InFOCμS Hard X-ray Imaging Telescope

InFOCμS is a new generation balloon-borne hard X-ray telescope with focusing optics and spectroscopy. We had a successful 22.5-hour flight from Fort Sumner, NM on September 16,17, 2004. In this paper, we present the performance of the hard X-ray telescope, which consists of a depth-graded platinum/carbon multilayer mirror and a CdZnTe detector. The telescope has an effective area of 49 cm² at 30 keV, an angular resolution of 2.4 arcmin (HPD), and a field of view of 11 arcmin (FWHM) depending on energies. The CdZnTe detector is configured with a 12 × 12 segmented array of detector pixels. The pixels are 2 mm square, and are placed on 2.1 mm centers. An averaged energy resolution is 4.4 keV at 60 keV and its standard deviation is 0.36 keV over 128 pixels. The detector is surrounded by a 3-cm thick CsI anti coincidence shield to reduce background from particles and photons not incident along the mirror focal direction. The inflight background is 2.9 × 10⁻⁴ cts cm⁻² sec⁻¹ keV⁻¹ in the 20–50 keV band.     

A Si/CdTe Compton Camera for gamma-ray lens experiment

Cadmium telluride (CdTe) and cadmium zinc telluride (CdZnTe) have been regarded as promising semiconductor materials for hard X-ray and γ-ray detection. However, a considerable amount of charge loss in these detectors results in a reduced energy resolution. We have achieved a significant improvement in the spectral properties by forming the Schottky junction on the Te side of the CdTe wafer. With the further reduction of leakage current by an adoption of guard ring structure, we have demonstrated a CdTe pixel detector with high energy resolution and full charge collection capabilty. The detector has a pixel size of a few mm and a thickness of 0.5 $-$ 1 mm. We apply this high resolution detector to a new silicon and CdTe Compton Camera which features high angular resolution. We also describe a concept of the stack detector which consists of many thin CdTe layers and provides sufficient efficiency for hard X-rays and gamma-rays up to several hundred keV maintaining good energy resolution. A narrow-FOV Compton telescope can be realized by installing a Si/CdTe Compton Camera inside the deep well of an active shield. This configuration is very suitable as focal plane detector for future focusing gamma-ray missions.     

Compact Two-Dimensional Array of Stressed GE:GA Detectors

We have developed a 4 × 8 array of stressed Ge:Ga detectors. This array detector has a high density format ofentrance pupils so that we can minimize the size of the cameraoptics. The cutoff wavelength of the detector is about 170 m, and the detector's NEP is better than 10¹⁶ WHz^-1/2. We are going apply this array detector toballoon-borne astronomical observations. Furthermore, we aredeveloping this detector into a 5 × 15 array detector that will be placed onboard the IRIS satellite to be launched in 2003.     

Feasibility and design of a solid state gamma-ray detector

For conventional radiation detectors fabricated from compound semi-conductors, the wide disparity between the transport properties of the electron and holes, means that detector performances are limited by the carrier with the poorest mobility-lifetime product (μτ). Finite drift lengths introduce an energy dependent depth term into the charge collection process, which effectively limit maximum detection volume to tens of mm³ – entirely unsuitable for the detection of gamma-rays. The recent introduction of the coplanar-grid charge-sensing techniques has overcome this problem by essentially discarding the carrier with the poorest transport properties, thus permitting high spectral resolution and high detection efficiency. For example, energy resolutions of 2% full-width half-maximum at 662 keV have been demonstrated with coplanar-grid CdZnTe detectors of volumes up to 2 cm³. Further improvements in detector performance and yield are being pursued through refinements in electrode design and material quality. Because coplanar-grid CdZnTe detectors can operate at room temperature, they are ideally suited for applications requiring portability, small size, or low power consumption such as planetary space missions. Other potential applications include well logging, medical diagnostics, and gamma-ray astronomy. We discuss the feasibility and design of a solid state gamma-ray detector based on CdZnTe and compare its performance to a large volume Ge detector. As will be shown, a significant improvement can be made if T1Br is used as the detection medium.     

A 256×256 pixel InSb FPA for SIRTF and other astronomical applications

As part of the NASA Space Infrared telescope Facility (SIRTF), a low noise multiplexer has been developed. The hybridization of this multiplexer to a high indium antimonide (InSb) photodiode array has resulted in a MWIR detector of outstanding performance. The multiplexer is made of a 256×256 array of source follower amplifiers on a 30 m square pitch. Random access binary decoders are used to access each pixel of the array, allowing any read-out scheme to be implemented. Dark current has been measured at temperatures ranging from 4K to 77K. Generation recombination currents dominate above 45K. With 100mV of reverse bias applied, less than 3×10^-17 A is typical below 50K with 8×10^-19 A (5 e^-/s) at 4K. Under the same conditions 0.25 pA was measured at 77K. Read noise has been measured as low as 186 e^- using non-correlated techniques. Detector QE is 50 to 80 % through the entire 1 to 5 m band.     

Prospects of hard X-ray polarimetry with Astrosat-CZTI

Astrosat is the first Indian satellite mission dedicated for astronomical studies. It is planned for launch during 2014 and will have five instruments for multi-wavelength observations from optical to hard X-rays. Cadmium Zing Telluride Imager (CZTI) is one of the five instruments aiming for simultaneous X-ray spectroscopy and imaging in the energy range of 10 keV to 100 keV (along with all sky photometric capability unto 250 keV). It is based on pixilated CZT detector array with total geometric area of 1024 cm². It will have two-dimensional coded mask for medium resolution X-ray imaging. The CZT detector plane will be realized using CZT detector modules having integrated readout electronics. Each CZT detector module consists of 4 cm × 4 cm CZT with thickness of 5 mm which is further pixilated into 16 × 16 array of pixels. Thus each pixel has size of 2.5 mm × 2.5 mm and thickness of 5 mm. Such pixilated detector plane can in principle be used for hard X-ray polarization measurements based on the principle of Compton scattering by measuring azimuthal distribution of simultaneous events in two adjacent pixels. We have carried out detailed Geant4 simulations for estimating polarimetric capabilities of CZTI detector plane. The results indicate that events in the energy range of 100 keV to 250 keV, where the 5 mm thick CZT detector has significant detection efficiency, can be used for polarimetric studies. Our simulation results indicate the minimum detectable polarization (MDP) at the level of ～ 10% can be achieved for bright Crab like X-ray sources with exposure time of ～500 ks. We also carried out preliminary experiments to verify the results from our simulations. Here we present detailed method and results of our simulations as well as preliminary results from the experimental verification of polarimetric capabilities of CZT detector modules used in Astrosat CZTI.     

The infrared space observatory camera

At the focal plane of the ISO 60 cm telescope, ISOCAM will take images of the sky in the wavelength range 2.5 to 17 m. It features two independent channels, containing each a 32×32 array detector. The long wavelength (4–17 m) detector, developed by LIR-LETI specifically for ISOCAM, is a Si:Ga array bonded with indium bumps to a DV readout circuit. The short wavelength (2.5–5.5 m) detector, developed by SAT, is an InSb array with a CID readout.Only one channel operates at a time. The channel selection is obtained with two field mirrors supported by a wheel. The field mirror in operation is in the focal plane of the telescope, covering the 3 arcmin field of view, and reflecting the light into one or into the other channel.In each channel, 10 or 12 discrete band pass filters and CVFs with resolution 45 are mounted on a wheel. A second wheel supports a set of lenses, which reimage the focal plane of the telescope on the array, with a pixel field of view of 1.5, 3, 6 or 12 arcsec.Two integrating spheres, illuminated through small holes by black body sources, are mounted on the selection wheel, for flat fielding and calibration purposes.The flight model of ISOCAM, fully tested and calibrated, has been, delivered to ESA and is ready to be integrated in the satellite, whose expected launch date is September 1995.     

Legri Operations. Detectors and Detector Stability

Two years after launch (04.21.97), LEGRI is operating on Minisat-01 in a LEO orbit. The LEGRI detector plane is formed by two type of gamma-ray solid state detectors: HgI₂ and CdZnTe. Detectors are embedded in a box containing the FEE and DFE electronics. This box provides an effective detector passive shielding. Detector plane is multiplexed by a Coded Aperture System located at 54 cm and a Ta Collimator with a FCFOV of 22° and 2° angular resolution. The aim of this paper is to summarize the detector behaviour in three different time scales: before launch, during the in-orbit check-out period (IOC), and after two years of routine operation in space. Main results can be summarized as follows:A large fraction of the HgI₂ detectors presented during LEGRI IOC very high count ratios from their first switch-on (May 1997). Therefore, they induced saturation in the on-board mass memory. After some unsuccessful attempts to reduce the count ratios by setting up different thresholds during LEGRI IOC, all of them were switched off except nine detectors in column 4, with a higher degree of stability.Oppositely, the 17 CdZnTe detectors present a remarkable stability in both their count-ratios and spectral shapes. Details about CdZnTe ground energy calibration, in-flight calibration (using the Crab) and detector stability are discussed hereafter. Detector efficiency function has been computed with the fixed flight threshold used within the calibrated energy range (20-80 KeV). It presents a maximum at 60 KeV, and decreasing efficiencies in the lower and upper energy range ends. Both, non-linear threshold cutting and the drop in the detector efficiency explain the CdZnTe computed operational efficiency response.     

The advanced Moon micro-imager experiment (AMIE) on SMART-1: Scientific goals and expected results

The advanced Moon micro-imager experiment (AMIE) is the imaging system on board ESA mission to the Moon SMART-1; it makes use of a miniaturised detector and micro-processor electronics developed by SPACE X in the frame of the ESA technical programme. The AMIE micro-imager will provide high resolution CCD images of selected lunar areas and it will perform colour imaging through three filters at 750, 915 and 960 nm with a maximum resolution of 46 m/pixel at the perilune of 500 km. Specific scientific objectives will include (1) imaging of high latitude regions in the southern hemisphere, in particular the South Pole Aitken basin (SPA) and the permanently shadowed regions close to the South Pole, (2) determination of the photometric properties of the lunar surface from observations at different phase angles (physical properties of the regolith), (3) multi-band imaging for constraining the chemical and mineral composition of the surface, (4) detection and characterisation of lunar non-mare volcanic units, (5) study of lithological variations from impact craters and implications for crustal heterogeneity. The AMIE micro-imager will also support a Laser-link experiment to Earth, an On Board Autonomous Navigation investigation and a Lunar libration experiment coordinated with radio science measurements.     

Spin-dependent limits from the DRIFT-IId directional dark matter detector

Data are presented from the DRIFT-IId detector operated in the Boulby Underground Science Facility in England. A 0.8 m³ fiducial volume, containing partial pressures of 30 Torr CS₂ and 10 Torr CF₄, was exposed for a duration of 47.4 live-time days with sufficient passive shielding to provide a neutron free environment within the detector. The nuclear recoil events seen are consistent with a remaining low-level background from the decay of radon daughters attached to the central cathode of the detector. However, charge from such events must drift across the entire width of the detector, and thus display large diffusion upon reaching the readout planes of the device. Exploiting this feature, it is shown to be possible to reject energy depositions from these Radon Progeny Recoil events while still retaining sensitivity to fiducial-volume nuclear recoil events. The response of the detector is then interpreted, using the F nuclei content of the gas, in terms of sensitivity to proton spin-dependent WIMP-nucleon interactions, displaying a minimum in sensitivity cross section at 1.8 pb for a WIMP mass of 100 GeV/c². This sensitivity was achieved without compromising the direction sensitivity of DRIFT.     

Analyses of the Light Variations of OJ 287, 0716+714 and BL Lacertae

OJ 287 is a BL Lac object which exhibits intense activities of low peak-frequencies. Its energy spectrum in low frequency band is quite similar with those of two other TeV BL Lac objects (i.e., 0716+714 and BL Lacertae). However, the Cerenkov telescope did not detect its TeV rays. By using the observational data of these three heavenly bodies and comparing the discrepan- cies of their minimal periods of light variations and delays at 22 GHz, 37 GHz and B-waveband, we have further investigated the possible reason why the TeV gamma-rays of OJ 287 have not been observed. The results of analyses are as fol- lows. (1) For the minimal periods of light variations, the periods of OJ 287 at 37 GHZ and B-waveband are short. At 22 GHz the results of OJ 287 and 0716+714 are comparable, but the period of OJ 287 is much shorter in comparison with that of BL Lacertae, and this shows that its activity is more intense. However, the TeV gamma-rays of OJ 287 have not been detected, which implies that the radiation of OJ 287 in TeV waveband may have no connection with the minimal periods of light variations in these three low-energy wavebands. (2) In respect of delays, the delay of OJ 287 in the B waveband with respect to 37 GHz is longer than that of 0716+714, but shorter than that of BL Lacertae. Its delay at 37 GHz in respect to 22 GHz is shorter than that of 0716+714. Meanwhile, the delay of BL Lacertae at 37 GHz in respect to 22 GHz is negative, which implies that 22 GHz precedes 37 GHz. Via the comparison and analysis of delays, no obvious differences between OJ 287 and 0716+714 as well as BL Lacertae have been found. On the side of energy spectra, it is quite possible that due to the steep energy spectrum of OJ 287 in TeV waveband, the Cerenkov telescope has not detected the gamma radiation of OJ 287. However, nowadays it is still not clear whether the steep energy spectrum in TeV energy range has some influence on the light variations in low energy realm.     

The Lateral Trigger Probability function for the Ultra-High Energy Cosmic Ray showers detected by the Pierre Auger Observatory

In this paper we introduce the concept of Lateral Trigger Probability (LTP) function, i.e., the probability for an Extensive Air Shower (EAS) to trigger an individual detector of a ground based array as a function of distance to the shower axis, taking into account energy, mass and direction of the primary cosmic ray. We apply this concept to the surface array of the Pierre Auger Observatory consisting of a 1.5 km spaced grid of about 1600 water Cherenkov stations. Using Monte Carlo simulations of ultra-high energy showers the LTP functions are derived for energies in the range between 10¹⁷ and 10¹⁹ eV and zenith angles up to 65°. A parametrization combining a step function with an exponential is found to reproduce them very well in the considered range of energies and zenith angles. The LTP functions can also be obtained from data using events simultaneously observed by the fluorescence and the surface detector of the Pierre Auger Observatory (hybrid events). We validate the Monte Carlo results showing how LTP functions from data are in good agreement with simulations.     

The High-energy Burst Spectrometer for SMESE Mission

The SMall Explorer for Solar Eruptions (SMESE) is a small satellite being developed jointly by China and France. It is planed to launch around the next solar maximum year (∼ 2011) for observing simultaneously the two most violent types of eruptive events on the sun (the coronal mass ejection (CME) and the solar flare) and investigating their relationship. As one of the 3 main payloads of the small satellite, the high energy burst spectrometer (HEBS) adopts the upto- date high-resolution LaBr3 scintillation detector to observe the high-energy solar radiation in the range 10 keV—600 MeV. Its energy resolution is better than 3.0% at 662 keV, 2-fold higher than that of current scintillation detectors, promising a breakthrough in the studies of energy release in solar flares and CMEs, particle acceleration and the relationship between solar flares and CMEs.     

Polarisation measurements with a CdTe pixel array detector for Laue hard X-ray focusing telescopes

Polarimetry is an area of high energy astrophysics which is still relatively unexplored, even though it is recognized that this type of measurement could drastically increase our knowledge of the physics and geometry of high energy sources. For this reason, in the context of the design of a Gamma-Ray Imager based on new hard-X and soft gamma ray focusing optics for the next ESA Cosmic Vision call for proposals (Cosmic Vision 2015-2025), it is important that this capability should be implemented in the principal on-board instrumentation. For the particular case of wide band-pass Laue optics we propose a focal plane based on a thick pixelated CdTe detector operating with high efficiency between 60–600keV. The high segmentation of this type of detector (1–2mm pixel size) and the good energy resolution (a few keV FWHM at 500keV) will allow high sensitivity polarisation measurements (a few % for a 10mCrab source in 10⁶s) to be performed. We have evaluated the modulation Q factors and minimum detectable polarisation through the use of Monte Carlo simulations (based on the GEANT 4 toolkit) for on and off-axis sources with power law emission spectra using the point spread function of a Laue lens in a feasible configuration.     

A Study of the [OII]λ3727/Hα Flux Ratio of Emission Line Galaxies

With the sample of 10⁵ emission line galaxies selected from the Sloan Digital Sky Survey Data Release 4 (SDSS DR4), we have investigated the relations of the [O_II]λ3727/H_α flux ratio with the dust extinction, the ionization state of interstellar gas and the metal abundance of galaxies. It is found that the dust extinction correction has a significant effect on the [O_II]λ3727/H_α flux ratio. Before and after the dust distinction correction is made, the mean [O_II]λ3727/H_α flux ratios are 0.48 and 0.89, respectively. After the dust extinction is corrected, the dispersion of the distribution of F([O_II]λ3727) as a function of F(H_α) is obviously reduced. The [O_II]λ3727/H_α flux ratio of metal-poor galaxies decreases with the increasing ionization degree of interstellar gas, but this relation does not exist in metal-rich galaxies. Besides, it is found that the [O_II]λ3727/H_α flux ratio is correlated with the metal abundance. When 12 + lg(O/H) > 8.5, the [O_II]λ3727/H_α flux ratio decreases with the increasing metal abundance; for the galaxies of 12 + lg(O/H) > 8.5, the spectral flux ratio correlates positively with the metal abundance. Finally, by using the parameters of gas ionization degree and metal abundances of galaxies, the formulae for calculating the [O_II]λ3727/H_α flux ratios of different types of galaxies are given. With the [O_II]λ3727/H_α flux ratio given by these formulae, the star formation rate can be derived by using the [O_II]λ3727-line flux, for the galaxies of the redshift z > 0.4, such as the large number of galaxies to be observed by the LAMOST telescope.     

Latitudinal variations of HCN, HC3N, and C2N2 in Titan's stratosphere derived from Cassini CIRS data

Mid- and far-infrared spectra from the Composite InfraRed Spectrometer (CIRS) have been used to determine volume mixing ratios of nitriles in Titan's atmosphere. HCN, HC₃N, C₂H₂, and temperature were derived from 2.5 cm⁻¹ spectral resolution mid-IR mapping sequences taken during three flybys, which provide almost complete global coverage of Titan for latitudes south of 60° N. Three 0.5 cm⁻¹ spectral resolution far-IR observations were used to retrieve C₂N₂ and act as a check on the mid-IR results for HCN. Contribution functions peak at around 0.5-5 mbar for temperature and 0.1-10 mbar for the chemical species, well into the stratosphere. The retrieved mixing ratios of HCN, HC₃N, and C₂N₂ show a marked increase in abundance towards the north, whereas C₂H₂ remains relatively constant. Variations with longitude were much smaller and are consistent with high zonal wind speeds. For 90°-20° S the retrieved HCN abundance is fairly constant with a volume mixing ratio of around 1 × 10⁻⁷ at 3 mbar. More northerly latitudes indicate a steady increase, reaching around 4 × 10⁻⁷ at 60° N, where the data coverage stops. This variation is consistent with previous measurements and suggests subsidence over the northern (winter) pole at approximately 2 × 10⁻⁴ m s⁻¹. HC₃N displays a very sharp increase towards the north pole, where it has a mixing ratio of around 4 × 10⁻⁸ at 60° N at the 0.1-mbar level. The difference in gradient for the HCN and HC₃N latitude variations can be explained by HC₃N's much shorter photochemical lifetime, which prevents it from mixing with air at lower latitude. It is also consistent with a polar vortex which inhibits mixing of volatile rich air inside the vortex with that at lower latitudes. Only one observation was far enough north to detect significant amounts of C₂N₂, giving a value of around 9 × 10⁻¹⁰ at 50° N at the 3-mbar level.     

The Cryogenic AntiCoincidence detector for ATHENA X-IFU: a scientific assessment of the observational capabilities in the hard X-ray band

ATHENA is a large X-ray observatory, planned to be launched by ESA in 2028 towards an L2 orbit. One of the two instruments of the payload is the X-IFU: a cryogenic spectrometer based on a large array of TES microcalorimeters, able to perform integral field spectrography in the 0.2–12 keV band (2.5 eV FWHM at 6 keV). The X-IFU sensitivity is highly degraded by the particle background expected in the L2 orbit, which is induced by primary protons of both galactic and solar origin, and mostly by secondary electrons. To reduce the particle background level and enable the mission science goals, the instrument incorporates a Cryogenic AntiCoincidence detector (CryoAC). It is a 4 pixel TES based detector, placed < 1 mm below the main array. In this paper we report a scientific assessment of the CryoAC observational capabilities in the hard X-ray band (E > 10 keV). The aim of the study has been to understand if the present detector design can be improved in order to enlarge the X-IFU scientific capability on an energy band wider than the TES array. This is beyond the CryoAC baseline, being this instrument aimed to operate as anticoincidence particle detector and not conceived to perform X-ray observations.     

Minimum estimates of the amount and timing of gases released into the martian atmosphere from volcanic eruptions

Volcanism has been a major process during most of the geologic history of Mars. Based on data collected from terrestrial basaltic eruptions, we assume that the volatile content of martian lavas was typically ∼0.5 wt.% water, ∼0.7 wt.% carbon dioxide, ∼0.14 wt.% sulfur dioxide, and contained several other important volatile constituents. From the geologic record of volcanism on Mars we find that during the late Noachian and through the Amazonian volcanic degassing contributed ∼0.8 bar to the martian atmosphere. Because most of the outgassing consisted of greenhouse gases (i.e., CO₂ and SO₂) warmer surface temperatures resulting from volcanic eruptions may have been possible. Our estimates suggest that ∼1.1 × 10²¹ g (∼8 ± 1 m m⁻²) of juvenile water were released by volcanism; slightly more than half the amount contained in the north polar cap and atmosphere. Estimates for released CO₂ (1.6 × 10²¹ g) suggests that a large reservoir of carbon dioxide is adsorbed in the martian regolith or alternatively ∼300 cm cm⁻² of carbonates may have formed, although these materials would not occur readily in the presence of excess SO₂. Up to ∼120 cm cm⁻² (2.2 × 10²⁰ g) of acid rain (H₂SO₄) may have precipitated onto the martian surface as the result of SO₂ degassing. The hydrogen flux resulting from volcanic outgassing may help explain the martian atmospheric D/H ratio. The amount of outgassed nitrogen (∼1.3 mbar) may also be capable of explaining the martian atmospheric ¹⁵N/¹⁴N ratio. Minor gas constituents (HF, HCl, and H₂S) could have formed hydroxyl salts on the surface resulting in the physical weathering of geologic materials. The amount of hydrogen fluoride emitted (1.82 × 10¹⁸ g) could be capable of dissolving a global layer of quartz sand ∼5 mm thick, possibly explaining why this mineral has not been positively identified in spectral observations. The estimates of volcanic outgassing presented here will be useful in understanding how the martian atmosphere evolved over time.     

The deep water abundance on Jupiter: New constraints from thermochemical kinetics and diffusion modeling

We have developed a one-dimensional thermochemical kinetics and diffusion model for Jupiter’s atmosphere that accurately describes the transition from the thermochemical regime in the deep troposphere (where chemical equilibrium is established) to the quenched regime in the upper troposphere (where chemical equilibrium is disrupted). The model is used to calculate chemical abundances of tropospheric constituents and to identify important chemical pathways for CO-CH₄ interconversion in hydrogen-dominated atmospheres. In particular, the observed mole fraction and chemical behavior of CO is used to indirectly constrain the jovian water inventory. Our model can reproduce the observed tropospheric CO abundance provided that the water mole fraction lies in the range (0.25-6.0) × 10⁻³ in Jupiter’s deep troposphere, corresponding to an enrichment of 0.3-7.3 times the protosolar abundance (assumed to be H₂O/H₂ = 9.61 × 10⁻⁴). Our results suggest that Jupiter’s oxygen enrichment is roughly similar to that for carbon, nitrogen, and other heavy elements, and we conclude that formation scenarios that require very large (>8× solar) enrichments in water can be ruled out. We also evaluate and refine the simple time-constant arguments currently used to predict the quenched CO abundance on Jupiter, other giant planets, and brown dwarfs.     

The Unusual Absorption Line Spectrum of Quasar SDSS J2220+0109

We report the Balmer broad absorption lines (BALs) in the quasar SDSS J2220 + 0109 discovered from the SDSS data, and present a detailed analysis of the peculiar absorption line spectrum, including the He I* multiplet at λλ3189, 3889 arising from the metastable 2³s-state helium and the Balmer Hα and Hβ lines from the excited hydrogen H I of n = 2 level, which are rarely seen in quasar spectra, as well as many absorption lines arising from the excited Fe II* of the levels 7 955 cm⁻¹, 13 474 cm⁻¹ and 13 673 cm⁻¹ in the wavelength range 3100∼3300 Å. Ca II H, K absorption line doublets also clearly appear in the SDSS spectrum. All absorption lines show a similar blueshifted velocity structure of Δv ≈ − 1500 ∼ 0 km·s⁻¹ relative to the quasar's systematic redshift determined from the emission lines. Detailed analysis suggests that the Balmer absorption lines should arise from the partially ionized region with a column density of N_HI ≈ 10²¹ cm⁻² for an electron density of ne ∼ 10⁶ cm⁻³; and that the hydrogen n = 2 level may be populated via collisional excitation with Lyα pumping.