CHESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy
Keri Hoadley, Kevin France, Nicholas Nell, Robert Kane, Brian Fleming, Allison Youngblood, Matthew Beasley
NNoname manuscript No. (will be inserted by the editor)
CHESS: an Innovative Concept for High-Resolution,Far-UV Spectroscopy
Instrument Design, Inception, and Results from the FirstTwo Sounding Rocket Flights
Keri Hoadley · Kevin France · NicholasNell · Robert Kane · Brian Fleming · Allison Youngblood · Matthew Beasley
Received: date / Accepted: date
Abstract
The space ultraviolet (UV) is a critical astronomical observing win-dow, where a multitude of atomic, ionic, and molecular signatures provide cru-cial insight into planetary, interstellar, stellar, intergalactic, and extragalacticobjects. The next generation of large space telescopes require highly sensitive,moderate-to-high resolution UV spectrograph. However, sensitive observationsin the UV are difficult, as UV optical performance and imaging efficiencies havelagged behind counterparts in the visible and infrared regimes. This has his-torically resulted in simple, low-bounce instruments to increase sensitivity. Inthis study, we present the design, fabrication, and calibration of a simple, highresolution, high throughput FUV spectrograph - the
Colorado High-resolutionEchelle Stellar Spectrograph (CHESS). CHESS is a sounding rocket payloadto demonstrate the instrument design for the next-generation UV space tele-scopes. We present tests and results on the performance of several state-of-the-art diffraction grating and detector technologies for FUV astronomicalapplications that were flown aboard the first two iterations of CHESS. The
Keri HoadleyCalifornia Institute of Technology, Dept. of Physics, Mathematics, & Astronomy, CahillCenter for Astronomy & Astrophysics, Pasadena, CA 91125, USADavid & Ellen Lee Postdoctoral Fellow in Experimental Physics at CaltechUniversity of Iowa, Dept. of Physics & Astronomy, Van Allen Hall, Iowa City, IA 52242 USAE-mail: [email protected] France · Nicholas Nell · Brian Fleming · Allison YoungbloodLaboratory for Atmospheric and Space Physics, University of Colorado, UCB 392, Boulder,CO 80309, USARobert KaneBlue Canyon Technologies, 2550 Crescent Drive, Lafayette, CO 80026, USAMatthew BeasleySouthwest Research Institute, Department of Space Studies, 1050 Walnut Ave., Boulder,CO, 80302, USA a r X i v : . [ a s t r o - ph . I M ] A ug Keri Hoadley et al.
CHESS spectrograph was used to study the atomic-to-molecular transitionswithin translucent cloud regions in the interstellar medium (ISM) through ab-sorption spectroscopy. The first two flights looked at the sightlines towards α Virgo and (cid:15)
Persei and flight results are presented.
Keywords
Instrumentation: spectrographs - ISM: abundances, clouds,molecules - stars: individual ( (cid:15)
Per (HD 24760))
Ultraviolet (UV) observations are critical for addressing a many key questionsin nearly all aspects of astrophysics. UV observations probe the flow of energet-ics in the universe as UV photons affect atoms, molecules, ions, and dust. TheUV is the peak of the hot star spectral energy distribution (SED), making itthe ideal probe of massive (recent) star-formation and galactic star formationhistories (e.g., Martin et al. 2005). A multitude of atomic, ionic, and molecularspecies have strong resonances with UV radiation, which pump and producestrong emission and absorption lines seen against the UV background. TheUV is home to emission lines in collisional ionization equilibrium at formationtemperatures up to ∼ × K (Tumlinson et al., 2017), critical for assess-ing hot interstellar, circumgalactic, and intergalactic environments. The onlyother wavelength regime with the potential to observe so many phases of gasand dust simultaneously is the far-IR, which similarly requires access to space.In cooler regimes of the interstellar medium (ISM), dust scatters and absorbsfar-UV (FUV) photons, which provides a UV background that changes thechemistry of galactic ISMs (e.g., Hamden et al. 2013). While dust is oftenseen as a hindrance to UV observations, such sightlines can be used to sensi-tively measure fine dust chemistry and distributions in different astrophysicalenvironments (e.g., Blasberger et al. 2017; Ma et al. 2020). Similarly, UV spec-troscopy of relatively high reddedning (E(B-V) > International UltravioletExplorer ( IUE ) and Copernicus revolutionizing our understanding of the dif-fuse and translucent ISM physical state and chemistry using UV absorptionline spectroscopy (e.g., Spitzer et al. 1973; Morton 1975; Savage et al. 1977;Bohlin et al. 1983, and many others). The UV is the only wavelength regimewe gain access to cold molecular hydrogen (H ; T(H ) <
500 K), and this initself is a critical diagnostic for many astrophysical environments and proper-ties that rely on accurate measures of H density or mass. UV observationsprovide a critical look at the characteristics of many astrophysical systems,both nearby and afar.Present-day ultraviolet spectroscopic facilities on community space tele-scopes (currently, only the Hubble Space Telescope ( HST ) Cosmic Origins Spec-trograph (COS; Green et al. 2012) and Space Telescope Imaging Spectrograph(STIS; Kimble et al. 1998) provide spectroscopic coverage in the FUV) haveprovided a unique view of many astrophysical systems, but they have their lim-itations and cannot address all our remaining questions. Current space facility
HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 3 spectrographs rely on old technologies in reflective mirror coatings, gratings,and detectors that all perform moderately in the UV - many gains in improve-ment can be made in almost all realms of UV instrumentation to probe deeperand fainter objects in our universe. Many of the most interesting, intriguing,and mysterious diffuse objects discovered by the
Galaxy Evolution Explorer ( GALEX ) a decade ago (e.g., Martin et al. 2007; Sahai & Chronopoulos 2010;Hoadley et al. 2020) are unattainable with UV spectrographs on
HST , whichdo not have the sensitivity to observe them. Highly sensitive, high-resolutionUV spectroscopy is necessary to address astronomical questions that are be-yond the reach of present-day instruments. In this paper, we discuss a conceptfor a high-throughput, high-resolution (R > Colorado High-resolution Echelle Stellar Spectrograph (CHESS) is an ob-jective echelle spectrograph designed to achieve a minimum resolving power(R) > − ≤ Keri Hoadley et al.
Translucent clouds reside in the transition between the diffuse (tradi-tionally defined as the visual dust extinction (A V ) <
1) and dense (A V >
3) phases of the interstellar medium (ISM). It is in this regime where theUV portion of the average interstellar radiation field plays a critical role inthe photochemistry of the gas and dust clouds that pervade the Milky Waygalaxy. One powerful technique for probing the chemical structure of translu-cent clouds is to combine measurements of H with knowledge of the full carboninventory (CI, CII, and carbon monoxide (CO)) along a given line of sight.Snow & McCall (2006) argue that an analysis of the carbon budget should bethe defining criterion for translucent clouds, rather than simple measurementsof visual extinction. Moderate resolution 1000-1120 ˚A spectra from FUSE andhigher-resolution spectra from HST/STIS can confirm whether or not a sight-line is consistent with the existence of translucent material in the frameworkof current models of photodissociation regions of the ISM (CO/H > − andCO/CI ∼
1; Burgh et al. 2007, 2010).The bandpass of CHESS contains absorption lines of H (1000 − − < > The Local Interstellar Medium (LISM) provides an opportunity tostudy general ISM phenomena up close and in three dimensions, includinginteractions of different phases of the ISM, cloud collisions, cloud evolution,ionization structure, thermal balance, and turbulent motions (Redfield, 2006).Our immediate interstellar environment also determines the structure of theheliosphere, or the momentum balance of the solar wind and the surroundingISM. Several physical characteristics of the LISM are measurable, including theionization structure. Since many clouds in the LISM are optically thin, the dis-tribution of ionizing sources (i.e., hot stars) determines the three-dimensionalionization structure of the LISM.Measurements of different ionization species are required to probe differ-ent phases of the LISM. In addition to local ionization structure, temperatureand elemental depletion structure are also critical to understanding the three-dimensional morphology of the LISM. The temperature distribution of theLISM can place constraints on models of the evolution of the local solar neigh-borhood. Determining these temperatures requires high spectral resolution sothat contributions from thermal and turbulent motions can be distinguished,a capability that is achievable with CHESS.
HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 5
CHESS is an objective f/12.4 echelle spectrograph. The instrument designincluded the development of two novel grating technologies and flight-testingof a cross-strip anode microchannel plate (MCP) detector (Beasley et al.,2010). The high-resolution instrument is capable of achieving resolving powers ≥ λ / ∆λ across a bandpass of 1000 − – A mechanical collimator, consisting of an array of 10.74 mm × × , a field of view (FOV) of 0.67 ◦ , and allows on-axisstellar light through to the spectrograph. A mechanical collimator designwas chosen over a traditional telescope design for practicality, while stillachieving the necessary sensitivity for the spectrograph. A collimating tele-scope would be difficult to align to the spectrograph and remain alignedduring the course of its flight. A wire mesh collimator was considered, butthe total loss of collecting area was too high to implement. The large FOVof the mechanical collimator was dictated by the physical dimensions ofcommercially-available anodized aluminum tubes. Because CHESS looksat bright FUV point sources (stars) to provide a “backlight which illumi-nates the interstellar gas in the sightline, there was very little chance ofsource confusion over this FOV, given the effective area of the spectrographand the brightness of the target stars. – A square echelle grating (ruled area: 100 mm ×
100 mm), with a designedgroove density of 69 grooves/mm and angle of incidence (AOI) of 67 ◦ , inter-cepts and disperses the FUV stellar light into higher diffraction terms (m= 266 − – Instead of using an off-axis parabolic cross disperser (Jenkins et al., 1988),CHESS employs a holographically-ruled cross dispersing grating with atoroidal surface figure and ion-etched grooves, maximizing first-order ef-ficiency. The cross disperser is ruled over a square area (100 mm × ρ ) = 2467.96mm. The grating spectrally disperses the echelle orders and corrects forgrating aberrations (Thomas, 2003). – The cross-strip MCP detector (Vallerga et al., 2010; Siegmund et al., 2009)is circular in format, 40 mm in diameter, and capable of total global countrates ∼ counts/second. The cross-strip anode allows for high resolutionimaging, with the location of a photoelectron cloud determined by the Keri Hoadley et al. centroid of current read out from five anode “fingers” along the x and yaxes.Because the echelle disperses light into high orders and the cross disperserseparates light sharing the same echelle diffraction order solutions, the finaldata product is a series of spectra, where each echelle spectrum provides a smallfraction of the total spectral coverage of the instrument. At the same time, eachspectral snippet in the full raw data is able to be sampled at high resolution.This is how CHESS is able to achieve both large wavelength coverage and highresolution.The CHESS instrument also includes an optical system (Xybion ModelISS-255 low-light video camera ), which was used to align the spectrograph toan independent star tracking system during calibrations and flight. The aspectcamera relies on the positions of the instrument gratings to direct zeroth-orderlight to a visible-light camera. Fig. 1 shows a Zemax ray trace of collimated(star) light through the entire spectrograph.CHESS was separated into two sections to fly aboard a sounding rocketmission: a vacuum (spectrograph) and non-vacuum (electronics) section. Thetwo sections are separated by a hermetic bulkhead. The detector was mountedwith a hermetic seal on the electronics side of the vacuum bulkhead, facinginto the spectrograph section. The overall length of the payload was 226.70cm from mating surface to mating surface, with a total weight of about 365lbs. The opto-mechanical design of the spectrograph consisted of a carbon-fiberspace frame attached to the vacuum side of the hermetic bulkhead. The carbonfiber frame (five 2.54 cm diameter x 182.88 cm long tubes) holds three light-weighted aluminum disks which suspend the mechanical collimator, echellegrating and cross-disperser in place. The aspect camera is attached to thevacuum side of the hermetic bulkhead. A Solidworks rendering of the payloadis shown in Figure 2.2.2 Components & PerformanceThe CHESS instrument is comprised of three optical components: two gratings(one high-dispersion echelle grating and one cross-dispersing grating) and onedetector (cross-strip anode MCP). Table 1 presents technical details abouteach optical element used for the first two flights of the instrument. Echelle gratings are distinguishable by their low line densities (20 - 300 lines/mm)and use at steep facet angles ( θ : 20 ◦ - 80 ◦ ). Both qualities allow echelle gratings At the time of CHESS-1 and 2, the Xybion Electronics System (XES) low-light camerawas deprecated and the NASA Sounding Rocket Operations Contract (NSROC) was in theprocess of replacing it with a newer model of space-qualified low-light camera system. In2018, NSROC officially replaced the XES Model ISS-255 with the Stanford Photonics Inc.XR/M model:
HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 7
Table 1
Instrument Specifications for the CHESS Sounding Rocket PayloadMechanical Collimator SpectrographFOV: 18.5 (cid:48) × (cid:48) Bandpass (˚A): 1000 - 1650Dimensions (mm): Resolving power:10.74 × × ∼ ): 40.0 Demonstrated R ≤ F f /12.4Echelle (CHESS-1) Echelle (CHESS-2) Cross Disperser DetectorVendor: LightSmyth Vendor: Bach Research Vendor: HORIBA Jobin-Yvon Vendor: Sensor SciencesShape: Flat Shape: Flat Shape: Toroidal Type: Open-face MCPBlaze angle( ◦ ): 73.0 Blaze angle( ◦ ): 64.3 Radius (mm): 2500.25/2467.96 Pixel format: 8k × µ m):71.7 53.85 351 25Ruling: Lithographic Ruling: Mechanical Ruling: Holographic Anode: Cross-stripCoating: Al+LiF Coating: Al+LiF Coating: Al+LiF Photocathode: CsIDimensions (mm): Dimensions (mm): Dimensions (mm): Outer dimension (mm): 40100 × × × ×
16 100 × ×
30 Global count rate (Hz): 10 Material: Silicon Material: Zerodur Material: Fused Silica to theoretically achieve high dispersion, high efficiency at or near the Littrowconfiguration, or where the angle of incidence equals the diffraction angle ( α = β = θ ), and high resolution with low polarization effects.As a part of the instrument design, the echelle was meant to be an exper-imental technology demonstration piece for two different grating fabricationprocesses: the first was a lithographic-ruling process, provided by LightSmyth,Inc., and the second an electron-beam etching technique, fabricated by theMicrodevices Laboratory at JPL. The lithographic ruling process starts witha substrate (silicon) with a thin oxidation layer, over which a photoresist andphotomask with the desired groove pattern is overlaid. Using extreme-UV lightto etch into the photoresist, the photomask is removed and the oxide layer isfurther etched via chemical agents that do not harm the photoresist. After re-moving the rest of the photoresist, the etched substrate is left with the desiredgroove density and thickness. The lithography process theoretically allows forthe creation of uniform, low-scatter gratings at arbitrary groove densities withsub-100 nm surface deviations. The electron-beam etch scans a focused beamof electrons across the surface of the optic, which is covered with an electronsensitive film. The electron beam changes the solubility of the resist, whichremoves exposed regions of the resist in a solvent (McCord & Rooks, 2000).This technique also enables controlled line spacing on the grating and sub-10nm surface deviations for low scatter grooves.We present results from in-house determination of the groove efficiency ofthe best echelle gratings fabricated by both the lithography and electron-beametching processes in Table 2 and Figure 3. The high angle of incidence andhigh diffraction orders we were working with in CHESS made the the fabrica-tion process of the echelle gratings extremely difficult, and both manufactur-ers were unable to produce echelle gratings that met our flight specification Keri Hoadley et al.
Table 2
Comparison of Echelle Performance for CHESS (Hoadley et al., 2016)Grating α Groove Density Ly α Efficiency(degrees) (grooves/mm)
CHESS, Designed 67.0 69.0 70.0 %
LightSmyth, CHESS-1 73.0 71.7 1.5 %JPL Echelle 65.5 100.0 4.5 %Bach, CHESS-2 (cid:63) (cid:63) in the limited time allowed before scheduled launches. For the first flight ofCHESS, we had not identified a back-up grating to use for flight, so we flewthe lithographically-ruled echelle. For CHESS-2, we identified two mechanical-ruling manufacturers that provided higher efficiency gratings for the secondflight. The Bach echelle was delivered in time for the second flight, while theRichardson echelle was used for subsequent flights (Kruczek et al., 2017, 2018).
The CHESS cross disperser grating is a 100 mm ×
100 mm ×
30 mm fusedsilica optic with a toroidal surface profile. The toroidal surface shape sepa-rates the foci of the tangential and sagittal axes of the dispersed light, whichcorrects astigmatic aberrations typically introduced by a more traditional off-axis parabolic design (e.g., Sasian 1997; Jenkins et al. 1996; Indebetouw et al.2001). The surface figure of the toroid focuses the echelle order widths at theposition as the spectral line widths, which are dictated by the grating solu-tions etched into the cross disperser. Both the sagittal and tangential foci ofthe toroidal optic do not intersect either the ion repeller or quantum efficiency(QE) grids, both of which are placed in front of the detector. The cross dis-persing optic is a novel type of imaging grating that represents a new family ofholographic solutions and was fabricated by Horiba Jobin-Yvon (JY). The linedensities are low (351 lines per mm, difficult to achieve with the ion-etchingprocess), and the holographic solution allows for more degrees of freedom thanwere previously available with off-axis parabolic cross dispersing optics. Theholographic ruling corrects for aberrations that otherwise could not be cor-rected via mechanical ruling. The grating is developed under the formalism oftoroidal variable line spacing (VLS) gratings (Thomas, 2003) and correspondsto a holographic grating produced with an aberrated wavefront via deformablemirror technology. This results in a radial change in groove density and a tra-ditional surface of concentric hyperboloids from holography (e.g., HST/COS;Green et al. 2012).Figure 4 shows the measured reflectivity (order efficiency times the reflec-tivity of the optical coating) of the cross dispersing optic for order m = -1,which is the dispersion order used in the CHESS instrument. We measurethe reflectivity of the cross disperser before the launch of CHESS-1, between
HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 9 the launch of CHESS-1 and CHESS-2, and after the launch of CHESS-2, andfound that it did not change significantly over time. The cross disperser iseffective at dispersing most of the on-axis light into the m = ± ± The cross-strip MCP detector was built and optimized to meet the CHESSspectral resolution specifications at Sensor Sciences (Vallerga et al., 2010; Sieg-mund et al., 2009). The detector has a circular format and a diameter of 40mm. The microchannel plates are lead silicate glass, containing an array of 10-micron diameter channels. They are coated with an opaque cesium iodide (CsI)photocathode, which provides QE = 15 −
40% at FUV wavelengths. When UVphotons strike the photocathode to release photoelectrons, the photoelectronsare accelerated down the channels by an applied high voltage ( ∼ µ m over an 8k x 8k pixel format. The detector quantumefficiency (DQE) was measured by Sensor Sciences and is shown in Figure 5(left). The designed throughput of CHESS was set assuming that certain milestonesin grating etching processes, specifically higher efficiency, low scattered lightechelle grating, were achieved. CHESS was designed as a 2-bounce spectro-graph to minimize the total reflective surfaces and maximum throughput ofthe instrument. The specified echelle diffraction efficiency into Littrow or-ders through the CHESS bandpass was defined at 70%, a performance whichmatches echelle counterparts in visible and infrared spectrographs. The cross-strip MCP was coated with a photocathode to increase the DQE to ∼ ∼
50% dispersion efficiency through the FUV. The reflective coatingon both gratings is lithium fluoride over an aluminum layer (Al+LiF), whichis ∼
70% reflective for λ > ∼ , the designed average effective area of CHESSwas meant to be 1.75 cm .The component-level performance and total throughput of CHESS-1 andCHESS-2, defined as the effective area (the total throughput of the instrument ×
40 cm collecting area), are shown in Figure 5. The low effective area of the instrument was driven by the efficiency of the echelle grating, which under-performed our specifications. The CHESS-2 effective area improved by roughlyan order of magnitude from CHESS-1 by using a traditionally-ruled echellegrating, and subsequent flights of CHESS have improved the throughput bycontinuing to exchange echelle gratings (Kruczek et al., 2017). Pre- and post-flight calibrations of CHESS were performed in a dedicatedvacuum chamber for sounding rocket payloads located in the AstrophysicalResearch Laboratory at the University of Colorado - Boulder. The chamberpumps the payload down to pressures of < − Torr, allowing for UV lightto transmit through the payload and safe operation of electronics requiringhigh voltage, such as the MCP detector. Because CHESS was designed as ahigh-resolution spectrograph with a broad UV bandpass, a hollow-cathode arclamp supplied with high purity H+Ar gas produced UV light used to focusand calibrate the instrument. This gas not only provided lines of hydrogen andargon, but a myriad of lines from H fluorescence. Figure 6 shows the best-focused calibration images from CHESS-1 and CHESS-2 pre-flight spectra.The broadest emission features show atomic lines of hydrogen and argon, whilethe narrower features are H fluorescence lines excited by interactions withenergetic electrons within the lamp.Calibration images like those shown in Figure 6 were used to convert theCHESS echellogram to a 1-dimensional spectrum, defining both the wave-length solution and line spread function (LSF) of the instrument. First, eachorder dispersed by the cross disperser was identified and extracted to collapseinto a 1D spectrum. Then, using adjacent order spectra, the orders were cross-correlated and stitched together to create a complete spectrum from 1000 -1600 ˚A. The cross-correlation was a (roughly) linear function from orders withshorter wavelengths to longer wavelength orders, so adjacent orders with nooverlapping features could still be co-added into the complete spectrum. Asynthetic H model of electron-impact fluorescence was used to match H fea-tures in the CHESS spectrum to the laboratory wavelengths of H emission -this created the wavelength solution of the instrument, as shown in Figure 7.Once a wavelength solution was found, we convolved each synthetic emis-sion line with a Gaussian of unknown width to match the CHESS emission lineprofiles, thereby creating LSF kernels through the CHESS bandpass. Each linewas best fit with a narrow and broad Gaussian component. Figure 8 shows aclose-up of a few emission lines with a two-component Gaussian fit of the H lines imaged by CHESS.The “broad” component of the CHESS LSF was unexpected and severelyimpeded the spectral resolution of the instrument. CHESS was designed withR ≥ ≤ HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 11 was the culprit of the degraded resolution, as it was etched onto a thin waferthat could bow or warp the surface of the grating.It was determined during the calibration of CHESS-2 that the broad emis-sion lines were the result of the cross dispersing grating - the toroidal shape ofthe optic was designed to focus and correct for optical aberrations, but the rul-ing of the grating occurred along the optical axis perpendicular to the radius ofcurvature, which meant that, while the order widths were minimized to avoidorder confusion, the spectral lines were out-of-focus. We verified this was theculprit by flipping the grating axis of the cross disperser in our Zemax models,which perfectly reproduced the line shapes and measured spectral resolutionof the instrument (Kruczek et al., 2018, 2019). While the cross disperser wasnever re-fabricated, Kruczek et al. 2018 addressed the spectral resolution issueby adding a slight curvature to the high-dispersion echelle grating, achievingR ∼ < α Virgo
The first flight of CHESS targeted the β Scorpii ( β Sco ; HD 144217) sightline. β Sco is spectroscopic binary comprised of a B0.5V and a B1.5V spectral typestar at distance of 161 pc with intermediate reddening (E(B-V) = 0.2, A V ∼ ) ∼ − (Savage et al., 1977)).However, given the poor performance of the exerimental echelle gratingused in CHESS-1, the science team elected to define a back-up target to moveto if the signal-to-noise (S/N) of β Sco was insufficient to produce a science-quality spectrum during the limited time of the flight. The back-up target, α Virgo ( α Vir; HD 116658; “Spica”), is a spectroscopic binary, consisting of aB1III-IV star + B2V star at a distance of 43 pc (Hoffleit & Jaschek, 1982),well within the Local Bubble (E(B-V) = 0.03, A V ∼ α Vir outputs 4 − β Sco and thus demonstratedthe capabilities of the CHESS instrument.CHESS-1 (NASA/CU mission 36.285 UG) was launched from White SandsMissile Range on 24 May 2014 at 01:35am using a two-stage Terrier/BlackBrant IX vehicle. Data was downlinked through the NASA telemetry systemin real-time as [ x , y , t , P HD ] photon lists, where [ x , y ] defines the digitalpixel position of the photon recorded on the detector, t is the time the photonwas recorded, and P HD is the pulse height recorded for the photon, which isrelated to detector gain. Overall, the mission was a comprehensive success andachieved all the goals it aimed to meet. The instrument successfully collecteddata over the allotted ∼
400 seconds of observing time. After the initial countrate on beta Sco was lower than expected, we elected to observe α Vir, andstayed on this target for the remainder of the flight. Count rate from α Vir wasalso lower than expected, but stellar absorption features in the echellogrammade it clear that we were observing the star. Additionally, count rate contri-bution from air glow contamination was lower than expected, which suggested that at least one of the optics became unaligned with the instrument betweenpre-flight operations and launch. Post-launch calibrations verified that one ofthe gratings shifted before flight – the echellogram did not appear on the de-tector until the instrument was shifted off-axis from the vacuum light sourceby ∼ α Vir from the CHESS-1 launch. Promi-nent features apparent in the echellogram include stellar Ly α , Si II (1193 ˚A),Si III (1206 ˚A), and C III (1175 ˚A). However, the S/N of the flight data werenot adequate to reconstruct an analysis-quality 1D spectrum. (cid:15) Persei
CHESS-2 was launched aboard a Terrier-Black Brant IX sounding rocket fromWhite Sands Missile Range on 21 February 2016 as a part of NASA/CU mis-sion 36.297 UG. CHESS-2 observed the line of sight to (cid:15)
Persei ( (cid:15)
Per; HD24760). (cid:15)
Per is a B0.5III star at d ≈
300 pc with low − intermediate red-dening (E(B-V) = 0.1; log(H ) ∼ , C I, CO, and C II were all detectedby Copernicus; however, higher sensitivity and spectral resolution is requiredfor a complete analysis of these types of sightlines (Federman et al., 1980).Observations by Copernicus and IUE have been used to measure the velocitystructure along the sightline to (cid:15) Per, and have found at least three sepa-rate cloud structures described by different kinematic behavior and molecularabundances (Bohlin et al., 1983). Resolving the various molecular clouds onthe (cid:15)
Per sightline is the primary goal of CHESS-2. Overall, the line of sightto (cid:15)
Per shows typical abundances of molecular material and ionized metalfound in translucent clouds, such as H , Fe II and Mg II (Bohlin et al., 1983),consistent with the sightline towards recent star-forming sites.The observed count rate on-target was ∼ − for a totalexposure time (t exp ) of 250 seconds, resulting in a signal-to-noise ratio (S/N)per spectral resolution element of ∼ λ = 1020 - 1060 ˚A and S/N (cid:38) λ > (cid:15) Per is shown in Figure 10.5.1 Analysis of CHESS-2 dataThe low S/N of the final flight data and diminished resolving power due to thefabrication error in the cross disperser meant that many of the primary sciencegoals of CHESS were not achieved. We were unable to resolve any carbon-complex lines or achieve the signal needed to distinguish CO band absorption.However, the CHESS-2 spectrum of (cid:15)
Per does show many absorption featuresbetween 1030 - 1220 ˚A, which we fit and analyzed to compare to previousobservations of this sightline taken with Copernicus and
IUE .Scattered light from geo-coronal contamination posed a non-trivial amountof noise in the raw echellogram of CHESS-2. A combination of simulations
HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 13 and laboratory measurements were used to estimate the scattered light back-ground in the raw data. Scattered light simulations were performed using theray-tracing software Zemax using the average flux of geo-coronal HI-Ly α atnight (63 × − erg cm − s − arcsec − ), and scattered light images weretaken using an uncollimated Bayard-Albert tube fed with hydrogen-argon gas(France, 2006). Both produced consistent scattered light profiles, which wereused to model the background noise of the flight echellogram. We note thateven after extracting the modeled scattered light profile, it appears that anunderlying scattered light residual still remained, which appears as excess fluxin absorption features through the 1D spectrum. For example, several knownsaturated absorption features are not entirely bottomed-out to the zero-pointof flux scale. HST -STIS suffers the same data reduction challenges due to thelarge scattered light profile of their mechanically-ruled gratings (Landsman& Bowers, 1997). Initial data reduction packages for STIS produced similarresults - for example, either the over- or under-subtraction of background light(e.g., McGrath et al. 1999). STIS handles the complicated echelle-dominatedbackground light profile with a 2-D algorithm that still has issues completelygetting rid of the scattered light noise (Valenti et al. 2003 and referencestherein). This excess background noise in the CHESS observations is difficultto completely get rid of due to the low S/N of the observation, so we add thisextra noise source into the errors in our analysis.
We visually inspect the CHESS-2 spectrum of (cid:15)
Per against the
IUE andCopernicus spectra of the sightline. From this inspection, we select a numberof stellar and interstellar absorption features that match well between all datasets. Next, we normalize the continuum of all data sets around the absorptionlines of interest by fitting an arbitrary quadratic function through continuumpoints surrounding the absorption feature. Each absorption feature is fit witha Gaussian line profile using a reduced- χ statistic. The initial conditions ofthe Gaussian fit are found to not drastically change the final result, so all lineprofiles begin with the same width parameter ( σ ). We then find the equivalentwidth ( W λ ) of the each absorption line by finding the area under each Gaussianprofile. This entire process is repeated for the same absorption lines found in IUE and Copernicus spectra.Overall, we find good agreement between all three data sets. Figure 12presents the normalized absorption lines for a large selection of metal featurefrom all three data sets, and Table 3 describes the quantities derived fromeach line profile, including the full width half maximum (FWHM) of eachabsorption line fit and W λ derived from the line fits. Table 4 presents columndensity estimates from CHESS-2 and our re-analysis of IUE and Copernicusspectra of (cid:15)
Per, along with the total column density of each species determinedby Martin & York (1982) when available. We estimate a total column densityof the species in the sightline ( N ( X )) by assuming the lines fit along the linearpart of the curve of growth. We find good agreement in our estimates of N ( X ) Table 3 (cid:15)
Persei Metal Absorption Line Diagnostics & Comparison with Archival
IUE andCopernicus Spectra CHESS-2
IUE
CopernicusIon λ f
FWHM W λ χ FWHM W λ χ FWHM W λ χ (˚A) (km/s) (m˚A) (km/s) (m˚A) (km/s) (m˚A)H I 1025.72 0.079 1011 3646.6 156.60 743 3373.2 17.00H I 1215.67 0.416 1835 7701.1 247.06 1620 6194.6 54.26C II 1334.53 0.129 131 190.8 1.21 99 347.6 0.51 124 360.0 1.13C II 1335.71 0.115 69 208.6 1.25 63 448.6 0.99 93 316.9 1.35C III 1175.26 0.272 440 624.1 31.73 561 2569.3 13.12 521 2763.5 11.8N I 1134.98 0.040 75 35.1 3.04 60 95.5 1.15N I 1199.26 0.133 37 36.3 3.27 40 150.5 7.67 53 84.7 6.23N II 1083.99 0.101 65 74.0 2.03 157 121.9 1.16N V 1238.82 0.156 113 209.3 12.12 283 339.6 10.28 325 407.9 5.27N V 1242.80 0.078 270 330.6 12.32 220 509.6 13.37 307 322.7 7.28O I 1302.17 0.052 218 302.8 15.59 160 430.1 10.41 242 467.2 50.29S II 1259.52 0.016 64 69.0 2.22 28 42.6 2.23 52 66.6 1.07Si II 1194.50 1.62 96 122.9 3.11 113 348.5 14.34 303 267.6 5.34Si II 1197.39 0.323 69 107.9 12.13 92 161.8 17.18 101 260.7 8.03Si III 1206.51 1.67 273 1151.2 2.94 473 1558.8 10.62 470 1337.1 5.78Si IV 1393.76 0.528 570 1999.4 9.48 648 2698.0 16.71 637 2673.1 9.78Si IV 1402.77 0.262 583 2077.3 5.68 481 2387.4 12.94 385 1906.1 5.15P II 1152.81 0.236 51 22.1 0.89P II 1301.87 0.017 37 38.4 2.06 45 37.5 0.04 50 68.2 0.48 Table 4 (cid:15)
Persei Metal Column Densitieslog N ( X ) log N ( X ) (total)Ion CHESS IUE
Copernicus Martin & York (1982)C II 16.3 +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . +0 . − . P II 13.6 +0 . − . +0 . − . +0 . − . (cid:63) Jenkins & Shaya (1979). and Martin & York (1982) for all lines that overlap. For saturated lines (e.g., HI), we use the intrinsic properties of the line to estimate the column density inthe sightline, which is the same approach used for our H analysis below (seeSection 5.1.2 for more details). We discuss the H I column density estimatefrom the CHESS-2 data in the following subsection. Absorption Profile Fitting, Rotation Diagrams, and Formation Rates
Molecular hydrogen (H ) is the most abundant constituent of interstellarclouds. Interstellar H is observed as absorption features throughout the far- HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 15
UV regime along sightlines toward hot stars. Many H interstellar sightlinesdisplay multiple temperature components (Spitzer et al., 1973), where thelow rotation ( J ) lines ( J = 0, 1) provide a measure of the kinetic (collision-dominated) gas temperature, while H probed in higher rotational levels ( J>
3) are sensitive to other physical processes, like UV-pumping (e.g., vanDishoeck & Black 1986), formation on dust grains (e.g., Jura 1974; Lacouret al. 2005) and heating by turbulence and/or shocks (e.g. Gry et al. 2002;Ingalls et al. 2011). The multi-component population structure provides im-portant diagnostics about which processes dominate the physical cloud condi-tions and individual interstellar cloud properties, such as the total hydrogendensity and the H formation rate.To analyze the properties of interstellar H in the (cid:15) Per sightline, we de-termine the H rotational structure following the procedure of France et al.2013. We fit a multi-component H absorption model to observed absorptionfeatures by combining the H ools optical depth templates (McCandliss, 2003)and the MPFIT least-squares minimization routine (Markwardt, 2009). Thismethod takes the theoretical line shape of each H energy level for a givenset of column density (N(H )) and Doppler- b values, convolves the syntheticspectrum with the line spread function (LSF) of the instrument, and adjustsN(H ) and b until a best-fit spectrum is found.The (cid:15) Per spectrum is normalized to our model spectra around select H absorption bands. We fit synthetic profiles simultaneously to the H (1 − λ (0 −
0) ( λ J (cid:48)(cid:48) = 0 − absorptionfeatures in our data set. To check the solutions of the H (1 −
0) and H (0 −
0) profile fits, we fit the modeled H column densities to the H (4 −
0) H band ( λ (cid:15) Per sightline may have upwards of fourinterstellar cloud components (Martin & York, 1982), the moderate spectralresolution of CHESS does not allow us to separate these different components.Therefore, assuming a single H cloud, we find a b -value of 3.6 km s − , de-termined from the H ( J (cid:48)(cid:48) = 2 −
6) rotational levels, which are sensitive tochanges in b . This value is consistent with previous curve-of-growth measure-ments of H in the (cid:15) Per sightline (e.g., Stecher & Williams 1967; Carruthers1971) and typical b -values for H in the local ISM (Lehner et al., 2003; Franceet al., 2013).We present the total H column density and densities in each J -level inTable 5. The normalized spectra and best-fit models for the H (4 −
0) andH (1 − (0 −
0) bands of (cid:15)
Per are shown in Figure 13. We determine thetotal column density of H ( N (H ) obs ) to be log N (H ) obs = 20.33 ± α absorption feature ( λ absorption fits (listed in Table 5). However, as noted in Diplas & Savage(1994), for B-stars, there is a non-negligible stellar contribution to the H -Ly α absorption profile, such that N (H ) obs = N (H ) stellar + N (H ) interstellar .For (cid:15) Per, Diplas & Savage (1994) found the total stellar H column density Table 5 (cid:15)
Persei H Parameters from CHESS-2 Observations and Model FitsH Level log N(H , v (cid:48)(cid:48) =0, J (cid:48)(cid:48) ) log N(H , v (cid:48)(cid:48) =0, J (cid:48)(cid:48) )MPFIT Model MCMC Model J (cid:48)(cid:48) = 0 19.20 +0 . − . J (cid:48)(cid:48) = 1 19.56 +0 . − . J (cid:48)(cid:48) = 2 17.35 ± J (cid:48)(cid:48) = 3 15.47 +0 . − . J (cid:48)(cid:48) = 4 14.75 +0 . − . J (cid:48)(cid:48) = 5 14.91 +0 . − . J (cid:48)(cid:48) = 6 13.12 +0 . − . J (cid:48)(cid:48) = 7 < N (H ) 19.72 ± ± N (H ) 20.31 ± ± ± ± exc ±
150 K b H ± − − f ( H ) 0.24 ± ± Rn × − s − n H
55 cm − to be log N (H ) stellar = 18.82, so the total interstellar column density ofneutral hydrogen is log N (H ) = 20.31 ± N (H ) and N (H )to calculate the molecular fraction ( f ( H )) of the interstellar cloud: f ( H ) = 2 N ( H ) N ( HI ) + 2 N ( H ) = 0 . ± . . (1)The homonuclear nature of the hydrogen molecule forbids radiative transitionsfrom J (cid:48)(cid:48) = 1 → J (cid:48)(cid:48) = 2 → A → ≈ × − s − ; Wolniewiczet al. 1998). Therefore, for a sightline with an appreciable density of hydrogenthrough the ISM, collisions are expected to control the level populations of thelower energy J (cid:48)(cid:48) states (0, 1, 2). We can therefore define a kinetic temperature, T , which describes the collision-dominated regime of the interstellar cloud. T is derived from the ratio of column densities in the J (cid:48)(cid:48) = 0 and J (cid:48)(cid:48) = 1levels: N ( J (cid:48)(cid:48) = 1) /N ( J (cid:48)(cid:48) = 0) = g g e ( − E /kT ) = 9 e ( − K/T ) (2)where g and g are the statistical weights in the J (cid:48)(cid:48) = 1 and 0 levels, respec-tively. We find that the kinetic temperature of the H in the (cid:15) Per sightline is T = 95 ± J (cid:48)(cid:48) >
3) can be fit withan “excitation” temperature, T exc , by determining the slope of the rotationdiagram between J (cid:48)(cid:48) = 3 −
7. A least-squares linear fit through these rotationlevels found T exc = 500 ±
150 K (Figure 14, left).We use the new results from the CHESS-2 observations to measure thesetwo interstellar physical cloud parameters in the (cid:15)
Per sightline. First, we de-rive Rn and n H from two empirical approaches: the first from Martin & York HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 17
Table 6
Physical Cloud Conditions of (cid:15)
Persei, Derived from Analytic SolutionsMethod: Martin & York (1982) Gry et al. (2002) Rn (s − ) 45.6 × − × − n aH (cm − ) 150 20 n bH (cm − ) 80 10 n cH (cm − ) 10 1.5 a R = 3.0 × − cm s − ; Jura (1975b). b R = 6.0 × − T / cm s − , assuming T = T = 95 K; Black & van Dishoeck (1987). c R = 7.3 × − (T / 100) / (cid:104) (cid:15) gr (cid:105) Σ − cm s − , assuming T = T = 95 K, Σ − =the grain surface area for silicate-graphite PAHs = 6.0 ( × − ) cm H − , and (cid:104) (cid:15) gr (cid:105) = the average efficiency for H formation on dust grains ∼ n H ; Weingartner & Draine (2001); Draine (2011). (1982), who use an updated prescription to a common approach (e.g., Jura1975b) to define diffuse ISM physical cloud conditions, and the second fromGry et al. (2002), who take into account additional physical mechanisms impor-tant for denser clouds, like dust and molecular shielding and irradiation. TheMartin & York (1982) analysis calculates Rn based on the formalism presentedin Jura (1975b): Rn ≈ N (H ; J=4) / N (H )(4.28 × ). Using the columndensities derived from the CHESS-2 observation, we find Rn = 45.6 × − s − . The Gry et al. (2002) method balances formation and destruction rates ofH : Rn = 0.5 f − f β (cid:104) S (cid:105) , where f is the molecular fraction of H in the inter-stellar cloud, β is the mean rate of H photo-dissociation in the diffuse localinterstellar medium, and (cid:104) S (cid:105) is the mean self-shielding factor of H againstphoto-dissociation from the UV radiation source. For the diffuse ISM, β isdefined as 5.0 × − s − (e.g., Habing 1968; Mathis et al. 1983). We find (cid:104) S (cid:105)≈ × − using Equation 37 in Draine & Bertoldi (1996), which dependson N (H ) and b derived from our synthetic absorption spectra (Section 5.1.2).We find Rn ≈ × − s − using the Gry et al. (2002) approach, whichis ∼ Rn derived using the Martin & York (1982) method.Table 6 presents the H formation rates and resulting realizations of n H using R found from Jura (1975b), Black & van Dishoeck (1987), and Draine (2011).There are roughly two dex in spread for n H between these two methods,depending on the adopted H formation rate. To better constrain Rn and n H in the (cid:15) Per sightline, we create stationary H equilibrium models, whichpopulate H energy levels based on the physical conditions of an interstel-lar cloud. We follow the framework presented by Draine & Bertoldi (1996)(Equations 1 - 14), Gry et al. (2002), and references therein. The models bal-ance excitation and de-excitation processes working on H , including formationpumping, H dissociation, photo-excitation, spontaneous decay, and collisionalde-excitation.The UV radiation field is assumed to be dominated by the blackbody UVcontinuum produced by (cid:15) Per. Interstellar H is excited to higher energy elec-tronic levels by this UV radiation, where it either fluoresces back to the groundelectronic energy state or dissociates. Energy levels, transition probabilities,and dissociation probabilities for fluorescent transitions are adapted from Ab- grall & Roueff (1989), Abgrall et al. (1992, 1993a), and Abgrall et al. (1993b).Spontaneous decay rates of ground-level quadrupole transitions are taken fromWolniewicz et al. (1998). Finally, collisional de-excitation of H is calculatedfrom equations and coefficients presented by Mandy & Martin (1993) andMartin & Mandy (1995), where hydrogen is the dominate collisional partnerof H .We allow Rn and n H to float but constrain the values of n H between theminimum and maximum values presented in Table 6. We assume that n (H), n (H ), and n H are uniformly distributed across the interstellar cloud slab,which allows us to retrieve N (H ) and N (H I) from our simulations, and Rn does not vary across the modeled region. The models are run through a MarkovChain Monte Carlo (MCMC) routine, performed with the Python emcee pack-age (Foreman-Mackey et al., 2013) in the same manner as described in Hoadleyet al. (2017). The routine is run with 100 walkers over 500 individual steps.The MCMC analysis produces comparable molecular and atomic hydrogencolumn densities compared to our direct observational results: log N (H ) =19.83 ± N (H I) = 20.43 ± (H ) = 95 ± Rn = 49.5 × − s − and n H = 55cm − . Our results are slightly larger values than diffuse cloud conditions of (cid:15) Per determined by Jura (1975a) and Martin & York (1982) and much largerthan the denser interstellar cloud schematic of Gry et al. (2002), making ourresults consistent with a diffuse interstellar cloud origin (Jura, 1975b). Ourresults find that the volumetric formation rate of H in the (cid:15) Per sightline is R = 9.0 × − cm s − . Assuming the gas density is uniform, we find thatthe interstellar cloud probed in the sightline extends ∼ equilibrium models, along with the CHESS-2 H rotation diagram for (cid:15) Per, in Figure 14 (right). Our resulting H forma-tion rate is slightly higher than the classic formation rates presented by Jura(1975b) and Black & van Dishoeck (1987) (Table 6). This can be accounted forin a few ways: either the efficiency of H formation on silicate-dominated dustgrains is higher than expected ( (cid:104) (cid:15) gr (cid:105) ∼ (cid:15) Per sightline deviates from the canonical distribution assumedfor the average Milky Way ISM (Weingartner & Draine, 2001).
The Colorado High-resolution Echelle Stellar Spectrograph (CHESS) is a UVspectroscopic instrument meant to demonstrate high-throughput, high-resolutionfar-UV spectroscopy. Such concepts are especially important for pathfindersof upcoming space missions, where future observatories will have to addresskey observational capabilities when
HST is no longer available. In addition,CHESS provides an ideal platform to improve and test experimental technolo-gies meant to vastly improve the performance of diffraction gratings, mirrorcoatings, and detector efficiencies throughout the UV.
HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 19
For the first two launches of CHESS, we tested and flew different high-order diffraction gratings (echelles) for UV spectroscopy and demonstrated theimproved performance and dynamic range of state-of-the-art MCP detectors.In the end, both the underwhelming performance of the experimental echellegratings and an error in the ruling of the toroidal surface figure cross-dispersinggrating proved detrimental to the overall performance of the spectrograph.Still, the first two launches of the instrument gathered data on the stellarsightlines they observed, with the second flight acquiring adequate S/N toperform a detailed analysis of the interstellar sightline. We find many differentmetallic absorption lines in the (cid:15)
Per sightline, and we fit multiple H bandcomplexes with thermal and stationary equilibrium models to characterize theH in this diffuse ISM sightline. We find the column density and temperatureof cool H to be consistent with analyses performed with Copernicus and findthat the full stationary equilibrium model suitably fits the observed columndensities of all J states.The CHESS experiment went on to observe nearby early B-type stars tostudy the molecular properties of local interstellar clouds aboard two addi-tional NASA-funded sounding rocket missions: the third launch of CHESStook place at WSMR in June 2017 and successfully observed the sightline to-wards β Sco (Kruczek et al., 2017), while the fourth and final launch took offfrom the Kwajalein Test Range in April 2018 and observed γ Arae (Kruczeket al., 2018). A detailed analysis of both sightlines, paired with an analysisof archival spectral datasets obtained with Copernicus and
FUSE , has shownthat previous studies overestimate the average gas kinetic temperature of thediffuse molecular ISM by 12% (Kruczek et al., 2019).We acknowledge the hard work and dedication of the NASA WFF/NSROCpayload team, the Physical Sciences Laboratory at New Mexico State Univer-sity, and the Navy team at WSMR that supported the NASA/CU 36.297 UG.We thank the referee of this manuscript for their helpful suggestions and pointswhere further clarification was warranted. KH would like to acknowledge thegenerous support and guidance from Prof. Jim Green, Ted Schultz, MichaelKaiser, and the University of Colorado UV sounding rocket research group dur-ing the build and operations of CHESS-I and CHESS-II. KH would also like tothank Dr. Nicholas Kruczek, Jacob Wilson, Jack Swanson, and Nicholas Er-ickson for each of their individual contributions and achievements that led to asuccessful CHESS-II flight, subsequent CHESS successes, and their invaluablemoral support before, during, and after the mission. KH acknowledges sup-port by the David & Ellen Lee Postdoctoral Fellowship in Experiment Physicsat Caltech. This research was funded by the NASA Astrophysics Researchand Analysis (APRA) grant NNX13AF55G. AY acknowledges support by anappointment to the NASA Postdoctoral Program at Goddard Space FlightCenter, administered by the Universities Space Research Association througha contract with NASA.
References
Abgrall, H., Le Bourlot, J., Pineau Des Forets, G., Roueff, E., Flower, D. R.,& Heck, L. 1992, A&A, 253, 525Abgrall, H. & Roueff, E. 1989, A&AS, 79, 313Abgrall, H., Roueff, E., Launay, F., Roncin, J. Y., & Subtil, J. L. 1993a, A&A,101, 273—. 1993b, A&A, 101, 323Beasley, M., Burgh, E., & France, K. 2010, in Proc. SPIE, Vol. 7732, SpaceTelescopes and Instrumentation 2010: Ultraviolet to Gamma Ray, 773206Black, J. H. & van Dishoeck, E. F. 1987, ApJ, 322, 412Blasberger, A., Behar, E., Perets, H. B., Brosch, N., & Tielens, A. G. G. M.2017, ApJ, 836, 173Bohlin, R. C., Jenkins, E. B., Spitzer, Jr., L., York, D. G., Hill, J. K., Savage,B. D., & Snow, Jr., T. P. 1983, ApJS, 51, 277Burgh, E. B., France, K., & Jenkins, E. B. 2010, ApJ, 708, 334Burgh, E. B., France, K., & McCandliss, S. R. 2007, ApJ, 658, 446Carruthers, G. R. 1971, ApJ, 166, 349Diplas, A. & Savage, B. D. 1994, ApJ, 93, 211Draine, B. T. 2011, Physics of the Interstellar and Intergalactic MediumDraine, B. T. & Bertoldi, F. 1996, ApJ, 468, 269Federman, S. R., Glassgold, A. E., Jenkins, E. B., & Shaya, E. J. 1980, ApJ,242, 545Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. 2013, PASP,125, 306France, K., Beasley, M., Kane, R., Nell, N., Burgh, E. B., & Green, J. C.2012, in Proc. SPIE, Vol. 8443, Space Telescopes and Instrumentation 2012:Ultraviolet to Gamma Ray, 844305France, K., Fleming, B., & Hoadley, K. 2016a, Journal of Astronomical Tele-scopes, Instruments, and Systems, 2, 041203France, K., Hoadley, K., Fleming, B. T., Kane, R., Nell, N., Beasley, M., &Green, J. C. 2016b, Journal of Astronomical Instrumentation, 5, 1640001France, K., Nell, N., Kane, R., Burgh, E. B., Beasley, M., & Green, J. C. 2013,ApJL, 772, L9France, K. C. 2006, PhD thesis, The Johns Hopkins UniversityGreen, J. C., Froning, C. S., Osterman, S., Ebbets, D., Heap, S. H., Lei-therer, C., Linsky, J. L., Savage, B. D., Sembach, K., Shull, J. M., Sieg-mund, O. H. W., Snow, T. P., Spencer, J., Stern, S. A., Stocke, J., Welsh,B., B´eland, S., Burgh, E. B., Danforth, C., France, K., Keeney, B., McPhate,J., Penton, S. V., Andrews, J., Brownsberger, K., Morse, J., & Wilkinson,E. 2012, ApJ, 744, 60Gry, C., Boulanger, F., Nehm´e, C., Pineau des Forˆets, G., Habart, E., &Falgarone, E. 2002, A&A, 391, 675Habing, H. J. 1968, BAIN, 19, 421Hamden, E. T., Schiminovich, D., & Seibert, M. 2013, ApJ, 779, 180
HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 21
Hoadley, K., France, K., Arulanantham, N., Parke Loyd, R. O., & Kruczek,N. 2017, ArXiv e-printsHoadley, K., France, K., Kruczek, N., Fleming, B., Nell, N., Kane, R., Swan-son, J., Green, J., Erickson, N., & Wilson, J. 2016, in Proc. SPIE, Vol. 9905,Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Se-ries, 99052VHoadley, K., France, K., Nell, N., Kane, R., Schultz, T., Beasley, M., Green,J., Kulow, J., Kersgaard, E., & Fleming, B. 2014, in Proc. SPIE, Vol. 9144,Space Telescopes and Instrumentation 2014: Ultraviolet to Gamma Ray,914406Hoadley, K., Martin, D. C., Metzger, B. M., Seibert, M., McWilliam, A.,Shen, K. J., Neill, J. D., Stefansson, G., Monson, A., & Schaefer, B. E.2020, Nature, Accepted for PublicationHoffleit, D. & Jaschek, C. 1982, The Bright Star Catalogue. Fourth revisededition.Indebetouw, R., McLean, R., Wilkinson, E., Green, J., & Beasley, M. 2001,Review of Scientific Instruments, 72, 1850Ingalls, J. G., Bania, T. M., Boulanger, F., Draine, B. T., Falgarone, E., &Hily-Blant, P. 2011, ApJ, 743, 174Jenkins, E. B., Joseph, C. L., Long, D., Zucchino, P. M., & Carruthers, G. R.1988, in Proc. SPIE, Vol. 932, Ultraviolet technology II, ed. R. E. Huffman,213–229Jenkins, E. B., Reale, M. A., Zucchino, P. M., & Sofia, U. J. 1996, A&SS, 239,315Jenkins, E. B. & Shaya, E. J. 1979, ApJ, 231, 55Jenkins, E. B. & Tripp, T. M. 2001, ApJS, 137, 297Jura, M. 1974, ApJ, 191, 375—. 1975a, ApJ, 197, 581—. 1975b, ApJ, 197, 575Kane, R., Beasley, M., Green, J., Burgh, E., & France, K. 2011, in Proc. SPIE,Vol. 8145, Society of Photo-Optical Instrumentation Engineers (SPIE) Con-ference Series, 81450PKimble, R. A., Woodgate, B. E., Bowers, C. W., Kraemer, S. B., Kaiser, M. E.,Gull, T. R., Heap, S. R., Danks, A. C., Boggess, A., Green, R. F., Hutchings,J. B., Jenkins, E. B., Joseph, C. L., Linsky, J. L., Maran, S. P., Moos,H. W., Roesler, F., Timothy, J. G., Weistrop, D. E., Grady, J. F., Loiacono,J. J., Brown, L. W., Brumfield, M. D., Content, D. A., Feinberg, L. D.,Isaacs, M. N., Krebs, C. A., Krueger, V. L., Melcher, R. W., Rebar, F. J.,Vitagliano, H. D., Yagelowich, J. J., Meyer, W. W., Hood, D. F., Argabright,V. S., Becker, S. I., Bottema, M., Breyer, R. R., Bybee, R. L., Christon,P. R., Delamere, A. W., Dorn, D. A., Downey, S., Driggers, P. A., Ebbets,D. C., Gallegos, J. S., Garner, H., Hetlinger, J. C., Lettieri, R. L., Ludtke,C. W., Michika, D., Nyquist, R., Rose, D. M., Stocker, R. B., Sullivan, J. F.,Van Houten, C. N., Woodruff, R. A., Baum, S. A., Hartig, G. F., Balzano,V., Biagetti, C., Blades, J. C., Bohlin, R. C., Clampin, M., Doxsey, R.,Ferguson, H. C., Goudfrooij, P., Hulbert, S. J., Kutina, R., McGrath, M.,
Lindler, D. J., Beck, T. L., Feggans, J. K., Plait, P. C., Sandoval, J. L., Hill,R. S., Collins, N. R., Cornett, R. H., Fowler, W. B., Hill, R. J., Landsman,W. B., Malumuth, E. M., Standley, C., Blouke, M., Grusczak, A., Reed, R.,Robinson, R. D., Valenti, J. A., & Wolfe, T. 1998, ApJL, 492, L83Kruczek, N., France, K., Hoadley, K., Fleming, B., & Nell, N. 2019, ApJ, 878,77Kruczek, N., France, K., Hoadley, K., Fleming, B., Nell, N., Kane, R., &Beatty, D. 2017, in Space Telescopes and Instrumentation 2017, Proc. SPIEKruczek, N., Nell, N., France, K., Hoadley, K., Fleming, B., Ulrich, S.,Miller, A. D., Egan, A., Witt, E., & Kane, R. 2018, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 10699,Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Se-ries, 106990KLacour, S., Ziskin, V., H´ebrard, G., Oliveira, C., Andr´e, M. K., Ferlet, R., &Vidal-Madjar, A. 2005, ApJ, 627, 251Landsman, W. & Bowers, C. 1997, in The 1997 HST Calibration Workshopwith a New Generation of Instruments, ed. S. Casertano, R. Jedrzejewski,T. Keyes, & M. Stevens, 132Lehner, N., Jenkins, E. B., Gry, C., Moos, H. W., Chayer, P., & Lacour, S.2003, ApJ, 595, 858Ma, X.-Y., Zhu, Y.-Y., Yan, Q.-B., You, J.-Y., & Su, G. 2020, MNRASMandy, M. E. & Martin, P. G. 1993, ApJ, 86, 199Markwardt, C. B. 2009, in Astronomical Society of the Pacific ConferenceSeries, Vol. 411, Astronomical Data Analysis Software and Systems XVIII,ed. D. A. Bohlender, D. Durand, & P. Dowler, 251Martin, D. C., Fanson, J., Schiminovich, D., Morrissey, P., Friedman, P. G.,Barlow, T. A., Conrow, T., Grange, R., Jelinsky, P. N., Milliard, B., Sieg-mund, O. H. W., Bianchi, L., Byun, Y.-I., Donas, J., Forster, K., Heckman,T. M., Lee, Y.-W., Madore, B. F., Malina, R. F., Neff, S. G., Rich, R. M.,Small, T., Surber, F., Szalay, A. S., Welsh, B., & Wyder, T. K. 2005, ApJL,619, L1Martin, D. C., Seibert, M., Neill, J. D., Schiminovich, D., Forster, K., Rich,R. M., Welsh, B. Y., Madore, B. F., Wheatley, J. M., Morrissey, P., &Barlow, T. A. 2007, Nature, 448, 780Martin, E. R. & York, D. G. 1982, ApJ, 257, 135Martin, P. G. & Mandy, M. E. 1995, ApJL, 455, L89Mathis, J. S., Mezger, P. G., & Panagia, N. 1983, A&A, 128, 212McCandliss, S. R. 2003, PASP, 115, 651McCord, M. A. & Rooks, M. J. 2000, SPIE Handbook of Microlithography,Micromachining and Microfabrication, Vol. 1 (SPIE)McGrath, M. A., Busko, I., & Hodge, P. 1999, Calstis6: Extraction of 1-DSpectra in the STIS Calibration Pipeline, Tech. rep.Morton, D. C. 1975, ApJ, 197, 85Redfield, S. 2006, in Astronomical Society of the Pacific Conference Series,Vol. 352, New Horizons in Astronomy: Frank N. Bash Symposium, ed. S. J.Kannappan, S. Redfield, J. E. Kessler-Silacci, M. Landriau, & N. Drory, 79
HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 23
Sahai, R. & Chronopoulos, C. K. 2010, ApJL, 711, L53Sasian, J. 1997, Optical Engineering, 36, 183Savage, B. D., Bohlin, R. C., Drake, J. F., & Budich, W. 1977, ApJ, 216, 291Scowen, P. A., Tripp, T., Beasley, M., Ardila, D., Andersson, B.-G., Ma´ızApell´aniz, J., Barstow, M., Bianchi, L., Calzetti, D., Clampin, M., Evans,C. J., France, K., Garc´ıa Garc´ıa, M., Gomez de Castro, A., Harris, W., Har-tigan, P., Howk, J. C., Hutchings, J., Larruquert, J., Lillie, C. F., Matthews,G., McCandliss, S., Polidan, R., Perez, M. R., Rafelski, M., Roederer, I. U.,Sana, H., Sanders, W. T., Schiminovich, D., Thronson, H., Tumlinson, J.,Vallerga, J., & Wofford, A. 2017, PASP, 129, 076001Siegmund, O. H. W., Tremsin, A. S., & Vallerga, J. V. 2009, in Proc. SPIE, Vol.7435, UV, X-Ray, and Gamma-Ray Space Instrumentation for AstronomyXVI, 74350LSnow, T. P. & McCall, B. J. 2006, ARA&A, 44, 367Spitzer, L., Drake, J. F., Jenkins, E. B., Morton, D. C., Rogerson, J. B., &York, D. G. 1973, ApJL, 181, L116Stecher, T. P. & Williams, D. A. 1967, ApJL, 149, L29The LUVOIR Team. 2019, arXiv e-prints, arXiv:1912.06219Thomas, R. J. 2003, in Proc. SPIE, Vol. 4853, Innovative Telescopes and In-strumentation for Solar Astrophysics, ed. S. L. Keil & S. V. Avakyan, 411–418Tumlinson, J., Peeples, M. S., & Werk, J. K. 2017, ARA&A, 55, 389Valenti, J., Busko, I., Quijano, J. K., Lindler, D., & Bowers, C. W. 2003, inHST Calibration Workshop : Hubble after the Installation of the ACS andthe NICMOS Cooling System, ed. S. Arribas, A. Koekemoer, & B. Whit-more, 209Vallerga, J., Raffanti, R., Tremsin, A., Siegmund, O., McPhate, J., & Varner,G. 2010, in Proc. SPIE, Vol. 7732, Space Telescopes and Instrumentation2010: Ultraviolet to Gamma Ray, 773203van Dishoeck, E. F. & Black, J. H. 1986, ApJS, 62, 109Weingartner, J. C. & Draine, B. T. 2001, ApJ, 548, 296Wolniewicz, L., Simbotin, I., & Dalgarno, A. 1998, ApJ, 115, 293Woodruff, R. A., Danchi, W. C., Heap, S. R., Hull, T., Kendrick, S. E., Purves,L. R., Rhee, M. S., Mentzell, E., Fleming, B., Valente, M., Burge, J., Lewis,B., Dodson, K., Mehle, G., & Tomic, M. 2019, Journal of AstronomicalTelescopes, Instruments, and Systems, 5, 024006
List of Figures − × reflectivity of Al+LiF)of the cross dispersing grating in CHESS over time, overplottedwith simple spline curves to show the resemblance of each trial.The colored points and lines represent times before and afterthe CHESS-1 and CHESS-2 missions that the cross dispersergrating efficiency was measured: pre-36.285 (blue) is before thecross disperser was installed and aligned into the instrumentfor the CHESS-1 flight, post-36.285 (red) is right after we re-covered the instrument after the CHESS-1 flight and measuredthe post-flight efficiency of all the optical components, and pre-36.297 (green) is before the grating was installed and alignedinto the instrument structure for the CHESS-2 flight. We focuson the reflectivity of the m = -1 order, which is the disper-sion order used in the CHESS instrument. Because Al+LiF canexhibit efficiency degradations when not stored in a dry environ-ment, we measure how the order reflectivity changes betweenCHESS-1 and CHESS-2 without re-coating the optic. No signif-icant degradation of the coating has been measured between thefirst two flights of CHESS. Adapted from Hoadley et al. (2016). 31 HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 25 Left:
Performance (for each grating: peak order efficiency, andfor the detector: detector quantum efficiency) of all optical com-ponents of CHESS-2. For the first flight of CHESS, the onlycomponent performance that changed significantly was the echelleefficiency, which was << Right:
The CHESS-2 effective area, including throughput loss frombaffling, compared to the effective area of CHESS-1. The to-tal effective area of CHESS-2 is about an order of magnitudelarger than that of CHESS-1, owing primarily to the large gainin echelle order efficiency across the bandpass. Adapted fromHoadley et al. (2016). . . . . . . . . . . . . . . . . . . . . . . . 326 Presented are the raw images of the CHESS-1 (left) and CHESS-2 (right; edge effects have been cropped out) echellograms frompre-flight calibrations (March 2014 and December 2015) usingan arc lamp flowing 65%/35% H/Ar gas. The brightest featurein both images is H I-Ly α ( λ α in one echelle order, while the CHESS-2echellogram disperses Ly α photons into two adjacent echelle or-ders. The other broad feature(s) visible in the CHESS-2 echel-logram are HI-Ly β (1025.72 ˚A), about 1/4 of the way from thetop of the image, and HI-Ly γ (97.25 ˚A), barely visible abovethe Ly β features. The more discrete features dotted through-out the spectrum are H emission from electron-impact fluores-cence. Adapted from Hoadley et al. (2014) and Hoadley et al.(2016). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 The complete first-order wavelength solution for the pre-launchCHESS-2 calibration spectra from λλ
900 - 1750 ˚A. The finalwavelength solution was determined using H fluorescence emis-sion features and a functional extrapolation of the wavelengthwith a 6 th -order polynomial fit. Over-plotted in magenta is themodel H fluorescence inside the arc lamp (T eff = 800 K, N(H ) ∼ cm − , E electron = 50 eV). The spectrum is scaled to thehighest total counts of the H features; otherwise, Ly α woulddominate the spectrum and the H features would be washedout. To show how neighboring order spectra overlap and cor-relate to form the final 1D spectrum, individual order spectrahave been plotted in different colors. Adapted from Hoadleyet al. (2016). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348 The line spread function (LSF) fits of H emission features inone order of the pre-launch calibration spectrum of CHESS-2(echelle order m = 286). The order spectrum is shown in black.Red and blue Gaussian line fits are shown for the narrow andbroad Gaussian fits for each line, respectively. The green lineis the sum of all Gaussian components to reproduce the spec-trum. A modeled H fluorescence spectrum is shown in magenta.Adapted from Hoadley et al. (2016). . . . . . . . . . . . . . . . 35 α Vir. The purple/blackregions represent areas with lower concentrations of photoncounts, and blue/green pixels represent pixels with higher con-centrations of photons collected. Marked with green arrows andlabeled are the most prominent features in the echellogram. Be-cause of the low S/N of the flight data, the echellogram has beenbinned to 512 × (cid:15) Per, recorded over t exp ∼
250 sec on NASA/CU mission 36.297 UG (CHESS-2). Along theright, we mark the rough wavelength coverage as a function ofthe detector y axis. In an individual echelle order, wavelength in-creases to the right. Stellar continuum through the far-UV actsas a back-light behind the interstellar material. Dark streaksin the continuum show stellar and interstellar atoms, ions, andmolecules absorbing photons at specific wavelengths; prominentinterstellar and stellar features are labeled with arrows pointingto the absorption lines and the absorption species along the left.Adapted from Hoadley et al. (2016). . . . . . . . . . . . . . . . 3711 The flux-calibrated CHESS-2 spectra of (cid:15)
Per from λ = 1020- 1550 ˚A. Representative error bars are shown in red. TheCHESS-2 spectrum was flux-calibrated against IUE spectra of (cid:15)
Per, and the continuum shape was compared with
Coperni-cus spectra at λ < absorption features from v = 0, J = 0 - 7, orange linesshow H I absorption, green lines show carbon, oxygen, and ni-trogen species, and blue lines mark heavier metals (iron, silicon,sulfur, argon, magnesium, nickel, and copper). . . . . . . . . . . 3812 (a) We present the normalized absorption features of differentphases of interstellar carbon observed in the CHESS (left pan-els), Copernicus (middle panels), and IUE (right panels) FUVspectra of (cid:15)
Per. We fit all lines with a Gaussian profile (bluedashed line) and determine the equivalent width (W λ ) from thearea under the Gaussian (green hashed area). We determine W λ for many metal lines found throughout the FUV in all data setsover a variety of interstellar phases - a comprehensive list ofresults is presented in Table 3. . . . . . . . . . . . . . . . . . . 3912 (b) Continued; different phases of nitrogen observed by CHESS,Copernicus, and IUE , their Gaussian fits (blue dashed line), andW λ (green shaded area). . . . . . . . . . . . . . . . . . . . . . . 4012 (c) Continued; different phases of silicon observed by CHESS,Copernicus, and IUE , their Gaussian fits (blue dashed line), andW λ (green shaded area). . . . . . . . . . . . . . . . . . . . . . . 41 HESS: an Innovative Concept for High-Resolution, Far-UV Spectroscopy 27
12 (d) Continued; phosphorus, hydrogen, and sulfur observed byCHESS, Copernicus, and
IUE , their Gaussian fits (blue dashedline), and W λ (green shaded area). . . . . . . . . . . . . . . . . 4213 Synthetic H profile fits for the H (4 −
0) band ( left ) and H (1 −
0) and H (0 −
0) bands ( right ), shown in red, are overlaidon top of the (cid:15)
Per spectrum. Molecular rotational levels arelabeled with purple dashes. The best-fit Doppler velocities forall three spectral band fits is b = 3.6 km s − . . . . . . . . . . . 4314 We present the (cid:15) Per H rotation diagram with two differentmodel fits. Left:
The H rotation diagram is fit assuming thesightline has two temperature populations of H : a cool, ki-netic temperature, described by T (pink), and a warmer, “ex-citation” temperature, described by T exc (blue). Right:
The H rotation diagram is fit with an H equilibrium model (purple as-terisks), which includes affects from UV-photon pumping, col-lisions with other particles, and formation/destruction rates ofH in a diffuse medium. The kinetic temperature derived fromthese models is shown in orange. . . . . . . . . . . . . . . . . . 44 Fig. 1
The Zemax ray trace of CHESS, including the secondary aspect camera system.The mechanical collimator reduces stray light in the line of sight and feeds starlight to theechelle. The echelle disperses UV light into high-dispersion orders, which are focused bythe cross disperser onto the detector plane. The different colored lines represent a series ofwavelengths across the 1000 − Fig. 2
Schematic view of the Colorado High-resolution Echelle Stellar Spectrograph(CHESS). Dimensions defined are in centimeters. The optical path (dashed purple line)follows right to left, with the target light entering the instrument when the shutter door isopen during flight. Adapted from Adapted from Hoadley et al. (2014) and Hoadley et al.(2016).0 FIGURES
Fig. 3
A comparison of echelle gratings tested for use in the CHESS instrument. We includethe best-performing echelle gratings from the lithography etching R&D project undertakenby LightSmyth, Inc. (flown on CHESS-1, 36.285 UG), the e-beam samples fabricated byJPL, and two mechanically-ruled replica gratings from Bach Research, Inc. and RichardsonGratings, respectively. Both mechanically-ruled gratings out-performed the R&D echellesand met the CHESS minimum order efficiency threshold. Adapted from Hoadley et al.(2016).IGURES 31
Fig. 4
The measured reflectivity (order efficiency × reflectivity of Al+LiF) of the crossdispersing grating in CHESS over time, overplotted with simple spline curves to show theresemblance of each trial. The colored points and lines represent times before and after theCHESS-1 and CHESS-2 missions that the cross disperser grating efficiency was measured:pre-36.285 (blue) is before the cross disperser was installed and aligned into the instrumentfor the CHESS-1 flight, post-36.285 (red) is right after we recovered the instrument afterthe CHESS-1 flight and measured the post-flight efficiency of all the optical components,and pre-36.297 (green) is before the grating was installed and aligned into the instrumentstructure for the CHESS-2 flight. We focus on the reflectivity of the m = -1 order, which isthe dispersion order used in the CHESS instrument. Because Al+LiF can exhibit efficiencydegradations when not stored in a dry environment, we measure how the order reflectivitychanges between CHESS-1 and CHESS-2 without re-coating the optic. No significant degra-dation of the coating has been measured between the first two flights of CHESS. Adaptedfrom Hoadley et al. (2016).2 FIGURES Fig. 5
Left:
Performance (for each grating: peak order efficiency, and for the detector:detector quantum efficiency) of all optical components of CHESS-2. For the first flight ofCHESS, the only component performance that changed significantly was the echelle effi-ciency, which was << Right:
The CHESS-2 effective area,including throughput loss from baffling, compared to the effective area of CHESS-1. Thetotal effective area of CHESS-2 is about an order of magnitude larger than that of CHESS-1,owing primarily to the large gain in echelle order efficiency across the bandpass. Adaptedfrom Hoadley et al. (2016).IGURES 33
Fig. 6
Presented are the raw images of the CHESS-1 (left) and CHESS-2 (right; edge effectshave been cropped out) echellograms from pre-flight calibrations (March 2014 and December2015) using an arc lamp flowing 65%/35% H/Ar gas. The brightest feature in both imagesis H I-Ly α ( λ α in one echelle order,while the CHESS-2 echellogram disperses Ly α photons into two adjacent echelle orders. Theother broad feature(s) visible in the CHESS-2 echellogram are HI-Ly β (1025.72 ˚A), about1/4 of the way from the top of the image, and HI-Ly γ (97.25 ˚A), barely visible above theLy β features. The more discrete features dotted throughout the spectrum are H emissionfrom electron-impact fluorescence. Adapted from Hoadley et al. (2014) and Hoadley et al.(2016).4 FIGURES Fig. 7
The complete first-order wavelength solution for the pre-launch CHESS-2 calibrationspectra from λλ
900 - 1750 ˚A. The final wavelength solution was determined using H fluorescence emission features and a functional extrapolation of the wavelength with a 6 th -order polynomial fit. Over-plotted in magenta is the model H fluorescence inside the arclamp (T eff = 800 K, N(H ) ∼ cm − , E electron = 50 eV). The spectrum is scaled tothe highest total counts of the H features; otherwise, Ly α would dominate the spectrumand the H features would be washed out. To show how neighboring order spectra overlapand correlate to form the final 1D spectrum, individual order spectra have been plotted indifferent colors. Adapted from Hoadley et al. (2016).IGURES 35 Fig. 8
The line spread function (LSF) fits of H emission features in one order of the pre-launch calibration spectrum of CHESS-2 (echelle order m = 286). The order spectrum isshown in black. Red and blue Gaussian line fits are shown for the narrow and broad Gaussianfits for each line, respectively. The green line is the sum of all Gaussian components toreproduce the spectrum. A modeled H fluorescence spectrum is shown in magenta. Adaptedfrom Hoadley et al. (2016).6 FIGURES Fig. 9
The False-color representation of the flight echellogram from CHESS-1, taken on 24May 2014, of α Vir. The purple/black regions represent areas with lower concentrations ofphoton counts, and blue/green pixels represent pixels with higher concentrations of photonscollected. Marked with green arrows and labeled are the most prominent features in theechellogram. Because of the low S/N of the flight data, the echellogram has been binned to512 × Fig. 10
The raw, false-color echellogram of (cid:15)
Per, recorded over t exp ∼
250 sec onNASA/CU mission 36.297 UG (CHESS-2). Along the right, we mark the rough wavelengthcoverage as a function of the detector y axis. In an individual echelle order, wavelengthincreases to the right. Stellar continuum through the far-UV acts as a back-light behindthe interstellar material. Dark streaks in the continuum show stellar and interstellar atoms,ions, and molecules absorbing photons at specific wavelengths; prominent interstellar andstellar features are labeled with arrows pointing to the absorption lines and the absorptionspecies along the left. Adapted from Hoadley et al. (2016).8 FIGURES F l ux ( − e r g s c m − s − Å − ) Wavelength (Å)
Fig. 11
The flux-calibrated CHESS-2 spectra of (cid:15)
Per from λ = 1020 - 1550 ˚A. Repre-sentative error bars are shown in red. The CHESS-2 spectrum was flux-calibrated against IUE spectra of (cid:15)
Per, and the continuum shape was compared with
Copernicus spectra at λ < absorption features from v = 0, J = 0 - 7, orange lines showH I absorption, green lines show carbon, oxygen, and nitrogen species, and blue lines markheavier metals (iron, silicon, sulfur, argon, magnesium, nickel, and copper).IGURES 39 Fig. 12 (a) We present the normalized absorption features of different phases of interstellarcarbon observed in the CHESS (left panels), Copernicus (middle panels), and
IUE (rightpanels) FUV spectra of (cid:15)
Per. We fit all lines with a Gaussian profile (blue dashed line)and determine the equivalent width (W λ ) from the area under the Gaussian (green hashedarea). We determine W λ for many metal lines found throughout the FUV in all data setsover a variety of interstellar phases - a comprehensive list of results is presented in Table 3.0 FIGURES Fig. 12 (b) Continued; different phases of nitrogen observed by CHESS, Copernicus, and
IUE , their Gaussian fits (blue dashed line), and W λ (green shaded area).IGURES 41 Fig. 12 (c) Continued; different phases of silicon observed by CHESS, Copernicus, and
IUE , their Gaussian fits (blue dashed line), and W λ (green shaded area).2 FIGURES Fig. 12 (d) Continued; phosphorus, hydrogen, and sulfur observed by CHESS, Copernicus,and
IUE , their Gaussian fits (blue dashed line), and W λ (green shaded area).IGURES 43 ε Per: 1045 − 1060 Å N o r m a ili ze d F l ux J" = H ModelCHESSH ModelCHESSlog N(H ) = 19.72 cm −2 ε Per: 1085 − 1120 Å N o r m a ili ze d F l ux J" = H ModelCHESSH ModelCHESSlog N(H ) = 19.72 cm −2 Fig. 13
Synthetic H profile fits for the H (4 −
0) band ( left ) and H (1 −
0) and H (0 − right ), shown in red, are overlaid on top of the (cid:15) Per spectrum. Molecular rotationallevels are labeled with purple dashes. The best-fit Doppler velocities for all three spectralband fits is b = 3.6 km s − .4 FIGURES [0,J]) (K)1214161820 l og ( N ( H [ , J ]) / g J ) ε PerT (H ) = 95 KT exc (H ) = 500 K b = 3.6 km s −1 [0,J]) (K)1214161820 l og ( N ( H [ , J ]) / g J ) ε Per (CHESS−2)log N(H ) = 19.83, T (H ) = 95 KH ISM Models b = 3.6 km s −1 Fig. 14
We present the (cid:15)
Per H rotation diagram with two different model fits. Left:
TheH rotation diagram is fit assuming the sightline has two temperature populations of H : acool, kinetic temperature, described by T (pink), and a warmer, “excitation” temperature,described by T exc (blue). Right:
The H rotation diagram is fit with an H equilibriummodel (purple asterisks), which includes affects from UV-photon pumping, collisions withother particles, and formation/destruction rates of H2