THE NEXT DECADE
OF STELLAR CYCLES
RESEARCH
RECOMMENDATIONS
DEVELOPED AT A WORKSHOP HELD AT
Jeffrey Hall (
G. Wesley Lockwood (
TABLE OF CONTENTS
List of Attendees – 3
Abstract – 4
Executive Summary – 5
Review and Current Status – 6
Recommendations for the Next Decade –
15
Summary – 23
Commentary by Judith L. Lean – 24
Commentary by John A. Eddy – 26
References – 27
WORKSHOP ATTENDEES
Sydney
Barnes (Florida Institute of
Technology)
Juan
Fontenla (
Peter Foukal (Heliophysics, Inc.)
Claus
Fröhlich (
Mark
Giampapa (National Solar Observatory
–
Jeffrey Hall (
Gregory
Henry (
Michael
Knoelker (High Altitude Observatory)
G. Wesley
Lockwood (
Richard
Radick (National Solar Observatory –
Sunspot)
Gary Rottman (
Steven Saar (
O. R. White (High Altitude Observatory)
ABSTRACT
Long-term, synoptic observations
of the spectroscopic and photometric behavior of Sun-like stars has been
performed at select observing sites for nearly 40 years. Most of the spectroscopic data have been
collected at the Mount Wilson Observatory (MWO), beginning in March 1966 with
Olin Wilson’s initial observations of the cores of Ca II H&K lines in a set
of 139 Sun-like stars. Since 1994, Ca II
H&K and echelle data of the Sun and 300 Sun-like stars have been gathered
at Lowell Observatory using the Solar-Stellar Spectrograph (SSS), complementing
the MWO target set and spectral coverage. Synoptic photometry was carried out
for 18 years at
We convened a workshop at Lowell
Observatory on October 9-11, 2003 to plan the next decade of work in this field.
Several speakers gave presentations
about the status of various projects, but the emphasis during the sessions was
on open discussion of the relevant issues. The product of the workshop is this
document: a review of the development and current status of stellar cycles
research, and a summation of the recommendations of the attendees for essential
studies to be performed over the next decade.
For our purposes, “stellar cycles research” is taken in its full modern
context, encompassing not only observations of Sun-like stars, but also solar
variability on multiple timescales and its relevance to terrestrial climate
change.
We first briefly review
stellar cycles research: its historical development and the present status of
the field. We trace the important
threads of the past 50 years, and review current relevant programs, as
presented on the first day of the workshop by several of the participants. The perspectives and discussions of
outstanding problems that emerge from this review clarify the motivation for
the workshop, as well as why we need to do the work outlined in the
recommendations. This background material
can be found on pages 4 to 12 of this document.
With the background in place,
we present the current results and recommendations brought forward by workshop
attendees. These recommendations present
a road map for lines of study likely to be fruitful over the next 10 years, and
begin on page 13.
EXECUTIVE SUMMARY
Research on the nature,
morphology, and physics of stellar activity cycles has had a fruitful, 40-year
history. Beyond its many applications to pure stellar astrophysics, the field
has developed broad relevance to a variety of timely issues.
·
Understanding the
nature of activity cycles in the Sun’s nearest stellar cousins is essential to understanding the nature of
the solar activity cycle itself, as well as illuminating probable solar
behavior during periods of prolonged quiescence such as the Maunder Minimum of
1645-1715.
·
Improved
understanding of stellar irradiance variability allows development of more
realistic solar irradiance constructions. This impacts assessment of global
warming due to solar forcing, thereby playing a crucial role in supporting the
directives of NASA’s Living With a Star (LWS) program.
·
With the launch
of satellites such as the Solar Radiation and Climate Experiment (SORCE), which
provides ongoing observations of the spectral distribution of solar irradiance
variability, continued analogous observations of stellar variability in
complementary bandpasses, both from the ground and space, becomes imperative to
·
Identification of
the most nearly Sun-like stars provides critical guidelines for upcoming
searches for true “solar systems,” i.e., stars with habitable, Earth-like
planets with stable orbits and climates.
·
Assessment of the
overall variability of a large ensemble of nearby stars guides astrobiology
efforts by placing limits on the range of conditions likely to exist on such
planets as they may have.
Synoptic observations of the
Sun and Sun-like stars have been carried out at diverse locations such as Mount
Wilson,
REVIEW AND CURRENT
STATUS
INTRODUCTION
Research on the astrophysics of activity and variability of Sun-like stars covers an extensive literature. Recent work on the so-called “solar-stellar connection,” the solar irradiance data and terrestrial climate reconstructions, and the effect of solar variations on terrestrial climate have diversified the literature even further. The connections between the various threads are complex enough that reviewing the field by any one criterion – chronology, method of observation, class of object – does not necessarily provide the path of least obfuscation. For this review, therefore, we will use results presented at our workshop as springboards for overviews of how we have reached those results during the modern era of stellar cycles research, where we define the “modern era” to begin in 1957. That year marks the appearance of two seminal papers: the observational discovery of the dependence of Ca II H&K emission line widths on luminosity (Wilson & Bappu 1957), and the first of a series of theoretical papers concerning the NLTE source function (Thomas 1957)[1]; it also incorporates all observations and theory now commonly referenced, and defines a 46-year period that divides neatly into two parts that lend a useful historical perspective to the motivation for our workshop.
General developments in the field over this nearly five decade period appear in Figure 1 on the next page, taken from a slide shown at the opening of the workshop. This diagram shows some (though certainly not all) of the observational programs of the last 40 years, with arrows indicating their durations, some of the essential observational quantities (in red), and a few of the references and major reference series (in green). Shaded ellipses in the background highlight major turning points in the field. We argue in the sections below that one of these turning points is now, and this explains the timing of the workshop. We will now briefly describe the major programs and developments in the field, and then turn to the essential results.
MAJOR PROGRAMS AND DEVELOPMENTS
Long-Term Programs, 1960 - 2000
Long-term programs devoted to observations of the stars
appear in blue in Figure 1. Olin Wilson launched these programs in
the mid 1960s with at the Mount Wilson Observatory (MWO).[2] After his initial investigations of the now
well-known age-activity (Wilson 1963, 1964) and rotation-activity (Wilson 1966)
relationships, he began synoptic observations of the cores of the Ca II H&K
lines a set of 139 stars “for the purpose of initiating a search for stellar
analogues of the solar cycle” (Wilson 1968).
The program continued under his guidance until his publication of the
first magnum opus stellar cycles paper (
FIGURE 1.
An overview of
stellar cycles research. Far more is
omitted than shown, but many of the essential developments and projects are
listed. Major programs appear at left,
with descending arrows indicating their duration. The observational quantities used by workers
in the field are shown in red, and some of the essential references appear in
green at right. The shaded blue
ellipses in the background mark major turning points in our approach to stellar
cycles research.
The “
In the late 1980s, an
instrument called the Solar-Stellar Spectrograph (SSS) was developed at the
High Altitude Observatory and installed at the Lowell Observatory, designed to
complement the Mount Wilson HK photometer, the SSS incorporates both a
single-order instrument covering the HK region as well as an echelle covering
the optical and near IR from λλ 5100 to 9000 Å with 70% spectral
coverage. It is also uniquely equipped
to observe both the Sun and the stars directly.
Regular observations with the SSS commenced in 1994, and the initial
description of the system and the data reduction techniques appears in Hall
& Lockwood 1995.
Long-term observations of G dwarfs in M67, some nine magnitudes fainter than the Wilson stars, have been obtained with the WIYN telescope at Kitt Peak by Mark Giampapa, beginning in 1996 (Giampapa et al. 2000; Giampapa et al. 2004).
Complementing these ground based stellar and solar-stellar
observing campaigns is the long-running series of observations from stations of
the National Solar Observatory at
Developments, 1977 - 1982
Stellar cycles research underwent critical changes between 1977 and 1982, about halfway between publication of the famous Wilson-Bappu relation and the present. This important period in the field is indicated by the second of three background ellipses in Figure 1, and a perspective on its impact is critical to projecting fruitful lines of work over the next decade.
Perhaps the most significant of these changes was the advent
of the first spaceborne total solar irradiance (TSI) measurements, which
commenced with observations from the Nimbus 7 satellite, launched in late
October 1978, and the Solar Maximum Mission (SMM, launched on Valentine’s Day
1980), beginning a unbroken ongoing time series of TSI data that continues today
using instruments aboard the Solar Radiation and Climate Experiment (SORCE). For the first time, it became possible to
directly compare the irradiance variability of the Sun with its chromospheric
activity, and these growing data sets were part of the impetus behind the
initiation of the
In 1978, Johannes Hardorp launched another important line of study by beginning a systematic search for solar analogs – the stars most closely resembling the Sun (Hardorp 1978, 1980a). The literature quickly became contentious (Clements & Neff 1979, Hardorp 1980b), and Hardorp’s methods did result in his reaching poor conclusions about some stars (positing Van Buren 64, for example, as a solar analog; see the comments by Garrison 1985). However, Hardorp’s work inspired a second set of investigations (Cayrel de Strobel et al. 1980, Cayrel de Strobel & Bentolila 1988, Friel et al. 1989), an exhaustive review of the topic (Cayrel de Strobel 1996), and more recently the identification of the current best (and only) solar twin, 18 Sco (Porto de Mello & da Silva 1997, Hall 1998). Most significantly for stellar cycles workers, the search for “the Sun among the stars,” as Hardorp termed it, led to greatly increased interest in the so-called “solar-stellar connection,” and the idea that understanding putative effects of solar variability on terrestrial climate can be aided by analyses of activity cycles and irradiance variations of an ensemble of the most nearly Sun-like stars.
These advances – flux-calibrated data from satellite
observations of stars, the growing solar TSI database, and increased efforts to
identify the most solar-like stars – led to commensurate evolution in
interpretation of the data between 1978 and 1984, as is evident in the
observational quantities listed in red in Figure 1.
Today “Mount Wilson S” is universally recognized as a
standard measure of stellar activity, but it contains a color term (due to its
dependence on nearby continua) that render it unsuitable for direct physical
comparison of stars of different temperature, or for interpretation in the
context of absolutely calibrated data or models, such as the theoretical
flux-calibrated HK profiles being produced at that time by the Boulder group in
their Stellar Model Chromospheres (“SMC” in Figure 1) series of papers (Kelch,
Linsky, & Worden 1979, Linsky et al. 1979, Giampapa et al. 1981). Thus, in 1982, we find the first of the papers
that develop the by-now familiar quantity R’HK
(shown in Figure 1 as one of the significant steps in the development of
stellar activity observables). The concept was first formulated by Middelkoop
1982, in paper IV of the
The growing IUE, EINSTEIN, HEAO, and TSI databases made it also imperative to understand S in terms of physical flux, pursuant to the investigation of “flux-flux relations” to illuminate the processes driving energy transfer between the chromosphere and outer atmospheres in cool stars (indicated by the “FHK – Fλ” entry in Figure 1). Middelkoop (1982) provided the first prescription, but here the picture, as did the solar analog picture, became murky. The conversion from color-independent S to flux was refined several times (Oranje 1983, Rutten 1984, Schrijver et al. 1989), with the effect that many references in the early 1990s avoid the issue by using the dimensionless, color-independent form of S, which is at least linearly related to flux. The issue was further examined by Hall & Lockwood 1995 in connection with the SSS project, with the conclusion that Middelkoop’s original prescription was correct.
To add a final – but enormously significant – thread to the field, interest was renewed in solar-terrestrial interactions (Eddy 1976, Eddy 1977, Hays 1977), in the volume “The Solar Output and its Variation,” edited by Dick White. While interest in solar influences on climate had generated lengthy treatises earlier in the 20th century (e.g., Abbott 1929), the impending availability of quantitative solar irradiance data finally made formal study tractable. Understanding Sun-like stars as possible proxies of solar activity, especially in connection with periods such as the solar Maunder Minimum, brought stellar cycles work to the full attention of the climate community.
By 1984, therefore,
stellar cycles research had changed significantly from its state in 1977.
Solar Irradiance and Climate, 1980-2000
Many of the recent developments and efforts in stellar cycles research stem from the rapid resurgence of interest in Sun-climate interactions, especially in light of the rapid global temperature rise observed since about 1970, as well as from the availability of solar irradiance observations that prompted broader synoptic stellar observations than pure HK studies (the Lowell photometry program and the effort at HAO to build the SSS were both outgrowths of this).
Olin Wilson himself, in fact, presaged these recent broad
trends in a comment he made toward the end of an IAU colloquium on stellar
chromospheres (
Thus, we find a slightly revised form of
Contributing to the interest in broadband solar and stellar variability proxies were the accumulating records of solar irradiance data from a succession of satellites. The composite record appears below.
FIGURE 2.
The latest composite
record of solar irradiance (not yet including the recent SORCE data). From presentation
by C. Fröhlich, PMOD.
As these solar irradiance data accumulated through the 1980s, models of the solar luminosity variations appeared in the literature, typically employing a two-component model in which the positive correlation between solar activity level and total irradiance is explained by a excess facular brightening slightly dominating sunspot darkening (Foukal & Lean 1988). Contributions by additional non-facular components have been proposed (Kuhn & Libbrecht 1991), though Lean et al. (1998) find that these components are not required to recover the observed TSI variations, and Radick et al. (1998) assumed a two component model in their analysis of facular versus spot domination of irradiance variations in Sun-like stars of differing ages.
Closely connected with the solar-stellar irradiance analyses has been the application of these data in reconstructing terrestrial climate variability on both recent and long timescales. Intense interest and scrutiny surrounded the publication of an apparent extremely tight correlation between the global northern hemisphere temperature and the length of the solar cycle (Friis-Christensen & Lassen 1991), as well as a finding, based on Mount Wilson time series of non-cycling stars, that a Maunder Minimum-like phase may entail a brightness decrease of as much as 0.4%, well in excess of the current solar cycle variation (Baliunas & Jastrow 1990). However, others have not recovered the Friis-Christensen & Lassen result and Laut (2003) has published a detailed discrediting of the work. Hall & Lockwood (2004) have likewise been unable to recover the Baliunas & Jastrow distribution of cycling versus non-cycling stars, and Giampapa, in his presentation at the Stellar Cycles workshop, showed that he also does not recover a bimodal distribution of cycling versus flat stars in a 45-star sample of G dwarfs in M67. It appears that while solar forcing does affect climate and can reproduce much of the climate record until as recently as the slight cooling from 1940-1960, the magnitude of the effect is not known and is not yet well illuminated by present stellar results.
A New Turning Point, 2004
As indicated by the lowest ellipse in Figure 1, we are at a timely point to evaluate the status of stellar cycles research, and to identify future directions for the field. We will first very briefly summarize some of the results presented at the workshop.
As discussed above, there are two synoptic spectroscopic
observing programs underway, the 40-year
Complementing these studies are the ongoing photometric
observations of Sun-like stars being carried out at Fairborn Observatory.
Significantly, between the Mount Wilson, SSS, and
FIGURE 3.
Representative Ca II
H&K series of Sun-like stars from the
FIGURE 4. Representative Ca II H&K
series of stars of near-solar color from the Solar-Stellar Spectrograph
program, presented at the workshop by Jeffrey Hall. Activity is expressed in terms of the HK
index (left axis), flux (right axis), and derived S (yellow left axis). Blue
bands in each diagram represent the approximate excursion in S for the star as measured directly by
FIGURE 5. Yearly mean differential
magnitudes of HD 146233 = 18 Sco from the
If solar analog candidates are to be identified, surveys are the way to begin, and they are fortunately underway. T. J. Henry et al. (1996) have surveyed the Ca II H&K emission in 800 southern stars, which David Soderblom discussed at the workshop (Figure 6).
FIGURE 6. Chromospheric activity in 800 southern G dwarfs. The quantity log R’HK is a common measure for expressing the activity level; the Sun lies at B – V = 0.65 and log R’HK ≈ -4.95, in the middle of the “inactive” star classification band. From presentation by D. Soderblom, STScI.
Soderblom also discussed a large, volume-limited survey he
has undertaken of northern G dwarfs to about 50 pc, or roughly V=9, as determined from HIPPARCOS parallaxes. Surveys such as this one are an essential
starting point for identification of additional targets for the next decade of
stellar cycles work.
The paucity of good solar analogs, however, highlights a deficiency in the current stellar databases, at least as far as comparing solar irradiance with direct stellar counterparts goes: there is only one star brighter than V=7 that seems to truly resemble the Sun. If stellar cycles work is specifically to provide insight into solar and Sun-climate work – and this seems a sensible objective given the present uncertainty in the luminosity effects of Maunder minima, the overall role of solar forcing in climate change, and federal programmatic objectives in Sun-climate connections and even extrasolar planet searches and exobiology – then some important gaps in our observational approach must be addressed.
A scan of the review above will reveal a nearly monomaniacal
fixation on Ca II H&K – for good reason.
The features in question are accessible from the ground and respond at
an easily detectable level to changes in solar and stellar activity. However, they also lie well off the Sun-like
star blackbody peak, with the line cores at roughly 8% of the already weak
continuum. The MWO and SSS measurements
have precisions of no better than a few percent, and, as
RECOMMENDATIONS FOR
THE NEXT DECADE
The purpose of this
workshop was to provide a limited, well-directed set of goals for the field of
stellar cycles research over the next decade. The second day of the workshop consisted
exclusively of open discussions intended to identify these goals, and the
recommendations below have been distilled from the discussions that took
place. Representatives of the
Each of the six recommendations
below is presented in the same format. A
summary statement appears first, followed by the reasoning that led to the
recommendation and any qualifying points of view.
Œ
A next-generation, automated observatory dedicated to
synoptic spectroscopic observations of Sun-like stars should be constructed in
the next 3-6 years.
REASONS
- Existing
synoptic spectroscopic programs are becoming prohibitively expensive to
maintain in their present, non-automated operating mode. W. Lockwood reported the SSS program is
understaffed for the amount of observing needed, and S. Baliunas reported
(via Lockwood) that
- Equipment
used in the existing programs is aging and/or outdated. The HKP-2 photometer is more than two
decades old. The SSS uses small,
noisy CCDs and relies on a hardware system that has many interrelated,
irreplaceable parts; a critical hardware failure would be difficult or
impossible for current program staff to repair.
- The
recommendations listed below will (1) exacerbate the problems outlined
above or (2) cannot be achieved with current instrumentation.
DISCUSSION
The call for an automated facility
came most strongly from G. Henry, who operates the automated photometric
telescopes (APTs) at Fairborn Observatory.
Attendees concurred with his opinion that coordinated photometric and
spectroscopic observations are vital to progress over the next decade, and that
neither the
Questions followed regarding the
feasibility of automating current programs.
Attendees concurred that automation of the
FIGURE 7. The
Fairborn Observatory in sourther Arizona’s Patagonia Mountains, where Tennessee
State University, operates eight 0.4m to 2.0m telescopes.
Attendees considered what would
constitute the ideal spectroscopic facility to complement the
Observations using existing synoptic spectroscopic facilities
(
REASONS
- Although
future programs will most productively expand both the set of stars being observed
and the spectral coverage beyond that of
- Overlap
of the
- The
HK time series have long since reached the point of being usable for
application to studies of differential rotation (Baliunas et al. 1985) and
stellar dynamos (e.g., Baliunas
et al. 1996, Brandenburg, Saar, & Turpin 1998; Saar & Brandenburg
1999). While many of the
fundamental astrophysical questions posed about the frequency and
morphology of stellar activity cycles by originators of the field have
been answered, useful insights into both the dynamo behavior and changes
in the length, amplitude, or baseline level of the activity cycle in a
single target on multi-cycle timescales (as is well documented for the
Sun) are still possible with continued HK operation of these programs.
DISCUSSION
Program
continuity is considered essential. A
prime example of the problems in reconciling non-overlapped data sets appears
in Figure 2 above, in which the “PMOD” reconstruction of the solar irradiance
composite shows no change in irradiance from the minimum of solar cycles 21 and
22 to that of cycles 22 and 23, while the “ACRIM” composite shows a distinct
rise. Willson (1997) finds that the two
observed solar minima show a secular rise, while Fröhlich & Lean (1998)
find that it does not. Existing data
series must be directly comparable to those from any new program.
Ž
Substantial effort must be directed in the next 1-3
years to ensuring the cross-calibration, internal consistency, and
accessibility of the existing synoptic data sets.
REASONS
- The
evolution of instrumental performance is critical in the reduction and
analysis of data obtained by the long-term programs. Both the SSS and MWO data sets are
known to have systematic artifacts.
The SSS data show a slight long-term evolution manifested in
gradually increasing intensities in the wings of the Ca K line. Though this evolution is small, it
results in measurable systematic increases in the measured K indices
(Figure 8). Baliunas reported, via
Lockwood, that one of the MWO priorities was to “rework calibration for increased
precision and improved archival access.”
Lockwood noted that the S indices of the MWO series of Sun-like
stars show a systematic decrease (Figure 9).
- Although
the calibration of MWO S to SSS
derived S is good, discrepancies
that remain for some stars need to be reconciled.
- Results
from stellar cycles work are presented in myriad ways (K index, S, log R’HK, F), and the various archives are
either scattered about the Web or not accessible at all. Improved cross-calibration of these
quantities, and access and documentation of reduction and analysis
procedures for any of the relevant programs (not necessarily just the
spectroscopic efforts), is vital for broad evaluation of the results.
FIGURE 8. Systematic trend in SSS data. Shown is the mean intensity in the Ca K line wings in SSS spectra of the solar twin 18 Sco, approximately ±0.5 Å from the line core. A slight rise of about 0.005 of continuum intensity over 7 observing seasons is apparent. This same rise appears in the time series of many of the SSS stars. From presentation by Hall, Lowell.
FIGURE 9. Systematic trend in MWO data. All stars in this sample of Sun-like stars
with significant secular changes in S
trend downward. From presentation by Lockwood, Lowell.
DISCUSSION
The general
opinion of the workshop was that although many detailed investigations had been
carried out to resolve cross-calibration discrepancies, there is still no
consensus on the correct reconciliation of the databases. Several conversions of S to flux have been presented, and even the widely used R’HK has two different formulations.
The large
data archives are also generally not published in electronic format. Soderblom noted that the results of his large
Ca HK survey will be published on the web, and attendees encouraged other
programs to follow suit. To some extent,
the task of publishing each archive rests with the individual investigators,
though there was some discussion of a central Web site that, even if not
actually containing all the individual data, contained summaries, essential
results, and a concise set of links to external sites containing the archival
data from the various stellar cycles programs.
Coordination between complementary stellar cycles
programs, both ground-based and space-based, must be improved, and should be
incorporated in future proposals by the various groups involved.
REASONS
- The
newest generation of space-based solar irradiance observatories also
incorporates spectroscopic instruments covering spectral regions directly
comparable to those observed by current ground-based programs, opening a
new way to directly compare data sets of solar and stellar spectral
variability
- Important
problems remain in our understanding of the spectral distribution of solar
cycle variability, such as the discrepancy between the cycle amplitude of
solar coronal emission and that observed by ROSAT for many stars, which is
the new factor-of-three problem in solar astrophysics.
DISCUSSION
The most
recent solar irradiance mission, the Solar Climate and Radiation Experiment
(SORCE) has not only a TSI monitor on board, but a suite of instruments
including the Spectral Irradiance Monitor (SIM), which have begun synoptic
spectroscopic observations of the Sun from space. The spectral coverage of these instruments
and the irradiance variability models one can construct using these data were
discussed by Gary Rottman and Juan Fontenla of the SORCE project.
Future ground-based observations must take advantage
of large-format CCDs to provide ongoing ing of currently unexplored regions of
the spectrum particularly that from λ4000 to λ5000.
REASONS
- Carefully
chosen proxies other than HK, obtained in regions of the spectrum where
greater S/N is possible than in the HK line cores, have the potential to
provide more robust measures of activity than single-line proxies
- To
more fully understand the origin of luminosity variations in stars, we
need to observe a variety of proxies, as has been done for the Sun by
Livingston and his collaborators.
Long-term observing of features of varying degrees of sensitivity
to brightness changes in magnetic structures, for example, will let us
directly examine the relative importance of these structures in stellar
irradiance variations and, by extension, TSI variability.
DISCUSSION
The increased interest in using spectral features other than
Ca II H&K as activity proxies (e.g., Livingston, White, & Wallace 1987,
Hall & Lockwood 2000) led to extended discussion of the utility of these
features in stellar cycles work. Foukal mentioned
the CH G band at ≈λ4300 as a likely example of such a proxy, and
drew our attention to a fortuitously timed paper (Schussler et al. 2003) on
this very feature that appeared shortly after the workshop. Observations of this and other proxies with
varying degrees of sensitivity to magnetic structures, and with different
places of origin (e.g., chromosphere vs. quiet photosphere), if then compared
with the star’s HK cycle and irradiance variability, might more fully quantify
the origin of luminosity variations over the course of the activity cycles.
The
discussion also highlighted some important gaps in our current stellar cycles
database, summarized in Figure 10 on the next page. The spectrum between λ4000 and
λ5000 contains numerous features that might productively be used as new
activity proxies (including the G band and Hβ), yet spectroscopic time
series from this part of the spectrum do not exist. The SSS echelle does cover the region from
λ5100 to λ9000, but with gaps in the spectral coverage due to the
small size (512 x 512) of the CCD. Given
the ongoing b – y observations from
‘
The most nearly Sun-like stars down to visual
magnitude 10-11 must be identified via surveys and targeted for frequent
observation.
REASONS
- The
broad interest in the use of stellar observations to provide perspective
on solar irradiance variations and proxy reconstructions of long-term
solar variability makes identification of a statistically significant set
of Sun-like stars imperative. Statistics
indicate that the observations must be pushed to at least V=10 for this to be possible.
FIGURE 10. Deficiencies in the current overall
ground-based database, displayed on a time series – V magnitude – wavelength
phase space. The spectroscopic programs
cover a long duration with minimal spectral coverage (MWO) or broader coverage
with a much shorter time series (SSS).
Broad wavelength coverage on the time scale of stellar cycles and
Maunder minima is largely lacking, and the important region between 4000 and
5000 Å is unexplored spectroscopically.
The “
DISCUSSION
Identification of the closest solar analogs per se was largely considered merely a means to an end. Soderblom considered finding a true solar twin extremely unlikely, and similar suggestions about the scarcity of such stars are found in the literature (Cayrel de Strobel 1996, and in her presentation in Hall 1998).
However, the current state of knowledge regarding the current only solar twin, 18 Sco = HD 146233, suggests that finding more stars as nearly similar to the Sun as possible is worthwhile. Originally identified as the best solar twin by de Mello and da Silva (1997), 18 Sco was confirmed as a solar twin, on the basis of a spectral snapshot revealing its gross parameters, at the Solar Analogs workshop at Lowell (Hall 1998).
Time series observations of 18 Sco’s chromospheric behavior
confirmed that it has a reasonably Sun-like activity cycle (Hall & Lockwood
2000), and the
Several essential questions may be addressed with comprehensive, multiwavelength observations of good solar analogs. Do solar age stars with gross parameters extremely close to those of the Sun also exhibit highly Sun-like photometric quiescence, as 18 Sco does? Do they also in general exhibit similar chromospheric cycles, as 18 Sco does? If we find an 18 Sco counterpart that does not cycle, is this a true stellar Maunder Minimum, and if so, can we observe the spectral distribution of its irradiance variability with some confidence that this reflects the real Sun during its own period(s) of quiescence?
While the broad astrophysical goals of synoptic observations of cool stars must not be ignored, the specific potential for future observations of truly Sun-like stars presents at least one well-defined objective for the field in the next decade, which is roughly the timescale for a complete solar cycle, and which is well aligned with current broad programmatic goals.
SUMMARY
The stellar cycles workshop
developed six broad programmatic recommendations that, if followed, should
produce (1) a unified database of stellar parameters, activity records, and
irradiance variability and (2) maximally productive observing programs during
the next decade. Attendees specifically
suggested:
·
Ground-based
spectroscopic programs dedicated to observations of stellar cycles must be made
more compatible with other modern observing programs through construction of a
dedicated, automated observing facility, employing a 1-2 meter class telescope,
and a spectrograph providing as nearly complete spectral coverage from the near
UV to near IR as possible. Existing
programs should continue at least until this facility is in operation, to
ensure accurate cross-calibration of the data sets is possible.
·
Existing ground-based
data sets should be rendered into a publicly accessible final format as the
programs wind down. Centralized electronic
access to the various datasets pertinent to the field should be made available.
·
The
next-generation target set should focus first on the most nearly Sun-like stars
available, to a limiting magnitude of at least V = 10. The target set
should include stars on the
The research begun 40 years ago by
Olin Wilson is increasingly relevant today, and unification of extant databases
with the observing facilities and observational requirements specified for
future programs supports several timely science goals. Foremost among these is surely the role of
solar forcing of terrestrial climate, particularly in the context of the recent
rapid global warming, which is a critical scientific and public policy
issue. Understanding the Sun in the
context of an expanded sample of Sun-like stars, particularly in terms of total
stellar irradiance, spectral distribution of stellar variability, long-term
(multi-cycle) trends in stellar irradiance, and the behavior of stars during
periods of minimal activity, is a vital complement to solar and climatic
research. Additional relevance exists in
identifying optimum targets for extrasolar planet searches, and for understanding
the ensemble properties of Sun-like stars and the attendant implications for
astrobiological conditions in their habitable zones. The attendees of the workshop look forward to
what the next decade of observations will show us.
COMMENTARY
Judith L. Lean
[Editors’ note: This invited commentary is reproduced almost
exactly as received. Some brief comments pointing out minor errors in the original
report text have been edited following corrections. The editors thank Dr. Lean
for her thoughtful commentary.]
This concise synopsis of the evolution and current state of primarily ground-based stellar monitoring is both illuminating and instructive. It is to the authors (and workshop attendees) credit that they have attempted such a broad assessment of their field, and are motivated by the need to define priorities for the next decade of research.
The following comments are offered from an outsider of this field, and are intended not to detract from the report’s overall strengths, but to provide what is a hopefully a helpful additional and even broader perspective.
1) Spectral Irradiance Variability
The report aims at elucidating stellar cycles research in “its full modern context”. Yet the text primarily discusses just two types of stellar variability, namely total (bolometric) irradiance and the Ca HK flux. These are the two types of stellar monitoring that have been mainly conducted thus far. But they do not compose “the full modern context.” In fact, the Ca K (or Ca HK) flux is used in solar irradiance variability modeling as a proxy for the bright faculae that alter not just total solar irradiance at visible and IR wavelengths, but also the UV and EUV irradiance. The variability of spectral irradiance at different wavelengths depends on a balance between the bright facular and the dark sunspots, and this balance is strongly wavelength-dependent. It is also a function of time– the balance differs during rotation versus the activity cycle, and it apparently changes also on very long time scales; younger more active stars are dominated by spots, whereas older, less active stars are dominated by facular.
Thus quantifying the total irradiance and the Ca HK time series is actually a subset of what is a broader goal – to understand the mechanisms/sources of solar and stellar spectral irradiance variability at all wavelengths, on all time scales. Included in this are the UV, EUV and X ray fluxes which the report mentions only briefly (but does appropriately include as one of the recommendations), since it does not describe much about the space-based stellar variability research. The text would benefit from a coherent unifying statement of current knowledge and future needs of stellar spectral irradiance cycles, including observations from space-based observatories of X rays and EUV fluxes, and of time scales. For example, the X-ray solar stellar comparison are mentioned briefly (see recent paper by Judge et al, Ap. J. 2003) yet the relative relationships of irradiance variations from the photosphere, chromosphere and corona is potentially even more powerful in discerning solar-stellar relationships than a study of just the photosphere (total) and chromosphere (UV).
The discussion about what
The report would benefit from a table, for example, giving the amplitude and time scales of variability at different wavelengths in the Sun. This would identify the type of variability being sought in the stellar cycles.
2) Solar Analog
Much is made about what constitutes a “solar analog” and the answer seems to be that there really is no such thing. One solar analog is named. But this is not the same solar analog as, for example, Judge et al (2003) cite when considering X ray fluxes. And the behavior of only one star is unlikely to be, in the end, totally convincingly relevant for the Sun’s behavior. A great value of stellar cycles research is the ability to access many stars at different states of activity, and to statistically quantify how they behave so as to place the Sun in perspective. Even if no star is perfectly identical to the Sun, the insights of average stellar behavior has much value for understanding solar variability. It is, for example, interesting to know that younger stars are thought to be dominated by spots and older stars by faculae. Knowledge like this can help in developing scenarios for long-term solar variability needed for solar terrestrial research. Rather than continue the apparently not very fruitful search for THE solar analog (or two), perhaps stellar cycles research should seek broader categorization of cycles, and their relevance of the Sun. In a sense, this is reversing the focus of the past decade to goals that appeared to inform the first few decades of stellar monitoring.
3) Relevance for the Earth
The report makes a strong point of the relevance of stellar variability for understanding the Sun’s influence on climate change. The information that the stellar cycle research can provide is limits on the range of variability that we expect for the Sun. Even if actual values are elusive, plausible, defendable limits on solar irradiance variations are important to have – both lower and upper limits. An important aspect of such upper and lower limits is that their level of certainty (or uncertainty) be quantified. The original distribution of sun-like stars that Baliunas and Jastrow reported has now been discredited. But why was it so in error? What mistaken assumptions were made? In retrospect, what caveats might have been applied? What were the uncertainties of that distribution? Why should the revised distribution be more believable? There are lessons to be learned from this, for reporting future stellar distributions results.
Also important is the frequency with which the Sun can be expected to reside in, for example, Maunder-type states (or super active states?) relative to “normal” cycling conditions. Aside from the magnitude of the cycle, might stellar cycles research be able to determine such probabilities of high, medium, low activity? Baliunas and Jastrow actually attempted this, and compared their stellar distribution with the cosmogenic isotope record of Maunder-type events in the Sun. More generally, the focus of this report for future stellar cycles record is on utilizing the contemporary solar irradiance datasets, but much can also be learned from studying the very long-term proxies of solar activity (14C and 10Be) since they provide distributions of activity for the Sun.
Furthermore, the relevance of stellar cycles research is broader than just sun-climate – it also applies to the Sun’s influence on the ozone layer, and the upper atmosphere and space weather. Knowing how the upper atmosphere and ionosphere differed in the Maunder Minimum is equally interesting as knowing how climate changed. This means that plausible limits be discerned from stellar data for cycles and long-term variability for the entire solar spectrum – including the RUV, EUV and X rays.
4) Science Goals for the Future
The recommendations for the next decade are in each case a statement of hardware needs and measurement techniques. The report would be strengthened by translating these recommendations into science goals. The instrumentation and the monitoring techniques would then flow down from the science goals.
For example, the science goal of achieving a high statistical sample from which to make meaningful assessments about average stellar cycles requires as much observing time as possible, and hence automated equipment. The science goal of quantifying the spectral irradiance variations in stellar cycles requires a spectral device with wide wavelength coverage and relatively high spectral resolution, rather than a few broad band monitors. The science goal of detecting true trends requires that all new observing devices be properly and carefully connected radiometrically to existing observations.
COMMENTARY
John A. Eddy
[Editors’ note: This invited commentary is reproduced
exactly as received. Dr. Eddy is chairman of NASA’s Living With a Star working
group. The editors thank him for his comments.]
The Stellar Cycles Workshop held at the Lowell Observatory in the autumn of 2003 was, it seems to me, science in its purest form: the kind of thing that Percival Lowell would have applauded. For one, it stepped up to a vexing question in pure astronomy that has pressing, practical implications for national environmental policies, through what other Sun-like stars tell us about our own. It succeeded in bringing together, in the right place, almost all of those who know the most about this important but seldom heralded area of research, to review and discuss and then propose a prudent course for action in the next decade. In the process it also helped illuminate the present limits of what stellar cycles can and cannot support, with certainty, of the long-term behavior of solar irradiation.
As we all know, panels and workshops are common phenomena in science today. Nor is there any shortage of the reports that follow, inevitably, in their wake: laying out, in pontifical terms and with the full Weight and Authority of Those in Attendance, what needs to be done, by whom, and how soon. I was once told that the principal sources of such publications—the NAS and NRC—release about one such report each day, year after year, much like a hen laying eggs.
The product of the Lowell Workshop—The Next Decade of Stellar Cycles Research—is different, for it reports on an effort that was called and organized and put together, on a shoestring, not by those who dwell in the hallowed halls of science, but those who have labored long in the fields. And who, we can hope, will continue their important work for another forty years.
REFERENCES
Abbott, C. G. 1929, The Sun and the Welfare of Man (
Baliunas, S. L., & Jastrow, R.
1990, Nature, 345, 520.
Baliunas, S. L., Nesme-Ribes, E.,
Sokoloff, D., & Soon, W. H. 1996,
ApJ, 460, 848.
Baliunas, S. L., et al. 1985, ApJ,
294, 310.
Baliunas, S. L., et al. 1995, ApJ,
438, 268.
Cayrel de Strobel, G. 1996, A&A
Rev., 7, 243.
Cayrel de Strobel, G., &
Bentolila, C. 1988, A&A, 211, 324.
Cayrel de Strobel, G., Knowles, N.,
Hernandez, G., & Bentolila, C. 1980, A&A, 94, 1.
Clements, G. L. & Neff, J. S.
1979, A&A, 75, 193.
de Mello, G. F. P., & da Silva,
L. 1997 ApJ, 482, L89.
Eddy, J. A. 1976, Science, 192,
1189.
Eddy, J. A. 1977, in The Solar Output and Its Variation, ed.
O. R. White (
Foukal, P., & Lean, J. 1988,
ApJ, 328, 347.
Friel, E., et al. 1993, A&A,
274, 825.
Friis-Christensen, E. & Lassen,
K. 1991, Science, 254, 698.
Fröhlich, C., & Lean, J. 1998,
GRL, 25, 4377.
Garrison, R. F. 1985, in The Calibration of Fundamental Stellar
Quantities, IAU Symp. #111, ed D. S. Hayes et al. (
Giampapa, M. S., Radick, R. R.,
Hall, J. C., & Baliunas, S. L. 2000, BAAS, 32, 832.
Giampapa, M. S., Radick, R. R.,
Hall, J. C., & Baliunas, S. L. 2004, ApJ, in prep.
Giampapa, M. S., Worden, S. P., Schneeberger,
T. J., & Cram, L. E. 1981, ApJ, 246, 502.
Hall, J. C. 1998, ed. Solar Analogs: Characteristics and Optimum
Candidates (Proceedings of the 2nd Lowell Fall Workshop), available at http://www.lowell.edu/users/jch/workshop/sa.html
Hall, J. C., & Lockwood, G. W.
2004, ApJL, in press.
Hall, J. C., & Lockwood, G. W.
2000, ApJ, 514, 436.
Hall, J. C., & Lockwood, G. W. 1995,
ApJ, 438, 404.
Hardorp, J. 1978, A&A, 63, 383.
Hardorp, J. 1980a, A&A, 88, 334.
Hardorp, J. 1980b, A&A, 91, 221.
Hays, J. D. 1977, in The Solar Output and Its Variation, ed.
O. R. White (
Henry, G. W. 1999, PASP, 111, 845.
Henry, T. J. 1996, AJ, 111, 439.
Johannesson, A., Marquette, W. H.,
& Zirin, H. 1998, Sol. Phys., 177, 265.
Keil, S. L., & Worden, S. P.
1984, ApJ, 276, 766.
Kelch, W. L., Linsky, J. L., &
Worden, S. P. 979, ApJ, 229, 700.
Kuhn, J., & Libbrecht, K. 1991,
ApJ, 381, L35.
Laut, P. 2003, JATP, 65, 801.
Lean, J 2000, in Solar
Variability and Climate, ed. E. Friis-Christensen et al. (
Lean, J., Cook, J.,
Linsky, J. L., Worden, S. P.,
McClintock, W., & Robertson, R. M.
979, ApJS, 41, 47.
Lockwood, G. W., et al. 2004, ApJ, in preparation.
Lockwood, G. W., et al. 2002, BAAS, 34, 651.
Middelkoop, F. 1982, A&A, 107,
31.
Oranje, B. J. 1983, A&A, 124,
43.
Radick, R. R., Thompson, D. T.,
Lockwood, G. W., Duncan, D. K., & Baggett, W. E. 1987, ApJ, 321, 459.
Radick, R. R., Lockwood, G. W.,
Skiff, B. A., & Baliunas, S. L. 1998, ApJS, 118, 239.
Rutten, R. G. M. 1984, A&A 130,
353.
Saar, S. H., &
Schrijver, C. J., Cote, J., Zwaan,
C., &
Schussler, M., et al. 2003, ApJ, 597, L173.
Schwarzschild, K. & Eberhard,
G. 1913, ApJ, 38, 292.
Thomas, R. N. 1957, ApJ, 125, 260.
White, O. R., & Livingston, W.
C. 1978, ApJ, 226, 679.
White, O. R., & Livingston, W.
C. 1981, ApJ, 249, 798.
White, O. R., Skumanich, A., Lean,
J., Livingston, W. C., & Keil, S. L.
1992, PASP, 104, 1139.
Willson, R. C. 1997, Science, 277,
1963.
Wilson, O. C. 1963, ApJ, 138, 832.
Wilson, O. C. 1964, ApJ, 140, 1401.
Wilson, O. C. 1966, ApJ, 144, 695.
Wilson, O. C. 1968, ApJ, 153, 221.
Wilson, O. C. 1973, in Stellar
Chromospheres, IAU Colloq. 19, ed. S. Jordan & E. H. Avrett, 305.
Wilson, O. C. 1978, ApJ, 226, 379.
Wilson, O. C., and Bappu, M. K. V.
1957, ApJ, 125, 66.
Worden, J. R., White, O. R., &
Woods, T. N. 1998, ApJ, 496, 998.