~\ref{eq:fluxratio2}
1120: into Eq.~\ref{eq:prod3a} to obtain
1121: \begin{eqnarray}
1122:  {R_{exp}^{B/S}(\mathrm{CN})   \over   R^{B/S}_{cal}(^{15}\mathrm{O},   \delta
1123: m_{12}^2,\theta_{12}))  }  =  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\nonumber  \\
1124: (1.120     \pm     0.003)     \left[     {R^{SK}_{exp}(^8\mathrm{B})     \over
1125: R^{SK}_{cal}(^8\mathrm{B},            \delta            m_{12}^2,\theta_{12})}
1126: \right]^{0.828}~~~~~~~~~~   \nonumber   \\   \times   \left[   1   \pm   2.6\%
1127: (\mathrm{resid.~envir.})  \pm 7.6\%  (\mathrm{nuclear})  \right] \nonumber  \\
1128: \times    \left(     {X(^{12}\mathrm{C})    \over    X(^{12}\mathrm{C})_{SSM}}
1129: \right)^{0.805}  \left(  {X(^{14}\mathrm{N})  \over  X(^{14}\mathrm{N})_{SSM}}
1130: \right)^{0.199}.~~~~~~~~~
1131: \label{eq:final}
1132: \end{eqnarray}
1133: The SK rate term is the  experimental ``thermometer'' we use to remove most of
1134: the solar  model ``environmental'' uncertainty,  leaving in the next  term SSM
1135: uncertainties  that   are  dominated  by  the  nuclear   physics.   But  these
1136: uncertainties are,  in some sense, under  our control, and will  be reduced as
1137: laboratory reaction measurements continue.   The last terms are the primordial
1138: abundances we would  like to constrain.  The role of the  SSM in this equation
1139: is to  define a set  of parameters  and thus a  set of reference  rates, about
1140: which  we then  explore possible  variations.  Those  variations  generate the
1141: environmental   and   nuclear   uncertainties   given  above,   according   to
1142: Eq.~\ref{eq:prod3a}.
1143: 
1144: The   $R_{cal}$   factors   in   Eq.   (\ref{eq:final})   contain   additional
1145: uncertainties, including one important one:
1146: \begin{itemize} 
1147: \item The shape of the normalized neutrino spectra: The $^{15}$O shape spectrum
1148: is allowed,  and thus  accurately known.  The  $^8$B spectrum is  less certain
1149: because the $\beta$ decay populates  a broad final-state resonance.  In the SK
1150: analysis this spectrum  error is among those included  in the systematic error
1151: budget,  so  it  should  not   be  counted  a  second  time.   It  contributes
1152: \citep{hosaka}  at the  $\sim$ 1\%  level.  This  is a  laboratory astrophysics
1153: uncertainty  that could  be lowered  by  improved measurements  of the  $^8$Be
1154: resonance.
1155: \item Uncertainties  in the  elastic scattering cross  section are  also small
1156: ($\sim$ 0.5\% \citealt{hosaka}), and furthermore tend to cancel between the two
1157: normalizing cross sections in Eq.(\ref{eq:final}).
1158: \item The principal uncertainty in  the cross section ratio is that associated
1159: with neutrino  oscillations. Apart  from the dependence  on the  solar density
1160: profile, one can  consider this to be another  type of laboratory uncertainty:
1161: oscillation parameters  can and  will be further  constrained by a  variety of
1162: accelerator and reactor experiments.   For example, KamLAND currently provides
1163: our best constraint on $\delta m^2_{12}$.
1164: \end{itemize}
1165: 
1166: The  LMA parameter  uncertainties  in Super-Kamiokande  and Borexino/SNO+  are
1167: anti-correlated.  Most of the  low-energy $^{15}$N neutrinos do not experience
1168: a level  crossing, residing instead  in a portion  of the MSW plane  where the
1169: oscillations are close to the vacuum oscillation limit:
1170: \begin{equation}
1171: P_{\nu_e}(E_\nu) \rightarrow 1 - {1 \over 2} \sin{2 \theta_{12}}
1172: \end{equation}
1173: Thus  an increase  in  the  vacuum mixing  angle  $\theta_{12}$ decreases  the
1174: $\nu_e$ survival  probability.  The higher energy $^8$B  neutrinos are largely
1175: within  the MSW  triangle,  described  by an  adiabatic  level crossing.   The
1176: limiting behavior for an adiabatic crossing is
1177: \begin{equation}
1178: P_{\nu_e}(E_\nu) \rightarrow {1 \over 2} (1 -\cos{2 \theta_{12}})
1179: \end{equation}
1180: so that an increase in $\theta_{12}$ increases the survival probability.  This
1181: anti-correlation thus leads to larger effects in the ratio.
1182: 
1183: We have evaluated the impact of this uncertainty on Eq. (\ref{eq:final}), using
1184: the  allowed  regions for  $\theta_{12}$  and  $\delta  m_{12}^2$ obtained  in
1185: KamLAND's  combined analysis  \citep{kamland,kamland2},  $0.833 \lsim  \sin^2{2
1186: \theta} \lsim  0.906$ and $ 7.38  \times 10^{-5}$ eV$^2  \lsim \delta m_{12}^2
1187: \lsim 7.80 \times 10^{-5}$ eV$^2$.  This yields
1188: \begin{eqnarray}
1189: {R^{B/S}_{cal}(^{15}\mathrm{O},     \delta     m_{12}^2,\theta_{12}))    \over
1190:  R^{SK}_{cal}(^8\mathrm{B},     \delta     m_{12}^2,\theta_{12})^{0.825}}    =
1191:  ~~~~~~~~~~~~ \nonumber \\ (1 \pm 0.049)\left[ {R^{B/S}_{cal}(^{15}\mathrm{O},
1192:  \delta   m_{12}^2,\theta_{12}))   \over   R^{SK}_{cal}(^8\mathrm{B},   \delta
1193:  m_{12}^2,\theta_{12})^{0.825}} \right]^{BV}
1194: \end{eqnarray}
1195: where $BV$ denotes the best-value ratio.
1196: 
1197: Thus  the overall  uncertainty budget  in Eq.  {\ref{eq:final}}  appears quite
1198: favorable,  with the SK  ``thermometer'' contributing  at 3\%,  residual solar
1199: environmental uncertainties at 1\%,  and LMA parameter uncertainties at 4.9\%.
1200: The  largest  of the  errors  is  that  from nuclear  S-factor  uncertainties,
1201: currently 7.6\%.  The overall  uncertainty in the ``theoretical'' relationship
1202: between a future SNO+ or Borexino CN-neutrino flux and core C/N metals is thus
1203: about 9.6\%.  As  the nuclear physics uncertainty dominates  the analysis, one
1204: would expect this relationship to become more precise when ongoing analyses of
1205: the full data set for $^{14}$N(p,$\gamma$) are completed.  An appropriate goal
1206: would  be 3.5\%  in this  S-factor, a  30\% improvement.   The  uncertainty in
1207: $^{14}$N(p,$\gamma$)  would  no  longer  dominate the  nuclear  physics  error
1208: budget, but instead would be comparable to the contributions from S$_{33}$ and
1209: S$_{34}$.  However, the current 9.6\% uncertainty is not a bad starting point,
1210: as first-generation CN-cycle neutrino experiments are expected to measure this
1211: flux to an accuracy of about  10\%.  That is, the theoretical uncertainty will
1212: not dominate the experimental uncertainty.
1213: 
1214: \section{Future Experiments and Summary}
1215: 
1216: One of the main motivations for  this paper is the development of new detector
1217: ideas  that  might  allow   a  high-statistics  measurement  of  the  CN-cycle
1218: neutrinos.    In   particular,   detectors   based  on   ultra-clean   organic
1219: scintillation liquids  have, at  least in principle,  the potential to  make a
1220: high-sensitivity real-time measurement of the CN-cycle neutrinos.
1221: 
1222: The  Borexino collaboration has  investigated this  possibility \citep{franco}.
1223: Borexino, currently operating in Gran Sasso, has a 300-ton liquid scintillator
1224: target housed in a 8.5m spherical  nylon membrane and shielded by a kiloton of
1225: buffer fluid.  Events  producing light within the detector  are detected by an
1226: array  of 2200  photomultiplier tubes.   The inner  100 tons  of  the detector
1227: comprise the  fiducial volume.   The events come  from elastic  scattering off
1228: electrons, a  reaction that  is sensitive to  both electron and  (with reduced
1229: sensitivity) heavy-flavor neutrinos.
1230: 
1231: Borexino is primarily focused on detecting neutrinos from  the pp chain,
1232: specifically the 862 keV line neutrinos from electron capture on
1233: $^7$Be (the 90\% branch).  However, as mentioned previously, the
1234: collaboration has proposed using the
1235: detection window of (0.8-1.3) MeV to pick up
1236: contributions from the pep line source (1.442 MeV) and the $^{13}$N and
1237: $^{15}$O $\beta$ decay sources. 
1238: 
1239: The primary  obstacle to such a measurement  by Borexino is the  {\it in situ}
1240: cosmogenic  production of  $^{11}$C,  a $\beta^+$  source.  High-energy  muons
1241: produced by primary cosmic ray interactions in the atmosphere can penetrate to
1242: great  depths, producing  $^{11}$C  by  knocking a  neutron  out of  $^{12}$C.
1243: Borexino is located  in Gran Sasso, which has a depth  of about 3.1 kilometers
1244: of water equivalent (kmwe), when converted to the equivalent depth below a flat
1245: surface \citep{meihime}.  The $^{11}$C production is still significant at this
1246: depth: an initial estimate of the associated background of 7.5 c/d/100 tons in
1247: Borexino was recently confirmed by a direct measurement \citep{back}, yielding 
1248: 13.0 $\pm$  2.6 (stat) $\pm$ 1.4  (sys) c/d/100 tons.  This  exceeds the solar
1249: neutrino signal  in the window of  interest.  Thus some means  of vetoing this
1250: background must be introduced.  This vetoing is nontrivial because of the long
1251: mean lifetime of $^{11}$C, 29.4 minutes:  a simple cut based on the muon would
1252: thus not be feasible.
1253: 
1254: The collaboration  has proposed  that a successful  veto might be  possible by
1255: exploiting a triple coincidence, the  initiating muon, the prompt capture of a
1256: neutron on protons in the  scintillator fluid, and the delayed $\beta^+$ event
1257: (\citealt{franco};  Deutsch,  private communication).   This  would allow  the
1258:   experimenters to cut out a 
1259: spherical  volume defined  by  the neutron  capture  vertex, rejecting  events
1260: within  that  volume for  a  time $\Delta$t  large  compared  to the  $^{11}$C
1261: lifetime.    The   simulations   performed   by  Borexino   suggest   that   a
1262: signal/background  ratio of  1.2  could  be achieved  with  20\% deadtime.   A
1263: CN-cycle neutrino  (and pep neutrino) flux  measurement could then  be made by
1264: subtraction of this background.
1265: 
1266: An alternative  to this  approach would be  to place  such a detector  at very
1267: great depth.  This  possibility has been discussed by  the SNO+ collaboration,
1268: which has  proposed placing a  one-kiloton scintillator experiment  in SNOLab,
1269: using the  cavity that was originally  excavated for SNO  \citep{chen}.  Such a
1270: detector  could be  used for  detecting $^7$Be,  pep, and  CN-cycle neutrinos,
1271: geoneutrinos, and double beta decay.  The proposed volume is about a factor of
1272: three greater than that of Borexino.  The great advantage of SNO+ would be its
1273: depth, 6.0 kmwe,  and consequently its much lower  cosmic ray muon background.
1274: The additional  3.0 kmwe, relative to  Gran Sasso, provides about  a factor of
1275: $\sim$  70 additional  attenuation  in the  muon  flux, so  that the  expected
1276: $^{11}$C production would be reduced to $\sim$ 0.1 c/d/100 tons, a few percent
1277: of the expect CN neutrino signal.
1278: 
1279: Figure \ref{fig:CNO_SNOPlus} shows a simulation of the expected SNO+ response,
1280: performed by the experimenters (Chen, private communication).  Note that
1281: the simulation is based on  the  BS05(OP) SSM and the best-fit LMA solution 
1282: to  the solar  neutrino problem, rather than the updated BPS08(AGS) used in this
1283: paper.  The CN-neutrino  event rate  for  an energy
1284: window  above 0.8 MeV  was found  to be  2300 counts/year.   The experimenters
1285: concluded that  SNO+ could  determine the CN-neutrino  rate to an  accuracy of
1286: approximately 10\%, after three years of running \citep{chen}.  This accuracy
1287: is  the  appropriate  goal  for  such  a  first-generation  CN-cycle  neutrino
1288: measurement, as it  would approach the accuracy with which  that flux could be
1289: related theoretically to  the Sun's primordial core C and  N abundances, as we
1290: have argued in this paper.
1291: 
1292: In  this  paper  we  have  suggested  a possible  strategy  for  using  future
1293: Borexino/SNO+ CN-neutrino  measurements as  a test of  the primordial C  and N
1294: abundances in the solar core.  The approach is based on using Super-Kamiokande
1295: as a solar thermometer, to  largely eliminate other SSM uncertainties, so that
1296: a clean  relationship between  these abundances and  CN neutrino rates  can be
1297: made.  SNO,  KamLAND, and other  neutrino oscillation experiments are  used in
1298: the  analysis   to  constrain  LMA  oscillation  parameters.    We  derived  a
1299: relationship  where the dominant  linear dependence  on C  and N  remains, but
1300: other solar-model dependences are  largely eliminated.  This approach exploits
1301: the logarithmic derivatives  that have been previously calculated  for the SSM
1302: (especially those for the separate metal abundances) that define the impact of
1303: SSM  parameter  variations, supplemented by Monte Carlo calculations (which
1304: treat explicitly any correlations that may exist among the parameter variations).
1305: Although  the  relationship  is  derived in  the
1306: context of  the SSM, we suspect  that it remains  valid for a larger  group of
1307: models, e.g., those where SSM parameters are varied well beyond their accepted
1308: SSM uncertainties: many investigators have  shown that the $^8$B neutrino flux
1309: remains a reliable thermometer, even  when such large SSM parameter variations
1310: are made.
1311: 
1312: We have found that the factors that limit the accuracy of Eq. (\ref{eq:final})
1313: are first, uncertainties in nuclear  cross sections ($\sim$ 7.6\%) and second,
1314: uncertainties  in LMA oscillation  parameters ($\sim$  4.9\%).  Both  of these
1315: uncertainties can and will be  reduced in future laboratory measurements.  One
1316: goal should be  the reduction, eventually, to the level  of uncertainty of the
1317: SK thermometer, $\sim$ 3\%.
1318: 
1319: In  summary,  it appears  possible  to  use  future experiments  sensitive  to
1320: CN-cycle neutrinos to constrain C and  N content of the Sun's primordial core.
1321: This would test an important assumption made in the SSM, that the zero-age Sun
1322: was  homogeneous,  with  a  core  metalicity  identical  to  that  of  today's
1323: photosphere.
1324: 
1325: Such a measurement  would also address a significant  controversy, that recent
1326: 3D models of photospheric absorption lines  have led to lower estimates of the
1327: abundances of the volatile metals.  These new analyses appear to be on a solid
1328: foundation,  substantially  improving  absorption  line  systematics  and  the
1329: consistency between the  Sun and other similar stars in  the local group.  Yet
1330: they also significantly alter SSM predictions  of the sound speed in the upper
1331: part  of the  Sun's radiative  zone, so  that  the SSM  is no  longer in  good
1332: agreement with  constraints imposed by  helioseismology.  For this  reason, an
1333: independent measurement of the C and  N abundances in the Sun's radiative core
1334: would be of great interest.
1335: 
1336: This measurement would also place  an important experimental constraint on the
1337: evolutionary history of the Sun.  While  the argument for a homogeneous Sun at
1338: the end of the pre-main-sequence  Hayashi phase appears credible, once the Sun
1339: begins to form a radiative core, there are no subsequent SSM epochs that would
1340: allow mixing of the full Sun.   Thus is principle any anomaly in the accretion
1341: of  metals onto the  Sun, either  during the  main sequence  or in  the Henyey
1342: phase, could produce chemical inhomogeneities.   Such a scenario was the basis
1343: of the low-Z model, one of the proposed solar solutions to the puzzle posed by
1344: the chlorine experiment.
1345: 
1346: Unlike the low-Z  model, a naive attempt to accommodate  both the sound speeds
1347: required  by helioseismology and  the new  photospheric abundances  requires a
1348: convective zone  depleted in  metals.  We have  noted that the  solar system's
1349: primary reservoir for  metals, the gaseous giant planets,  are thought to have
1350: formed  late relative to the evolution of the proto-Sun,
1351: incorporating an excess of metal  estimated at 40-90 M$_\oplus$.  This mass is
1352: similar to the deficit of metals in the convective zone, were one to interpret
1353: the helioseismology/photospheric abundance discrepancy  in the most naive way.
1354: The  raises an  provocative question:  is it  possible that  the  process that
1355: concentrated  metals in the  gaseous giants  also produced  a large  volume of
1356: metal-depleted  gas that  subsequently was  accreted onto  the  Sun's surface?
1357: While the  suggestion of  a common chemical  mechanism linking  the convection
1358: zone  and  the gaseous  giants  is  speculative, we  think  this  is one  more
1359: motivation for  exploiting the CN neutrinos  as a quantitative  probe of solar
1360: core metalicity.
1361: 
1362: \acknowledgments
1363: 
1364: We thank  M. Chen,  P. Goldreich, K.  Snover, Y.  Suzuki, and M.  Wiescher for
1365: discussions.   This  work was  supported  in part  by  the  Office of  Nuclear
1366: Physics,  US Department  of  Energy, under  grant  DE-FG02-00ER-41132. AMS  is
1367: supported  by the  IAS through  a John  Bahcall Fellowship  and NSF grant
1368: PHY-0503584. 
1369: 
1370: \bibliographystyle{apj}
1371: 
1372: 
1373: \begin{thebibliography}{60}
1374: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
1375: 
1376: \bibitem[Abdurashitov et al.(2003)]{sage} Abdurashitov, 
1377: J.~N., et al.\ 2003, Nuclear Physics B Proceedings Supplements, 118, 39 
1378: 
1379: \bibitem[Adelberger et al.(1998)]{adelberger} Adelberger, E.~G., 
1380: et al.\ 1998, Reviews of Modern Physics, 70, 1265
1381: 
1382: \bibitem[Aharmim et al.(2005)]{sno_salt} Aharmim, B., et al.\ 
1383: 2005, \prc, 72, 055502 
1384: 
1385: \bibitem[Ahmad et al.(2001)]{sno1} Ahmad, Q.~R., et al.\ 
1386: 2001, Physical Review Letters, 87, 071301
1387: 
1388: \bibitem[Ahmad et al.(2002a)]{sno2} Ahmad, Q.~R., et al.\ 
1389: 2002, Physical Review Letters, 89, 011301 
1390: 
1391: \bibitem[Ahmad et al.(2002b)]{sno3} Ahmad, Q.~R., et al.\ 
1392: 2002, Physical Review Letters, 89, 011302
1393: 
1394: \bibitem[{{Antia} \& {Basu}(2005)}]{antia}
1395: {Antia}, H.~M. \& {Basu}, S. 2005, \apjl, 620, L129
1396: 
1397: \bibitem[Araki et al.(2005)]{kamland} Araki, T., et al.\ 2005, 
1398: Physical Review Letters, 94, 081801
1399: 
1400: \bibitem[Arpesella  et  al.(2008)]{borexino}  Arpesella,  C.,  et  al.\  2008,
1401:   Physics Letters B, 658, 101 
1402: 
1403: \bibitem[{{Asplund} {et~al.}(2005){Asplund}, {Grevesse}, \& {Sauval}}]{ags05}
1404: {Asplund}, M., {Grevesse}, N., \& {Sauval}, A.~J. 2005, in Astronomical Society
1405:   of the Pacific Conference Series, Vol. 336, Cosmic Abundances as Records of
1406:   Stellar Evolution and Nucleosynthesis, ed. T.~G. {Barnes}, III \& F.~N.
1407:   {Bash}, 25
1408: 
1409: \bibitem[Back et al.(2006)]{back} Back, H., et al.\ 2006, 
1410: \prc, 74, 045805
1411: 
1412: \bibitem[{{Badnell} {et~al.}(2005){Badnell}, {Bautista}, {Butler}, {Delahaye},
1413:   {Mendoza}, {Palmeri}, {Zeippen}, \& {Seaton}}]{opacity}
1414: {Badnell}, N.~R., {Bautista}, M.~A., {Butler}, K., {Delahaye}, F., {Mendoza},
1415:   C., {Palmeri}, P., {Zeippen}, C.~J., \& {Seaton}, M.~J. 2005, \mnras, 360,
1416:   458
1417: 
1418: \bibitem[{{Bahcall}(1989)}]{bahcallbook}
1419: {Bahcall}, J.~N. 1989, {Neutrino astrophysics} (Cambridge and New York,
1420:   Cambridge University Press, 1989, 584 p.)
1421: 
1422: \bibitem[{{Bahcall} {et~al.}(2005{\natexlab{a}}){Bahcall}, {Basu},
1423:   {Pinsonneault}, \& {Serenelli}}]{bbps05}
1424: {Bahcall}, J.~N., {Basu}, S., {Pinsonneault}, M., \& {Serenelli}, A.~M.
1425:   2005{\natexlab{a}}, \apj, 618, 1049
1426: 
1427: \bibitem[{{Bahcall} \& {Davis}(1976)}]{davis_bahc}
1428: {Bahcall}, J.~N. \& {Davis}, R. 1976, Science, 191, 264
1429: 
1430: \bibitem[{{Bahcall} \& {Pinsonneault}(2004)}]{bp04}
1431: {Bahcall}, J.~N. \& {Pinsonneault}, M.~H. 2004, Physical Review Letters, 92,
1432:   121301
1433: 
1434: \bibitem[{{Bahcall} {et~al.}(2001){Bahcall}, {Pinsonneault}, \& {Basu}}]{bpb00}
1435: {Bahcall}, J.~N., {Pinsonneault}, M.~H., \& {Basu}, S. 2001, \apj, 555, 990
1436: 
1437: \bibitem[{{Bahcall} \& {Serenelli}(2005)}]{bs05}
1438: {Bahcall}, J.~N. \& {Serenelli}, A.~M. 2005, \apj, 626, 530
1439: 
1440: \bibitem[{{Bahcall} {et~al.}(2005{\natexlab{b}}){Bahcall}, {Serenelli}, \&
1441:   {Basu}}]{bsb05}
1442: {Bahcall}, J.~N., {Serenelli}, A.~M., \& {Basu}, S. 2005{\natexlab{b}}, \apjl,
1443:   621, L85
1444: 
1445: \bibitem[{{Bahcall} {et~al.}(2006){Bahcall}, {Serenelli}, \& {Basu}}]{bsb_mc}
1446: {Bahcall}, J.~N., {Serenelli}, A.~M., \& {Basu}, S. 2006, \apjs, 165, 400
1447: 
1448: \bibitem[{{Bahcall} \& {Ulrich}(1988)}]{bu88}
1449: {Bahcall}, J.~N. \& {Ulrich}, R.~K. 1988, Reviews of Modern Physics, 60, 297
1450: 
1451: \bibitem[Bemmerer et al.(2006)]{bemmerer} Bemmerer,  D., et al.\ 2006, Nuclear
1452:   Physics A, 779, 297 
1453: 
1454: \bibitem[{{Bodenheimer} \& {Lin}(2002)}]{bod}
1455: {Bodenheimer}, P. \& {Lin}, D.~N.~C. 2002, Annual Review of Earth and Planetary
1456:   Sciences, 30, 113
1457: 
1458: 
1459: \bibitem[{{Brown} {et~al.}(2007){Brown}, {Bordeanu}, {Snover}, {Storm},
1460:   {Melconian}, {Sallaska}, {Sjue}, \& {Triambak}}]{brown}
1461: {Brown}, T.~A.~D., {Bordeanu}, C., {Snover}, K.~A., {Storm}, D.~W.,
1462:   {Melconian}, D., {Sallaska}, A.~L., {Sjue}, S.~K.~L., \& {Triambak}, S. 2007,
1463:   \prc, 76, 055801
1464: 
1465: \bibitem[{{Brun} {et~al.}(1998){Brun}, {Turck-Chi{\`e}ze}, \& {Morel}}]{brun1}
1466: {Brun}, A.~S., {Turck-Chi{\`e}ze}, S., \& {Morel}, P. 1998, \apj, 506, 913
1467: 
1468: \bibitem[{{Brun} {et~al.}(1999){Brun}, {Turck-Chi{\`e}ze}, \& {Zahn}}]{brun2}
1469: {Brun}, A.~S., {Turck-Chi{\`e}ze}, S., \& {Zahn}, J.~P. 1999, \apj, 525, 1032
1470: 
1471: \bibitem[{{Castellani} {et~al.}(1994){Castellani}, {degl'innocenti},
1472:   {Fiorentini}, {Lissia}, \& {Ricci}}]{castellani}
1473: {Castellani}, V., {degl'innocenti}, S., {Fiorentini}, G., {Lissia}, M., \&
1474:   {Ricci}, B. 1994, \prd, 50, 4749
1475: 
1476: \bibitem[{{Castro} {et~al.}(2007){Castro}, {Vauclair}, \& {Richard}}]{castro}
1477: {Castro}, M., {Vauclair}, S., \& {Richard}, O. 2007, \aap, 463, 755
1478: 
1479: \bibitem[{{Chen}(2007)}]{chen}
1480: {Chen}, M.~C. 2007, in American Institute of Physics Conference Series, Vol.
1481:   944, American Institute of Physics Conference Series, 25--30
1482: 
1483: \bibitem[Confortola et al.(2007)]{confortola} Confortola, F., et 
1484: al.\ 2007, \prc, 75, 065803
1485: 
1486: \bibitem[{{Cs{\'o}t{\'o}} \& {Langanke}(2000)}]{csoto}
1487: {Cs{\'o}t{\'o}}, A. \& {Langanke}, K. 2000, Few-Body Systems, 29, 121
1488: 
1489: \bibitem[{{Davis} {et~al.}(1968){Davis}, {Harmer}, \& {Hoffman}}]{davis}
1490: {Davis}, R., {Harmer}, D.~S., \& {Hoffman}, K.~C. 1968, Physical Review
1491:   Letters, 20, 1205
1492: 
1493: \bibitem[{{Descouvemont} {et~al.}(2004){Descouvemont}, {Adahchour}, {Angulo},
1494:   {Coc}, \& {Vangioni-Flam}}]{descouvemount}
1495: {Descouvemont}, P., {Adahchour}, A., {Angulo}, C., {Coc}, A., \&
1496:   {Vangioni-Flam}, E. 2004, Atomic Data and Nuclear Data Tables, 88, 203
1497: 
1498: \bibitem[{{Fiorentini} \& {Ricci}(2003)}]{fr}
1499: {Fiorentini}, G. \& {Ricci}, B. 2003, ArXiv Astrophysics e-prints
1500: 
1501: \bibitem[Formicola et al.(2004)]{formicola} Formicola, A., et 
1502: al.\ 2004, Physics Letters B, 591, 61
1503: 
1504: \bibitem[Fukuda et al.(2001)]{sk1} Fukuda, S., et al.\ 
1505: 2001, Physical Review Letters, 86, 5651
1506: 
1507: \bibitem[Fukuda et al.(2002)]{sk2} Fukuda, S., et al.\ 
1508: 2002, Physics Letters B, 539, 179
1509: 
1510: \bibitem[Fukuda et al.(1996)]{kamioka} Fukuda, Y., et al.\ 
1511: 1996, Physical Review Letters, 77, 1683 
1512: 
1513: \bibitem[{{Grevesse} \& {Sauval}(1998)}]{gs98}
1514: {Grevesse}, N. \& {Sauval}, A.~J. 1998, Space Science Reviews, 85, 161
1515: 
1516: \bibitem[{{Guillot}(2005)}]{guillot}
1517: {Guillot}, T. 2005, Annual Review of Earth and Planetary Sciences, 33, 493
1518: 
1519: \bibitem[Gy{\"u}rky et al.(2007)]{gyurky} Gy{\"u}rky, G., et 
1520: al.\ 2007, \prc, 75, 035805
1521: 
1522: \bibitem[Hosaka et al.(2006)]{hosaka} Hosaka, J., et al.\ 
1523: 2006, \prd, 73, 112001
1524: 
1525: \bibitem[Imbriani et al.(2005)]{imbriani} Imbriani, G., et al.\ 
1526: 2005, European Physical Journal A, 25, 455
1527: 
1528: \bibitem[{{Kajino} \& {Arima}(1984)}]{kajino}
1529: {Kajino}, T. \& {Arima}, A. 1984, Physical Review Letters, 52, 739
1530: 
1531: \bibitem[{{Kirsten}(2003)}]{gallex}
1532: {Kirsten}, T.~A. 2003, Nuclear Physics B Proceedings Supplements, 118, 33
1533: 
1534: \bibitem[Lemut  et al.(2006)]{lemut}  Lemut, A.,  et  al.\ 2006,
1535:   Physics Letters B, 634, 483 
1536: 
1537: \bibitem[{{Mei} \& {Hime}(2006)}]{meihime}
1538: {Mei}, D.-M. \& {Hime}, A. 2006, \prd, 73, 053004
1539: 
1540: \bibitem[Mohapatra et al.(2004)]{APS} Mohapatra, R.~N., et 
1541: al.\ 2004, ArXiv High Energy Physics - Phenomenology e-prints, 
1542: arXiv:hep-ph/0412099
1543: 
1544: \bibitem[{{Montalb{\'a}n} {et~al.}(2004){Montalb{\'a}n}, {Miglio}, {Noels},
1545:   {Grevesse}, \& {di Mauro}}]{montalban}
1546: {Montalb{\'a}n}, J., {Miglio}, A., {Noels}, A., {Grevesse}, N., \& {di Mauro},
1547:   M.~P. 2004, in ESA Special Publication, Vol. 559, SOHO 14 Helio- and
1548:   Asteroseismology: Towards a Golden Future, ed. D.~{Danesy}, 574--+
1549: 
1550: \bibitem[{{Nara Singh} {et~al.}(2004){Nara Singh}, {Hass}, {Nir-El}, \&
1551:   {Haquin}}]{singh}
1552: {Nara Singh}, B.~S., {Hass}, M., {Nir-El}, Y., \& {Haquin}, G. 2004, Physical
1553:   Review Letters, 93, 262503
1554: 
1555: \bibitem[{{Pe\~na-Garay} \& {Serenelli}(2008)}]{bps08}
1556: {Pe\~na-Garay}, C. \& {Serenelli}, A.~M. 2008, in preparation
1557: 
1558: \bibitem[{{Podolak} {et~al.}(1993){Podolak}, {Hubbard}, \& {Pollack}}]{podolak}
1559: {Podolak}, M., {Hubbard}, W.~B., \& {Pollack}, J.~B. 1993, in Protostars and
1560:   Planets III, ed. E.~H. {Levy} \& J.~I. {Lunine}, 1109--1147
1561: 
1562: \bibitem[{{Runkle} {et~al.}(2005){Runkle}, {Champagne}, {Angulo}, {Fox},
1563:   {Iliadis}, {Longland}, \& {Pollanen}}]{runkle}
1564: {Runkle}, R.~C., {Champagne}, A.~E., {Angulo}, C., {Fox}, C., {Iliadis}, C.,
1565:   {Longland}, R., \& {Pollanen}, J. 2005, Physical Review Letters, 94, 082503
1566: 
1567: \bibitem[{{Saumon} \& {Guillot}(2004)}]{saumon}
1568: {Saumon}, D. \& {Guillot}, T. 2004, \apj, 609, 1170
1569: 
1570: \bibitem[{{Strom} {et~al.}(1993){Strom}, {Edwards}, \& {Skrutskie}}]{strom}
1571: {Strom}, S.~E., {Edwards}, S., \& {Skrutskie}, M.~F. 1993, in Protostars and
1572:   Planets III, ed. E.~H. {Levy} \& J.~I. {Lunine}, 837--866
1573: 
1574: \bibitem[{{The BOREXINO Collaboration}(2005)}]{franco}
1575: {The BOREXINO Collaboration}. 2005, Nuclear Physics B Proceedings Supplements,
1576:   145, 29
1577: 
1578: \bibitem[{{The KamLAND Collaboration}(2008)}]{kamland2}
1579: {The KamLAND Collaboration}. 2008, ArXiv e-prints, 801
1580: 
1581: \bibitem[Trautvetter et al.(2008)]{traut} Trautvetter, H.~P., 
1582: et al.\ 2008, Journal of Physics G Nuclear Physics, 35, 4019
1583: 
1584: \bibitem[{{Winnick} {et~al.}(2002){Winnick}, {Demarque}, {Basu}, \&
1585:   {Guenther}}]{winnick}
1586: {Winnick}, R.~A., {Demarque}, P., {Basu}, S., \& {Guenther}, D.~B. 2002, \apj,
1587:   576, 1075
1588: 
1589: \bibitem[{{Wuchterl} \& {Klessen}(2001)}]{wuchterl}
1590: {Wuchterl}, G. \& {Klessen}, R.~S. 2001, \apjl, 560, L185
1591: 
1592: \end{thebibliography}
1593: 
1594: 
1595: \clearpage
1596: 
1597: %% Use the figure environment and \plotone or \plottwo to include
1598: %% figures and captions in your electronic submission.
1599: %% To embed the sample graphics in
1600: %% the file, uncomment the \plotone, \plottwo, and
1601: %% \includegraphics commands
1602: %%
1603: %% If you need a layout that cannot be achieved with \plotone or
1604: %% \plottwo, you can invoke the graphicx package directly with the
1605: %% \includegraphics command or use \plotfiddle. For more information,
1606: %% please see the tutorial on "Using Electronic Art with AASTeX" in the
1607: %% documentation section at the AASTeX Web site,
1608: %% http://www.journals.uchicago.edu/AAS/AASTeX.
1609: %%
1610: %% The examples below also include sample markup for submission of
1611: %% supplemental electronic materials. As always, be sure to check
1612: %% the instructions to authors for the journal you are submitting to
1613: %% for specific submissions guidelines as they vary from
1614: %% journal to journal.
1615: 
1616: %% This example uses \plotone to include an EPS file scaled to
1617: %% 80% of its natural size with \epsscale. Its caption
1618: %% has been written to indicate that additional figure parts will be
1619: %% available in the electronic journal.
1620: 
1621: \begin{figure*}
1622: \includegraphics[width=12cm]{f1.pdf}
1623: \caption{ The pp-chain for hydrogen burning.  The relative termination rates
1624: of competing reactions correspond to the BPS08(AGS) SSM. \label{fig:one}} 
1625: \end{figure*}
1626: 
1627: \clearpage
1628: 
1629: %% Here we use \plottwo to present two versions of the same figure,
1630: %% one in black and white for print the other in RGB color
1631: %% for online presentation. Note that the caption indicates
1632: %% that a color version of the figure will be available online.
1633: %%
1634: 
1635: \begin{figure*}
1636: \includegraphics[width=12cm]{f2.pdf}
1637: \caption{The lower panel shows the CNO bi-cycle for hydrogen burning.
1638: The  upper panel  compares  the energy  produced  in the  CN  cycle with  that
1639: produced 
1640: in the pp-chain, as a function of temperature T$_7$, measured
1641: in units of 10$^7$ K.  The results are normalized to the 
1642: pp-chain energy production in the  Sun's central core and to solar metalicity,
1643: and 
1644: assume  the burning  is  in  equilibrium.  The  sharp  CN-cycle dependence  on
1645: temperature 
1646: is apparent.  If approximated as a  power law T$^x$, $x$ ranges between $\sim$
1647: 19 
1648: and  $\sim$  22   over  the  range  of  temperatures   typical  of  the  Sun's
1649: hydrogen-burning core. 
1650: The dot marks the point corresponding to the Sun's center,
1651: T$_7 = 1.57$. \label{fig:two}}
1652: \end{figure*}
1653: 
1654: % \clearpage
1655: % \begin{figure}
1656: % \includegraphics[width=8cm]{castellani.eps}
1657: % \caption{Neutrino fluxes are plotted at a function of the solar core
1658: % temperature $T_c$ 
1659: % for  calculations based on  the SSM, but  in which input SSM  parameters are
1660: %  varied  far  outside  their  assigned  uncertainties.   In  this  class  of
1661: % ``nonstandard'' solar models it 
1662: % is seen that the flux predictions continue to follow, to good accuracy, the
1663: %  $T_c$ power  laws characterizing  the SSM  (and its  more  modest parameter
1664: % variations).  From Castellani {\it et al.} \label{fig:castellani}.} 
1665: % \end{figure}
1666: 
1667: \clearpage
1668: 
1669: \begin{figure*}
1670: \includegraphics[width=16cm]{f3.pdf}
1671: \caption{Results  from  a   Monte  Carlo  simulation  of  SSM   where  the  11
1672:   environmental parameters  (see text) have  been varied. The two  left panels
1673:   show  the correlations between  the $^8{\rm  B}$ flux  and the  two CN-cycle
1674:   neutrino  fluxes  $^{13}{\rm N}$  and  $^{15}{\rm  O}$.  The slopes  of  the
1675:   correlations are  given in  the plots, together  with the  68.3\% confidence
1676:   level contours.  On the  right side  panels we show  the residuals  from the
1677:   fits,  2.8\% and  2.6\% for  the $^{13}{\rm  N}$ and  $^{15}{\rm  O}$ fluxes
1678:   respectively,  that  determine  the  residual environmental  uncertainty  in
1679:   Eqs.~(\ref{eq:prod4},\ref{eq:prod4b}) respectively.  \label{fig:correl}} 
1680: \end{figure*}
1681: 
1682: \clearpage
1683: 
1684: \begin{figure*}
1685: \includegraphics[]{f4.pdf}
1686: \caption{A simulation of events expected in SNO+, the proposed SNOLab
1687: experiment to measure low-energy solar and other neutrino sources.
1688: This figure is due to M. C. Chen \citep{chen}. \label{fig:CNO_SNOPlus}}
1689: \end{figure*}
1690: 
1691: \clearpage
1692: 
1693: \begin{deluxetable}{lcccccccccc}
1694: \tabletypesize{\small}
1695: \tablewidth{0pt}
1696: \tablecaption{Partial  derivatives  $\alpha(i,j)$  of  neutrino  fluxes  with
1697:     respect  to  solar environmental  and  nuclear  cross section  parameters.
1698:     \label{table:one}} 
1699: \tablehead{ & \multicolumn{4}{c}{Environmental $\beta_j$} &
1700: \multicolumn{6}{c}{Nuclear $\beta_j$}} 
1701: \startdata
1702: Source  & $L_\odot$  &  Opacity &  Age  & Diffusion~&  $S_{11}$  & $S_{33}$  &
1703: $S_{34}$ & $S_{17}$ & $S_{e7}$ & $S_{114}$ \\ 
1704: \hline
1705: $\phi$($^8$B) & 7.16 & 2.70 & 1.38 & 0.28 & -2.73 & -0.43 & 0.85 & 1.0 & -1.0 &
1706: -0.020 \\ 
1707: $\phi$($^{13}$N) & 4.40 & 1.43 & 0.86 & 0.34 & -2.09 & 0.025  & -0.053 & 0.0 &
1708: 0.0 & 0.71 \\ 
1709: $\phi$($^{13}$N)/$\phi$($^8$B)$^{0.599}$ & 0.11 & -0.19  & 0.03 & 0.17 & -0.45
1710: & 0.28 & -0.56 & -0.60 & 0.60 & 0.72 \\ 
1711: $\phi$($^{15}$O) & 6.00 & 2.06 & 1.34 & 0.39 & -2.95 & 0.018 & -0.041 & 0.0 &
1712: 0.0 & 1.00 \\ 
1713: $\phi$($^{15}$O)/$\phi$($^8$B)$^{0.828}$ & 0.07 & -0.18  & 0.20 & 0.16 & -0.69
1714: & 0.37 & -0.74 & -0.83 & 0.83 & 1.02 \\ 
1715: \enddata
1716: \tablecomments{Table   entries  are   the   logarithmic  partial   derivatives
1717: $\alpha(i,j)$  of the  solar  neutrino  fluxes $\phi_i$  with  respect to  the
1718: indicated solar  model parameter $\beta_j$,  taken about the SSM  best values.
1719: All fluxes are in units of  their SSM best values, and thus are dimensionless.
1720: The derivatives, taken from \citet{bps08}, are for the SSM 
1721: BPS08(AGS), which  employs the AGS abundances.  The  two flux ratios were
1722: determined   for  a   $\phi(^8$B)   exponent  that   minimizes  the   residual
1723: environmental error  in the prediction, including  the environmental variables
1724: here and  in Table \ref{table:two}.  As  explained in the text,  that error is
1725: weighted  according  to  the  uncertainties in  the  environmental  parameters
1726: $\beta_j$, given in Table~\ref{table:three}. 
1727: } 
1728: \end{deluxetable}
1729: 
1730: \clearpage
1731: 
1732: \begin{deluxetable}{lccccccccc}
1733: \tabletypesize{\small}
1734: \tablewidth{0pt}
1735: \tablecaption{Partial  derivatives  $\alpha(i,j)$  of  neutrino  fluxes  with
1736:     respect    to   fractional    abundances   of    the    primordial   heavy
1737:     elements. \label{table:two}} 
1738: \tablehead{       &      \multicolumn{2}{c}{C,       N       $\beta_j$}      &
1739:   \multicolumn{6}{c}{Environment Abundance $\beta_j$}} 
1740: \startdata
1741: Source & C& N & O & Ne & Mg & Si & S & Ar & Fe \\
1742: \hline
1743: $\phi$($^8$B) & 0.027 & 0.001 & 0.107 & 0.071 & 0.112 & 0.210 & 0.145 & 0.017
1744: & 0.520 \\ 
1745: $\phi$($^{13}$N) &  0.874 & 0.142 &  0.044 & 0.030 &  0.054 & 0.110  & 0.080 &
1746: 0.010 & 0.268 \\ 
1747: $\phi$($^{13}$N)/$\phi$($^8$B)$^{0.599}$ &  0.858 & 0.141 & -0.020  & -0.013 &
1748: -0.013 & -0.016 & -0.007 & 0.000 & -0.043 \\ 
1749: $\phi$($^{15}$O) &  0.827 & 0.200 &  0.071 & 0.047 &  0.080 & 0.158  & 0.113 &
1750: 0.013 & 0.393 \\ 
1751: $\phi$($^{15}$O)/$\phi$($^8$B)$^{0.828}$ &  0.805 & 0.199 & -0.018  & -0.012 &
1752: -0.013 & -0.016 & -0.007 & -0.001 & -0.038 \\ 
1753: \enddata
1754: \tablecomments{Heavy  elements are  divided into  ``environmental''  metals --
1755: those which  primarily influence the solar  core through their  effects on the
1756: opacity  and thus  the  core temperature  -- and  C  and N,  which govern  the
1757: production  of  $^{13}$N  and  $^{15}$O  solar  neutrinos  and  which  can  be
1758: determined,  in   principle,  from  measurements  of   these  fluxes.  Results
1759: correspond to the BPS08(AGS) model \citep{bps08}.} 
1760: \end{deluxetable}
1761: 
1762: \clearpage
1763: 
1764: 
1765: \begin{deluxetable}{lccccc}
1766: \tabletypesize{\small} 
1767: \tablewidth{0pt} 
1768: \tablecaption{Estimated  1$\sigma$  uncertainties  in  solar and  nuclear  SSM
1769: parameters,  taken  from  Bahcall,  Serenelli,  and  Basu  \cite{bsb_mc}  and
1770: Fiorentini  and Ricci  \cite{fr},  and their  influence  on flux  predictions,
1771: computed  from   the  partial  derivatives  of   Table  \ref{table:one}.   The
1772: experimental value for $S_{17}$ is taken from a series of recent measurements:
1773: this $S$-factor and $S_{114}$ are discussed in the text. \label{table:three}}
1774: \tablehead{$\beta_j$  &   Value  &  $\Delta   \beta_j/\beta_j$(\%)  &  $\Delta
1775:   \phi(^8$B)/$\phi(^8$B)(\%)  &   $\Delta  \phi(^{13}$N)/$\phi(^{13}$N)(\%)  &
1776:   $\Delta \phi(^{15}$O)/$\phi(^{15}$O) (\%)} 
1777: \startdata 
1778: L$_\odot$ & 3.842 $\times$ 10$^{33}$ ergs/s & 0.4 & 2.9 & 1.8 & 2.4 \\ 
1779: Opacity & 1.0 & 2.5 & 6.9 & 3.6 & 5.2 \\ 
1780: Age & 4.57 b.y. & 0.44 & 0.61 & 0.38 & 0.59 \\ 
1781: Diffusion & 1.0 & 15.0 & 4.0 & 4.9 & 5.7 \\ 
1782: p+p & 3.94 $\times$ 10$^{-25}$ MeV b & 0.4 & 1.1 & 0.84 & 1.2 \\ 
1783: $^3$He+$^3$He & 5.4 MeV b & 6.0 & 2.5 & 0.15 & 0.10 \\ 
1784: $^3$He+$^4$He & 0.564 MeV b & 4.9 & 4.1 & 0.25 & 0.20 \\ 
1785: p+$^7$Be & 20.6 eV b & 3.8 & 3.8 & 0.0 & 0.0 \\ 
1786: e+$^7$Be & & 2.0 & 2.0 & 0.0 & 0.0 \\ 
1787: p+$^{14}$N & 1.61 keV b & 5.0 & 0.1 & 3.5 & 5.0 \\ 
1788: \enddata
1789: \end{deluxetable}
1790: 
1791: \clearpage
1792: 
1793: \begin{deluxetable}{lcccc}
1794: \tabletypesize{\small} 
1795: \tablewidth{0pt} 
1796: \tablecaption{Estimated 1$\sigma$  historical (``conservative'') uncertainties
1797: in  AGS abundances,  as defined  in  \cite{bs05}.   The
1798: corresponding  uncertainties in  the  neutrino fluxes  are  computed from  the
1799: partial derivatives of Table~\ref{table:two}.\label{table:four}
1800: }
1801: \tablehead{$\beta_j$     &    $\Delta    \beta_j/\beta_j$(\%)     &    $\Delta
1802:   \phi(^8$B)/$\phi(^8$B)(\%)  &   $\Delta  \phi(^{13}$N)/$\phi(^{13}$N)(\%)  &
1803:   $\Delta \phi(^{15}$O)/$\phi(^{15}$O)(\%)
1804: }
1805: \startdata
1806: C & 29.7 & 0.70 & 25.5 & 24.0 \\ 
1807: N & 32.0 & 0.03 & 4.0 & 5.7 \\
1808: O & 38.7 & 3.6 & 1.4 & 2.3 \\
1809: Ne & 53.9 & 3.1 & 1.3 & 2.0 \\
1810: Mg & 11.5 & 1.2 & 0.59 & 0.87 \\
1811: Si & 11.5 & 2.3 & 1.2 & 1.7 \\
1812: S & 9.2 & 1.3 & 0.71 & 1.0 \\
1813: Ar & 49.6 & 0.69 & 0.40 & 0.53 \\
1814: Fe & 11.5 & 5.8 & 3.0 & 4.4 \\ 
1815: \enddata
1816: \end{deluxetable}
1817: 
1818: 
1819: 
1820: 
1821: 
1822: 
1823: 
1824: 
1825: 
1826: \end{document}
1827: 
1828: %%
1829: %% End of file `sample.tex'.
1830: