The Spectroscopy of Plasma Evolution from Astrophysical Radiation Mission

0.29 MB·English

by  J. Edelstein

Draftversion February5,2008
PreprinttypesetusingLATEXstyleemulateapjv.6/22/04
THE SPECTROSCOPYOF PLASMA EVOLUTION FROM ASTROPHYSICAL RADIATION MISSION
J. Edelstein1, K. W. Min2, W. Han3, E. J. Korpela1,
K. Nishikida1, B.Y. Welsh1, C. Heiles1, J. Adolfo1, M. Bowen1, W.M. Feuerstein1, K. McKee1,
J.-T. Lim2, K. Ryu2, J.-H. Shinn2, U.-W. Nam3, J.-H. Park3, I.-S. Yuk3, H. Jin3, K.I Seon3, D.H. Lee3, E. Sim4
1SpaceSciences Laboratory,UniversityofCalifornia,Berkeley,CA94720
2KoreaAdvanced InstituteofScienceandTechnology, 305-701,Daejeon, Korea
3KoreaAstronomyandSpaceScienceInstitute, 305-348,Daejeon,Koreaand
3KoreaAerospaceResearchInstitute, 305-333, Daejeon,Korea
Draft versionFebruary 5, 2008
ABSTRACT
6
TheSpectroscopyofPlasmaEvolutionfromAstrophysicalRadiation(ortheFar-ultravioletImaging
0
0 Spectrograph)instruments,flownaboardtheSTSAT-1satellitemission,haveprovidedthefirstlarge-
2 areaspectralmappingofthecosmicfarultraviolet(FUV,λ900-1750)background. Weobservediffuse
radiationfromhot(104–106K)andionizedplasmas,molecularhydrogen,anddustscatteredstarlight.
n
These data provide for the unprecedented detection and discoveryof spectral emissionfrom a variety
a
of interstellar environments, including the general medium, molecular clouds, supernova remnants,
J
and super-bubbles. We describe the mission and its data, present an overview of the diffuse FUV
6
sky’s appearance and spectrum, and introduce the scientific findings detailed later in this volume.
2
Subject headings: ultraviolet: ISM: general

1
v
7 1. INTRODUCTION the distribution of metals and driving evolutionary phe-
8 The Spectroscopy of Plasma Evolution from Astro- nomena such as star formation that depend on global
5 physical Radiation instruments (hereafter SPEAR, also morphology. On smaller scales, hot gas can power a
1 known as the Far-ultraviolet Imaging Spectrograph or variety of radiative, mechanical and chemical phenom-
0 ena by forming turbulent and evaporative structures,
FIMS)haveprovidedthefirstlarge-areaspectralskysur-
6 altering chemical abundance via dust ablation or va-
vey of cosmic far ultraviolet(FUV) radiation. The FUV
0 porizing dust condensates, and by radiatively influenc-
band (λ 900-1750) includes important astrophysical di-
/ ing the gas ionization balance. Candidate mechanisms
h agnostics such as strong atomic cooling lines from hot
p (T=104 K – 106 K) and photoionized plasmas as well (seeMcKee(1995)foranoverview)bywhichinterstellar
- as radiation from molecular hydrogen (H ) fluorescent plasmascoolinclude(1)energyexchangewithsurround-
o 2 ing media such as evaporation and conduction, (2) me-
emission and dust scattered starlight. We describe the
r chanical cooling such as adiabatic expansion, (3) direct
t SPEARmissionanditssciencemotivation,introducethe
s emissive radiation, and (4) indirect radiation by way of
a databypresentingthespatialdistributionoftheFUV(λ
heateddust-grainemission. Ourunderstandingofshock-
: 1360 -1710) sky brightness and the spectra of the entire
v ISM interactions and the global structure of ionized gas
observed sky, and provide a brief introduction to other
i in our Galaxy are far from complete. Several models
X SPEAR science results described in the accompanying
for the origin of this hot ionized gas have been pro-
papers in this Volume.
r posed (Slavin & Cox 1992, 1993; Shapiro & Benjamin
a SPEAR is the primary payload on the STSAT-1 (S-
1991;Borkowskiet al.1990;Shelton1998)thatmakedis-
1) satellite launched 2003 September 27. The mission
has thus far observed ∼80% of the sky and conducted tinctive,verifiablepredictions,yetnonecanbe ruledout
with current observations.
deeppointedobservationstowardnumerousselectedtar-
Very hot interstellar diffuse gas (T >= 106 K) was
gets. SPEAR contains dual imaging spectrographs opti-
first mapped by ROSAT soft X-ray (SXR) background
mizedforthemeasurementofdiffuseFUVemission. The
observations (Snowden et al. 1998) which show perva-
spectrographs,referredto as the Short wavelengthband
◦ ′ sive regions of SXR gas at high latitudes, conceivably
’S’ (λλ 900 – 1150, 4.0 x 4.6 view) and Long wave-
◦ ′ caused by hot gas injected into the Galactic halo, al-
length band ’L’: λλ 1350 – 1750, 7.4 x 4.3 view), each
have a spectral resolution of λ/∆λ ∼ 550 half-energy though uncertainty remains regarding the state of this
′ gas. A component of this SXR emission has been at-
width and an imaging resolution of 5. Each instru-
tributed to local hot gas but charge-exchange between
ment uses a collecting mirror, a diffraction grating and
theSolarwindandtheheliosphericenvironsmayconfuse
an open faced, photon counting micro-channel plate de-
the observationsby producingthe samedetected species
tector. The SPEAR instruments, their on-orbit perfor-
(Lallement 2004; Wargelin et al. 2004). Sanders et al.
mance, and the basic processing of instrument data are
(2001)andMcCammon et al.(2002)haverecordedSXR
described in detail in the following paper (Edelstein et
diffuse emission spectra that are inconsistent with pre-
al. ibid.).
dictions of standard models of hot ionized gas but these
2. ENERGETICPLASMAINTHEISM datamayalsobecontaminatedbySolarchargeexchange.
Supernovae and stellar winds produce shock-heated TheCHIPS mission(Hurwitz et al.2005)EUVobserva-
gas in the Galaxy. The activity powered by the ener- tions have established upper limits to local hot gas that
geticplasmashapesthestructureoftheGalaxy,effecting are an order of magnitude less than expected from the
2 EDELSTEIN et al.
postulated local SXR emitting gas, unless this plasma is dozen targets, using long (20-200 ksec) FUSE observa-
extremely depleted. (EUV observations are limited by tions of 30” fields (Dixon et al. 2001; Welsh et al. 2002;
interstellar absorption to nearby regions, with N(HI) < Shelton et al. 2001; Otte et al. 2003).
1018.5 cm−2.) The SPEAR mission was specifically designed to pro-
An overwhelming fraction ( > 85%) of the radia- vide a spectral imaging survey of diffuse FUV emission
tive cooling power from hot, thin interstellar plasma in from the ISM. We report the detection of such emission
emitted in the FUV (Landini & Monsignori Fossi 1990). from most of the sky (see Figure 1). These data provide
These FUV transitions, from the must abundant atoms the first extensive spectral observations of FUV cosmic
in prevalent ground states, provide important diagnos- diffuse emission. We have found diffuse FUV emission
tics for both collisional and photoionized species. FUV lines emanating from both atomic and molecular species
absorption observations have revealed the hot ionized in a variety of interstellar environments.
Galactic ISM. Measurements taken with IUE and HST
haveshownthat Siiv, Civ, andN V ions, characteristic 3. THESPEARMISSION,OPERATIONS&DATA
ofT=104.5–106Kgas,existthroughouttheGalaxywith SPEAR, aboard the S-1 satellite, was injected into a
scale heights up to 4–5 kpc (Sembach & Savage 1992). 700 km sun-synchronous orbit at 98.2◦ inclination with
Hotter gas, with T=105.2–106 K and indicated by O VI an orbital period of 98.5 minute and a ∼34 minute
λλ 1032 absorption, has been observed with Coperni- eclipse. Observationsarescheduledtobegin∼360safter
cus, Voyager, ORFEUS, HUT and now FUSE (Jenkins eclipse entry and end ∼300 s before eclipse exit. About
1978; Hurwitz et al. 1995; Davidsen 1993; Zsargo´ et al. 10dailyorbitsarescheduledforastronomyobservations.
2003). The O VI appears ubiquitous, although patchy One or two orbits per day are used for down-looking ob-
in distribution, and perhaps extends to severalkpc scale servations of the northern night-side aurora. The 3-axis
height (Savage et al. 2003). FUV absorption measure- controlled spacecraft platform can use a star tracker to
ments, limited to sight lines with suitable background achieve 5′ pointing knowledge, pointed accuracy of ∼6′,
sources,cannotaloneyieldthephysicalstateparameters and a stability of ∼12′.
ofthehotplasma(e.g. ne,T,pressure,fillingfactor)and Variouspointingmodesareusedforsurvey,targetand
are not suited to the detection of lines from gas with a calibration observations. Sky survey observations are
large velocity dispersion because placement of the adja- performed by rotating a spacecraft axis such that the
cent stellar continuum is problematic. SPEAR field of view is swept, in a “push broom” fash-
Detection of FUV interstellar emission lines from the ion, perpendicular to its long field of view and in a 180◦
ISM has proven to be difficult. Measurements with great circle from the north to south ecliptic pole via the
soundingrocketexperiments,short-livedorbitalmissions anti-Sundirection. Followingtheanti-Sunprogressionin
and small instruments on interplanetary missions have this way over one year would cause a full viewing of the
suffered from inadequate spectral or spatial resolution; skywithamaximumofoverlappedexposureattheeclip-
an inability to carefully correct for noise sources such ticpolesandaminimumofexposureattheeclipticplane.
as intense geocoronal emission, bright stars and dust- Calibration observations of stars or small (.10◦) fields
scattered stellar continuum; or from integration peri- are performed by a “back and forth” spacecraftrotation
ods insufficient to obtain sensitive results. (See Bowyer that sweeps the field of view over a limited angle. The
(1991) for a review of earlier work.) Space observato- calibration pointing mode, together with reports of the
rieswithFUVmeasurementcapability(e.g.,Copernicus, spacecraftrollrateandtheview-field’sground-measured
IUE, HUT, ORFEUS, GHRS, STIS, FUSE) have been width,providethemostaccuratepositionaldataandex-
optimizedforpointsourceobservationsandnotformea- posure times for the observation of point sources. Fixed
suring faint diffuse spectra over the large angular scales inertial pointing toward specific targets can also be per-
needed to characterize the Galactic plasma. Although formed. Pointings avoid the Sun by 45◦, the spacecraft
the GALEX mission (Martin et al. 2005) was designed velocityvectorby60◦ andarelimited tozenithanglesof
toobservelargeareasofthesky,ithasalimitedcapacity <110◦.
for mapping Galactic diffuse emission and must exclude TheSaTReCMissionOperationsCenter,Daejeon,Ko-
observations of regions including bright stars. GALEX reaisusedtocontrolthe missionandreceivedata. Data
records either a broad-band image including no spectral are decommutated into spacecraft-time marked attitude
information, or an objective-dispersion image that ob- and SPEAR-time marked science and engineering infor-
tains low resolution spectra of localized sources but can mation. The information is passed through a photon
confuse diffuse sources with angular extent. reduction processing pipeline (Edelstein et al. ibid) that
Despite these difficulties, the detection of interstel- performs(1)engineeringselectionofvalidphotonevents,
lar Civ and Oiii] FUV line emission was reported (2)correctionofdetectorelectronicdrift,(3)transforma-
(Martin & Bowyer 1990) 15 years ago (and not since, tion from detector to physical coordinates, (4) correc-
until now). These lines were observed at intensities of tion of detector distortions, and (5) marking of photons
severalthousand LU1 toward four locations, from which with Universal time. A mission data processing pipeline
thepropertiesofaninterstellarcomponentwerederived, creates (1) a time history of attitude knowledge, (2) a
with ne= 0.01 – 0.02 cm−3 and T = 104.7–105.3 K SPEARtospacecrafttimeassociationhistory,(3)atime-
(Shull & Slavin 1994). Emission from interstellar Ovi associatedskyexposurehistoryand(4)photonlistswith
has been convincingly detected more recently, with a time-associated attitude information. The mission data
doublet intensity of ∼3000 - 5000 LU toward about a products, in concordance with the photon data prod-
ucts, can then be combined to produce fluxed spectra
1 Lineintensityunits,“LU”,arephotons s−1 cm−2 sr−1. Con- and spectral sky maps.
tinuumintensityunits,“CU”,areLU˚A−1.
Each photon event is mapped to the sky by using a
SPEAR Mission 3
combination of the spacecraft attitude information, the sweeps. Thismethodhasbeenappliedtotheobservation
instrument’s bore-sight offset from the spacecraft axis, of individual targets and to ∼850 survey sweep orbits.
and the angular position of the event on the detector. Forobservationsthat cycle sweepsoverone field, the at-
Absolute bore-sight is determined by correlating recon- titude timing delay is varied until L-band stellar images
structed sky images with a field of known bright stars. converge. Absolute bore-sight is then determined by us-
The spacecraft attitude knowledge, scheduled for report inga2-dcorrelationoftheimagetobrightstarslistedin
at 1-2 s intervals, is determined by a star-tracker and a the TD-1 catalog (Carnochan & Wilson 1983). For sur-
gyroscope (gyro). The tracker updates are designed to vey or other single sweepobservations,a similar correla-
set the gyro knowledge to ∼2′ accuracy every 5 s. The tionisusedbetweenTD-1catalogobjectsandtheL-band
gyro knowledge drifts at 0.2′ s−1. The knowledge error, sweep image, recomputed for variable timing delays. A
derived from the drift rate and tracker update history, single star’s image can be reconstructed from multiple
′
is assigned by the pipe-line processing for each photon. viewings to an effective resolution of 10 FWHM with a
′
The skylocationattimes in-betweenspacecraftattitude 10 positional accuracy. Survey sweep positional correc-
reports is computed using a spherical coordinate inter- tions in the sweep track direction are found to be ±30′
polation. (2-σ)and±8′ (3-σ)inthe crosstrackdirection. Because
Sky exposure is derived according to valid exposure the S-1 attitude follows a time sequenced commandpro-
intervals(typically60s)thataredeterminedbyexamin- gram,SPEARcanproduceanimageofasweep-observed
ing operational and telemetry-interruption markers rks region even when attitude knowledge is compromised or
interleaved with the photon data. Time-marked expo- absent. We anticipate that the attitude correction for a
sure records are produced every 1.0 s within valid in- large fraction of the data with yet-uncorrected attitude
tervals for each 5’ increment within the spectrometers’ problems can be improved.
field-viewing angles. Each exposure event can then be
mappedto a skypositionina similarfashionasfor pho- 4. THEDIFFUSEFUVSKY:DATAREDUCTION
tons. Exposure events include a weighting factor to ac- We derive a spectral sky-map of diffuse emission from
count for fractional-second intervals that may occur at these data, from which we introduce the appearance of
the end of valid periods, for partial coverage of a sky the FUV sky and its total spectrum. Spectral sky maps
pixel,orforothereffectssuchasprocessingdead-timeor were created by binning photon and exposure events us-
angular vignetting (see Edelstein ibid). Time synchro- ingtheHEALPixtessellationscheme(G´orski et al.2005)
nization between the instrument and the spacecraft is with ∼15′ pixels and λ bins for the L and S bands of
basedupon aninterpolation between spacecrafttime re- 1.0˚A and 1.5˚A, respectively. The accumulations, made
portsandacorrectionthatusesprecise,synchronized1.0 only for times when the derived attitude error is ≤ 30′,
Hz and10 Hz timing marks whichhave been interleaved contains 1.2×107 and 1.4×107 photons for the L and
with the instrument data stream. The timing correc- S bands, respectively. The corresponding sky exposure
tion requires careful handling of clock interruptions or map for the L-band, shown in Figure 1, covers ∼80%
erroneoustelemetry. Photoneventsaremarkedtoapre- of the sky and includes features such as deep exposures
cision of 0.1 Hz with an estimated timing accuracy of (>10 ksec) toward calibration and pointed study fields,
0.25 s. Events with indeterminate time synchronization andexposuresof>500sdegree−2 nearthenorthecliptic
or shutter position, usually due to corrupted and miss- polewheremanysurveysweepsoverlap. Apparentarere-
ing telemetry, are eliminated and correspond to a ∼2% gionswherenocoverageexistsduetotheaforementioned
of the data. attitude problems, to operational interruptions, and to
detector-protective shutdowns while observing FUV in-
3.1. Mission Performance
tense regions suchas the Galactic plane. The integrated
SPEAR’s observational performance depends on sky exposure for the instantaneous field of view is 987 ks,
coverage, sky exposure and mapping accuracy. We re- with an average of 65 s degree−2 for observed sky re-
port on the first year’s (2003 November – 2004 Novem- gions. About15%oftheseobservationsweretakenusing
ber) observations. The S-1 mission was designed for a the 10% shutter aperture. The S-band exposure map is
2-yearmission,howeverspacecraftengineering problems similartothe L-bandmap,althoughcoverageisreduced
may preclude further science observations. During the due to the smaller field of view.
firstyear,2450orbitsofobservationswererecorded. The In orderto obtain true images of the cosmic FUV sky,
S-1 attitude control system behavior limits the quality these data were subject to further reduction to account
of attitude knowledge reconstruction. We find that the for artifacts, evident when simply dividing the photon
bore-sight offset is not constant on a per orbit basis due map by the exposure map, and to account for detector
to a variationin the reporting of attitude. Furthermore, background noise contribution. We proceed with arti-
the star tracker updates are sometimes lost, resulting in fact removal and analysis of the SPEAR L-band map,
an attitude error caused by gyro drift. Finally, espe- for 100% shutter-position observations, because of their
ciallytowardthe endoftheyear,startrackerupdatesor superior coverage and sensitivity in comparison to the
attitude knowledge reports are less frequently available. S-band data. The artifacts, e.g. streaking in the direc-
About 40% of our data suffer from some form of these tion of survey sweeps and localized over-intense regions,
problems. are due to systematic errors in exposure determination
The precision of attitude reconstruction from multi- andtotherecordingofdataduringtimesofhighairglow
ple, overlapping sky viewings is affected by the attitude orradiationbackgroundcontamination. Over-intensere-
system’s stability. We have developed a systematic ap- gionsdue to erroneousmapping were eliminatedby only
proach that uses SPEAR itself as a star tracker to cor- including orbits that contain exposure records and pho-
rect for bore-sight variability in survey and calibration tons that map to >95% of the identical sky pixels. Or-
4 EDELSTEIN et al.
bitshavingexcessiveairgloworparticleinduceddetector nal, a combination of direct unresolved and interstellar
noisewereidentifiedbyusingtheHLyman-β λ1027and dust-scatteredstarlight,andresidualinstrumentalback-
O I λ 1356 count rates as airglow monitors. Entire or- ground.
bits with an averageLyman-β rate exceeding 5 s−1 were The spectral fitting model of Korpela et al. (ibid) is
eliminated. For certain regions showing residual eclip- used to identify and estimate the strength of both the
tic streaking that have been observed repeatedly using astronomical and instrumental components within the
successive sweeps, entire orbits were rejected when their faint-sky signal. We find a median detector dark-noise
average monitor rate exceeded twice the dispersion of contribution of 460 CU, the geo-coronal emission line,
all the orbital rates about their median. The L-band OI λ 1356, and a number of prominent spectral features
spectral map, after artifact removal, contains 7.2×106 which we identify as diffuse astrophysical emission lines
photons and 810 ks of full-field exposure. from the ISM. Strong emission lines from abundant in-
To obtain the diffuse cosmic background, we attempt terstellar species in this bandpass can typically be iden-
to remove bright concentrated sources, e.g. resolved tified with confidence because they are usually well sep-
stars, from the map. Because the FUV diffuse inten- arated from each other and from night-time geocoronal
sity varies by orders of magnitude across the sky, stars lines, although features from unresolved stars or scat-
must be identified as locally intense pixels. An L-band teredstarlightandfrominterstellarH doaddsomecon-
2
total intensity map, integrated over λλ 1360 – 1710, fusion to the spectrum. Line identifications and inten-
was adaptively binned by sky area to attain a statisti- sities of prominent features are shown in Table 1, along
cal signal to noise (S/N) ≥45, value for each bin. In withtheintensities’statisticalerror(10-20%)andline-
thisscheme,thesky-binsizeisincreasedalongHEALPix fit modeling error.
pixel boundaries, i.e. for contiguous pixel groupings of Lines found that are likely to originate from the
n
size 4 , until sufficient counts exist within the bin. In warm/hotISM include the pronouncedCivλλ 1549fea-
this way, bright objects retain small angular dimensions ture and the Oiv]/ Siivλλ 1400/1403 blend, previ-
whilefaintregionsareaveragedto largerangulardimen- ously detected and tentatively identified, respectively,
sionswithvaluesofimprovedsignificance. Brightsources by Martin & Bowyer (1990) at similar intensities. In-
were identified in ∼3% of the pixels as having an inten- terstellar diffuse lines, newly discovered with SPEAR,
∗
sity that exceeds 2.5 times the median value of the en- include the Siii λ 1533 and Aliiλ 1672 resonance lines
veloping4◦×4◦ region(256HEALPixbins). (Wedidnot that could originate from either neutral or warm ion-
attemptto identify starsusingpoint-spreadfunctionde- ized ISM. We attribute the bright1533˚A feature to the
∗
tectionschemesinthispreliminaryworkbecause,ateach Siii λ fine structure ground state transition given that
point in the sky, the mapped point-spread function is a the Siii λ 1527 ground state transition should be opti-
composite of the elliptical instrumental spread-function cally thick for typical interstellar conditions and there-
whose orientation differs for each sweep direction.) fore undetectable. The Alii λ 1671 feature may have
The firstGalactic mapoftotaldiffuse SPEAR L-band been misidentified as the Oiii] λλ 1665 doublet in prior
FUVflux,showninFigure2,isobtainedasanendprod- lower resolution observations (Martin & Bowyer 1990).
uct of the data reduction by eliminating the locally in- Also discovered with SPEAR is diffuse 1657 ˚A emission
tense pixels from the starting continuum map, followed from Ci, a species that exists in the ISM despite pho-
by adaptively binning the pixels to S/N ≥10. toionization by the FUV background (Jenkins & Tripp
2001).
5. THEFUVSKYBRIGHTNESS&SPECTRA
The FUV H fluorescence emission, previously de-
2
The diffuse FUV L-band sky map (Figure 2) shows tected by Martin et al. (1990), presents prominent
thelargestintensitytowardtheGalacticplaneandother and recognizable features in the faint-sky spectra
regions where bright early type stars co-exist with sig- (Black & van Dishoeck1987),particularlyatλ1608and
nificant columns of interstellar dust, such as the obvi- λ1580. (SeetheSPEARobservationsofH inEridanus
2
ous features of the Sco-Cen association and the Mag- by Ryu et al. ibid, for examples of the H fluorescence
2
ellanic Clouds. Thus, the cosmic FUV flux distribu- spectral signature and band identification.) We identify
tion is consistent with what is is generally believed to the feature at 1639.1˚A as Heii λ 1640diffuse emission,
be its dominant component, starlightscattered by inter- although the fit line width, 50% larger than other iden-
stellar dust (Bowyer 1991), and is therefore similar to tified features,indicates spectralcontaminationfromH
2
maps of reddening,N(HI) and H-α (Schlegel et al. 1998; or from Oi airglow, although the night-time only ob-
Dickey & Lockman1990;Finkbeiner2003)becauseboth servations mitigates potential Oi contamination. Some
ofthesemapstraceinsomewayeitherinterstellardustor spectral features appear in both the bright-sky and the
the strong (local) FUV radiationfields that are required faint-sky spectra, such as those at λ 1485 and λ 1520,
to produce the dust-scattered FUV continuum. whichwetentativelyattributetoresidualstellarfeatures
Far UV spectra of the sky are obtained by spatially andsome possible contributionfromthe H fluorescence
2
binning theSPEAR L-bandspectralsky-map. The inte- bands. Thetotalintensityestimatedforallofthepromi-
grated faint-sky (bright-star subtracted) and bright-sky nentinterstellaratomicspectrallinesis∼10kLU,andfor
(star-only)FUV spectra are shownin Figure 3, together the distinct H fluorescence bands is ∼9k LU.
2
with example spectra for the type of stars expected to
dominate the direct and scattered stellar FUV back- 6. SPEARRESULTS
ground (Henry 2002). The bright-sky spectrum has a These FUV continuum and emission line data contain
λλ 1360- 1710median intensity of ∼30kCU. The faint- much astrophysical information. The continuum distri-
sky diffuse spectrumhasabandmedianintensityof2.3k bution depends upon the interstellar dust and radiation
CU that includes contributions from a true cosmic sig- field, the H fluorescence reveals the molecular distribu-
2
SPEAR Mission 5
tion and destruction rate, and the FUV emission lines flux while the cloud halo scatters local flux toward the
indicate the ionization equilibrium state and energetics observer.
oftheISM.Weshallexplorethesemattersinfuturepub- 6.3. Supernova Remnants and Superbubbles
lications and present SPEAR spectral-line sky maps,
SPEAR data provide unique spectral images of two
work which requires comprehensive analyses that is be-
nearby and well studied SNRs, Vela SNR (Nishikida et
yond the scope of this introductory letter. We refer the
alibid)andtheCygnusLoopSNR(Seonetalibid). Both
readerto accompanyingpapers followingin this volume,
authors find that the spatial distribution of FUV emis-
summarized below, that use SPEAR data and provide
sion cannot be simply predicted using visible Hα or X-
detaileddiscussionsontheidentificationandimportance
ray emission maps. The remnants’ emission line images,
of the FUV spectral components in different interstellar
recordedin Ciii, Oiii],Civ andOvifarUV lines, show
environments. whereradiativeshockswithvelocitiesrangingfrom∼100
- 200km s−1 prevail. The work directly verifies that the
6.1. The Diffuse ISM
FUV emission lines are important to SNR cooling. For
Korpela et al (ibid) present SPEAR S and L-band Vela, the combined luminosity of strong FUV lines ex-
spectra of the North Ecliptic Pole (NEP) region (β = ceeds the 1.0 – 4.0 keV X-ray luminosity by an order of
+30◦, N(HI) 2-8×1020 cm−2), a region that has no ob- magnitude, with a FUV to X-ray flux ratio that is a few
vious associations with active interstellar regions and times larger than that for the Cygnus Loop. The Vela
therefore represents a canonical sight-line through the enhancement is attributed in part to its higher covering
ISM. A large number of diffuse FUV atomic emission factor of relatively dense material that can give rise to
lines are detected in the NEP, representing the warm FUV bright shocks.
neutral to the hot ionized medium. The atomic lines, Kregenow et al. (ibid) reports on SPEAR spectra im-
rangingfrom1500to 6000LU in intensity, include those agedacrosstheshell-wallboundarysurroundingtheOri-
identifiedinthe L-bandall-skyspectrumpreviouslypre- Eri superbubble, an X-ray emitting interstellar cavity
sented, as well as Ovi, Ciii, Nii in the S-band, and thatmayhavebeencreatedbySNRorstellarwinds. The
Niv] in the L-band. Presumingthat the resonancelines FUV spectra are richand include lines of similar species
are not dominated by scattering of the interstellar radi- and intensities as found in the all-sky spectrum and in
ation field, it is found that the high ionization potential the NEP (Korpela et al. ibid). Kregenow’s study dis-
lines cannot be fit by a collisionalionization equilibrium covers a distinct correspondence of Ovi, Civ, and Siii∗
plasmamodel eventhoughthese lines, modeledona per with the shell-wall interface traced in Hα, an unprece-
species basis, have consistent emission measures, 0.001 denteddiagnosticforthistype ofevolvedstructure. The
to 0.005 cm−6pc, over the 104.5 to 105.5 K temperature work finds that the boundary emission is may be expli-
range. Therefore, it appears that photo-excitation, non- anedbyeitheraquiescentthermalinterfaceorashocked
equilibrium effects or abundance variations are impor- region.
tant in any explanation of the spectrum. In addition,
substantial H fluorescence emission is detected in the
NEP, despite2the region’s moderately low N(HI). 7. CONCLUSION
ThefirstFUVspectralimagingsurveyofthelargefrac-
6.2. FUV Hydrogen Fluorescence & Continuum tions of the sky has been taken with the SPEAR mis-
sion. The resulting map of total FUV radiation shows
SPEAR observations show that interstellar FUV H
2
thatdiffuse FUV cosmicradiationisconcentratedwhere
fluorescence is ubiquitous, consistent with FUV absorp-
both hot stars and scattering dust coexist, such as in
tion observations(Shull et al.2000; Gillmon et al. 2005)
the Galactic plane, young stellar associations, and the
that find H over large portions of the sky. Lee et al.
2
Magellanic clouds. The spectra of the total-sky contains
(ibid)observesH fluorescenceintheTauruscloud’shalo
2
prominent lines from ionized gas, including the previ-
but not from the dense cloud core, a fact attributed to
ously observed Civ and Siiv emission, and from newly
the core’s opacity excluding the FUV radiation needed
∗
discoveredCi,Siii ,Alii,Heii emission,aswellasfrom
to induce fluorescence. Ryu et al. (ibid) finds H fluo-
2
H fluorescence. Similar interstellar FUV emission lines
rescence emission over a large region about the Ori-Eri 2
are found in the general ISM, as sampled at the NEP,
super-bubbleandisabletoelucidatethegeometryofthe
and toward the Eri-Ori superbubble and its interface to
Ori-Eri region by comparing the FUV fluorescence with
theISM.IntenseFUVemissionlinesfromshocksarevis-
Hα and reddening maps and models an H2 excitation
temperature of≥1000K,largerthan generallyfound for ibleinnearbySNR.DiffuseH2 fluorescenceisubiquitous
across observed Galactic environments. Much work re-
molecular gas in the Galactic disk. The H fluorescence
2
mains to elaborate upon the character of the FUV sky
is also found in deepSPEAR observationsofthe Ori-Eri
andtoanalyzethephysicalconditionsofdetailedobjects.
super-bubbleinterfacetothe ambientISM(Kregenowet
al. ibid).
TheSPEARobservationoftheTauruscloud(Leeetal,
ibid) shows a counter-intuitive anti-correlation of FUV SPEAR / FIMS is a joint project of KASI & KAIST
continuum with visual extinction and IR dust emission. (Korea)andU.C., Berkeley(USA), fundedbythe Korea
The FUV continuum map of Taurus provides an optical MOST and NASA Grant NAG5-5355. We thank the
transfer relation over a wide range of depths that can teamfortheirremarkableefforttocreatethemissionand
be used to quantify the dust and illumination proper- families and friends for their love and support. Thanks
ties – the cloudcore appearsto block more distantFUV Phil.
6 EDELSTEIN et al.
REFERENCES
Black,J.H.&vanDishoeck,E.F.1987,ApJ,322,412 McKee, C. F. 1995, in ASP Conf. Ser. 80: The Physics of the
Borkowski,K.J.,Balbus,S.A.,&Fristrom,C.C.1990,ApJ,355, InterstellarMediumandIntergalactic Medium,292+
501 Nishikida,K.etal.2005, ApJ,thisissue
Bowyer,S.1991,ARA&A,29,59 Otte,B.,Dixon,W.V.,&Sankrit,R.2003,ApJ,586,L53
Carnochan,D.J.&Wilson,R.1983,MNRAS,202,317 Ryu,K.-S.etal.2005,ApJ,thisissue
Davidsen,A.F.1993,Science, 259,327 Sanders,W.T.,Edgar,R.J.,Kraushaar,W.L.,McCammon,D.,
Dickey,J.M.&Lockman, F.J.1990,ARA&A,28,215 &Morgenthaler,J.P.2001,ApJ,554,694
Dixon,W.V.D.,Sallmen,S.,Hurwitz,M.,&Lieu,R.2001,ApJ, Savage, B. D., Sembach, K. R.,Wakker, B. P., et al. 2003, ApJS,
552,L69 146,125
Edelstein,J.etal.2005, ApJ,thisissue Schlegel,D.J.,Finkbeiner,D.P.,&Davis,M.1998,ApJ,500,525
Finkbeiner,D.P.2003, ApJS,146,407 Sembach,K.R.&Savage, B.D.1992,ApJS,83,147
Gillmon,K.,Shull,J.M.,Tumlinson,J.,&Danforth,C.2005,apJ, Seon,K.-I.etal.2005,ApJ,thisissue
submitted Shapiro,P.R.&Benjamin,R.A.1991,PASP,103,923
Go´rski,K.M.,Hivon,E.,Banday,A.J.,Wandelt, B.D.,Hansen, Shelton,R.L.1998, ApJ,504,785
F.K.,Reinecke, M.,&Bartelmann,M.2005,ApJ,622,759 Shelton,R.L.,Kruk,J.W.,Murphy,E.M.,etal.2001,ApJ,560,
Henry,R.C.2002, ApJ,570,697 730
Hurwitz, M.,Bowyer, S., Kudritzki, R.-P.,& Lennon, D.J. 1995, Shinn,J.-H.etal.2005,ApJ,thisissue
ApJ,450,149 Shull,J.M.&Slavin,J.D.1994,ApJ,427,784
Hurwitz,M.,Sasseen,T.P.,&Sirk,M.M.2005,ApJ,623,911 Shull,J.M.etal.2000,ApJ,538,L73
Jenkins,E.B.1978,ApJ,219,845 Slavin,J.D.&Cox,D.P.1992,ApJ,392,131
Jenkins,E.B.&Tripp,T.M.2001,ApJS,137,297 —.1993, ApJ,417,187
Korpela,E.J.etal.2005,ApJ,thisissue Snowden, S. L., Egger, R., Finkbeiner, D. P., Freyberg, M. J., &
Kregenow,J.etal.2005, ApJ,thisissue Plucinsky,P.P.1998,ApJ,493,715
Lallement,R.2004,A&A,422,391 Wargelin, B. J., Markevitch, M., Juda, M.,et al. 2004, ApJ, 607,
Landini,M.&MonsignoriFossi,B.C.1990, A&AS,82,229 596
Lee,D.-H.etal.2005,ApJ,thisissue Welsh,B.Y.,Sallmen,S.,Sfeir,D.,etal.2002, A&A,394,691
Martin,C.&Bowyer,S.1990,ApJ,350,242 Zsargo´, J., Sembach, K. R., Howk, J. C., et al. 2003, ApJ, 586,
Martin,C.,Hurwitz,M.,&Bowyer,S.1990,ApJ,354,220 1019
Martin,D.C.etal.2005,ApJ,619,L1
McCammon,D.etal.2002,ApJ,576,188
SPEAR Mission 7
TABLE 1
Identification of faint-skylinefeatures
Species λ λfit Iλ Statistical Model
Error Uncertainty
˚A ˚A LU a LU LU
Siivc 1403 1403.1 736 162 281
Siii∗ 1533 1532.0 1909 201 275
Civ 1550 1548.7 2636 211 341
H2 1608 1606.5 1832 262 385
Heii 1640 1639.1 3577 378 1177
Ci 1657 1655.7 1316 276 487
Alii 1670 1669.7 1474 280 226
aLU=phs−1 cm−2 sr−1
bLU=phs−1 cm−2 sr−1
cblendedwithOIV]1400
8 EDELSTEIN et al.
Fig. 1.— Sky exposure forthe SPEAR L-band observations. The mapis made with0.5◦ pixels, a histogram equalized intensity scale
with a maximum of 500 s degree−2, Galactic Aitoff coordinates centered at l,b = 0◦,0◦ with longitude increasing toward the left, and is
shownwithlatitudeandlongitudelinesona30◦ grid.
SPEAR Mission 9
Fig. 2.—T
otaldiffuseintensitymapoftheskyfortheSPEAR L-band(λλ1360-1730)observations,afterremovaloflocallyintense pixels(stars).
Themap,inthesamecoordinates schemeasFigure1,hasahistogramequalizedlogarithmicintensityscalewithamaximumof20kCU,
andisadaptivelybinnedbyskyareatoaS/N≥10. EvidentfeaturesincludetheGalacticPlane,theSco-Cenassociation(e.g. zeta-Oph
atl,b=6◦,24◦),andtheLMCatl,b=280◦,-32◦.
10 EDELSTEIN et al.
Fig. 3.— TheL-bandspectra(top)fromthefaint-skyand(middle)fromthebright-sky,plottedwithfalsezeros. (bottom)IUEmeasured
stellarspectralforB3V,B5VandB8Vstarsinarbitraryunits. Dashedlinesmarkprominentfaint-skyspectralfeatures.