Publications (520)1048.72 Total impact
 [Show abstract] [Hide abstract]
ABSTRACT: We have constructed the first allsky map of the thermal SunyaevZeldovich (tSZ) effect by applying specifically tailored component separation algorithms to the 100 to 857 GHz frequency channel maps from the Planck survey. These maps show an obvious galaxy cluster tSZ signal that is well matched with blindly detected clusters in the Planck SZ catalogue. To characterize the signal in the tSZ map we have computed its angular power spectrum. At large angular scales ($\ell < 60$), the major foreground contaminant is the diffuse thermal dust emission. At small angular scales ($\ell > 500$) the clustered Cosmic Infrared Background (CIB) and residual point sources are the major contaminants. These foregrounds are carefully modelled and subtracted. We measure the tSZ power spectrum in angular scales, $0.17^{\circ} \lesssim \theta \lesssim 3.0^{\circ}$, that were previously unexplored. The measured tSZ power spectrum is consistent with that expected from the Planck catalogue of SZ sources, with additional clear evidence of signal from unresolved clusters and, potentially, diffuse warm baryons. We use the tSZ power spectrum to obtain the following cosmological constraints: $\sigma_8(\Omega_{\mathrm{m}}/0.28)^{3.2/8.1}=0.784 \pm 0.016 (68% C.L.). Marginalized bandpowers of the Planck tSZ power spectrum and the bestfit model are given. The nonGaussianity of the Compton parameter map is further characterized by computing its 1D probability distribution function and its bispectrum. These are used to place additional independent constraints on $\sigma_{8}$.03/2013; 
Article: Planck 2013 results X. Energetic particle effects: characterization, removal, and simulation
[Show abstract] [Hide abstract]
ABSTRACT: This paper presents the detection, interpretation and removal of the signal resulting from interactions of high energy particles with the Planck High Frequency Instrument (HFI). These interactions fall into two categories, heating the 0.1K bolometer plate and glitches in each detector time stream. Glitch shapes are not simple single pole exponential decays and fall into three families. The glitch shape for each family has been characterized empirically in flight data and removed from the detector time streams. The spectrum of the count rate per unit energy is computed for each family and a correspondence to where on the detector the particle hit is made. Most of the detected glitches are from galactic protons incident on the Si die frame supporting the micromachinced bolometric detectors. At HFI, the particle flux is � 5 cm2 s1 and is dominated by protons incident on the spacecraft with an energy >39MeV, leading to a rate of typically one event per second and per detector. Di�erent categories of glitches have di�erent signature in timestreams. Two of the glitch types have a low amplitude component that decays over nearly 1 second. This component produces an excess noise if not properly removed from the time ordered data. We have used a glitch detection and subtraction method based on the joint fit of population templates. The application of this novel glitch removal method removes excess noise from glitches. Using realistic simulations, we find this method does not introduce signal bias.03/2013;  03/2013;
 [Show abstract] [Hide abstract]
ABSTRACT: Based on CMB maps from the 2013 Planck Mission data release, this paper presents the detection of the ISW effect, i.e., the correlation between the CMB and largescale evolving gravitational potentials. The significance of detection ranges from 2 to 4 sigma, depending on which method is used. We investigate three separate approaches, which cover essentially all previous studies, as well as breaking new ground. (i) Correlation of the CMB with the Planck reconstructed gravitational lensing potential (for the first time). This detection is made using the lensinginduced bispectrum; the correlation between lensing and the ISW effect has a significance close to 2.5 sigma. (ii) Crosscorrelation with tracers of LSS, yielding around 3 sigma significance, based on a combination of radio (NVSS) and optical (SDSS) data. (iii) Aperture photometry on stacked CMB fields at the locations of known largescale structures, which yields a 4 sigma signal when using a previously explored catalogue, but shows strong discrepancies in amplitude and scale compared to expectations. Recent catalogues give more moderate results, ranging from negligible to 2.5 sigma at most, but with a more consistent scale and amplitude, the latter being still slightly above what is expected from numerical simulations within LCMD. Where they can be compared, these measurements are compatible with previous work using data from WMAP, which had already mapped these scales to the limits of cosmic variance. Planck's broader frequency coverage confirms that the signal is achromatic, bolstering the case for ISW detection. As a final step we use tracers of largescale structure to filter the CMB data, presenting maps of the ISW temperature perturbation. These results provide complementary and independent evidence for the existence of a dark energy component that governs the current accelerated expansion of the Universe.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: The multifrequency capability of the Planck satellite provides information both on the integrated history of star formation (via the cosmic infrared background, or CIB) and on the distribution of dark matter (via the lensing effect on the cosmic microwave background, or CMB). The conjunction of these two unique probes allows us to measure directly the connection between dark and luminous matter in the high redshift (1 < z <3) Universe. We use a threepoint statistic optimized to detect the correlation between these two tracers. Following a thorough discussion of possible contaminants and a suite of consistency tests, using lens reconstructions at 100, 143 and 217 GHz and CIB measurements at 100857 GHz, we report the first detection of the correlation between the CIB and CMB lensing. The well matched redshift distribution of these two signals leads to a detection significance with a peak value of 42 \sigma at 545 GHz and a correlation as high as 80% across these two tracers. Our full set of multifrequency measurements (both CIB auto and CIBlensing crossspectra) are consistent with a simple halobased model, with a characteristic mass scale for the halos hosting CIB sources of log_{10}(M/M_sun) = 10.5 \pm 0.6. Leveraging the frequency dependence of our signal, we isolate the high redshift contribution to the CIB, and constrain the star formation rate (SFR) density at z>1. We measure directly the SFR density with around 2 sigma significance for three redshift bins between z=1 and 7, thus opening a new window into the study of the formation of stars at early times.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: We present the Planck likelihood, a complete statistical description of the twopoint correlation function of the CMB temperature fluctuations. We use this likelihood to derive the Planck CMB power spectrum over three decades in l, covering 2 <= l <= 2500. The main source of error at l <= 1500 is cosmic variance. Uncertainties in smallscale foreground modelling and instrumental noise dominate the error budget at higher l's. For l < 50, our likelihood exploits all Planck frequency channels from 30 to 353 GHz through a physically motivated Bayesian component separation technique. At l >= 50, we employ a correlated Gaussian likelihood approximation based on angular crossspectra derived from the 100, 143 and 217 GHz channels. We validate our likelihood through an extensive suite of consistency tests, and assess the impact of residual foreground and instrumental uncertainties on cosmological parameters. We find good internal agreement among the highl crossspectra with residuals of a few uK^2 at l <= 1000. We compare our results with foregroundcleaned CMB maps, and with crossspectra derived from the 70 GHz Planck map, and find broad agreement in terms of spectrum residuals and cosmological parameters. The bestfit LCDM cosmology is in excellent agreement with preliminary Planck polarisation spectra. The standard LCDM cosmology is well constrained by Planck by l <= 1500. For example, we report a 5.4 sigma deviation from n_s /= 1. Considering various extensions beyond the standard model, we find no indication of significant departures from the LCDM framework. Finally, we report a tension between the bestfit LCDM model and the lowl spectrum in the form of a power deficit of 510% at l <~ 40, significant at 2.53 sigma. We do not elaborate further on its cosmological implications, but note that this is our most puzzling finding in an otherwise remarkably consistent dataset. (Abridged)03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: Planck has produced detailed allsky observations over nine frequency bands between 30 and 857 GHz. These observations allow robust reconstruction of the primordial cosmic microwave background (CMB) temperature fluctuations over nearly the full sky, as well as new constraints on Galactic foregrounds. This paper describes the component separation framework adopted by Planck. We test four foregroundcleaned CMB maps derived using qualitatively different component separation algorithms. The quality of our reconstructions is evaluated through detailed simulations and internal comparisons, and shown through various tests to be internally consistent and robust for CMB power spectrum and cosmological parameter estimation up to l = 2000. The parameter constraints on LambdaCDM cosmologies derived from these maps are consistent with those presented in the crossspectrum based Planck likelihood analysis. We choose two of the CMB maps for specific scientific goals. We also present maps and frequency spectra of the Galactic lowfrequency, CO, and thermal dust emission. The component maps are found to provide a faithful representation of the sky, as evaluated by simulations. For the lowfrequency component, the spectral index varies widely over the sky, ranging from about beta = 4 to 2. Considering both morphology and prior knowledge of the low frequency components, the index map allows us to associate a steep spectral index (beta < 3.2) with strong anomalous microwave emission, corresponding to a spinning dust spectrum peaking below 20 GHz, a flat index of beta > 2.3 with strong freefree emission, and intermediate values with synchrotron emission.03/2013;  03/2013;
 03/2013;
 [Show abstract] [Hide abstract]
ABSTRACT: Rotational transition lines of CO play a major role in molecular radio astronomy and in particular in the study of star formation and the Galactic structure. Although a wealth of data exists in the Galactic plane and some wellknown molecular clouds, there is no available CO high sensitivity allsky survey to date. Such allsky surveys can be constructed using the \Planck\ HFI data because the three lowest CO rotational transition lines at 115, 230 and 345 GHz significantly contribute to the signal of the 100, 217 and 353 GHz HFI channels respectively. Two different component separation methods are used to extract the CO maps from Planck HFI data. The maps obtained are then compared to one another and to existing external CO surveys. From these quality checks the best CO maps in terms of signal to noise and/or residual foreground contamination are selected. Three sets of velocityintegrated CO emission maps are produced: Type 1 maps of the CO (10), (21), and (32) rotational transitions with low foreground contamination but moderate signaltonoise ratio; Type 2 maps for the (10) and (21) transitions with a better signaltonoise ratio; and one Type 3 map, a line composite map with the best signaltonoise ratio in order to locate the faintest molecular regions. The maps are described in detail. They are shown to be fully compatible with previous surveys of parts of the Galactic Plane and also of fainter regions out of the Galactic plane. The Planck HFI velocityintegrated CO maps for the (10), (21), and (32) rotational transitions provide an unprecedented allsky CO view of the Galaxy. These maps are also of great interest to monitor potential CO contamination on CMB \Planck\ studies.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: We present the first results based on Planck measurements of the CMB temperature and lensingpotential power spectra. The Planck spectra at high multipoles are extremely well described by the standard spatiallyflat sixparameter LCDM cosmology. In this model Planck data determine the cosmological parameters to high precision. We find a low value of the Hubble constant, H0=67.3+/1.2 km/s/Mpc and a high value of the matter density parameter, Omega_m=0.315+/0.017 (+/1 sigma errors) in excellent agreement with constraints from baryon acoustic oscillation (BAO) surveys. Including curvature, we find that the Universe is consistent with spatial flatness to percentlevel precision using Planck CMB data alone. We present results from an analysis of extensions to the standard cosmology, using astrophysical data sets in addition to Planck and highresolution CMB data. None of these models are favoured significantly over standard LCDM. The deviation of the scalar spectral index from unity is insensitive to the addition of tensor modes and to changes in the matter content of the Universe. We find a 95% upper limit of r<0.11 on the tensortoscalar ratio. There is no evidence for additional neutrinolike relativistic particles. Using BAO and CMB data, we find N_eff=3.30+/0.27 for the effective number of relativistic degrees of freedom, and an upper limit of 0.23 eV for the summed neutrino mass. Our results are in excellent agreement with big bang nucleosynthesis and the standard value of N_eff=3.046. We find no evidence for dynamical dark energy. Despite the success of the standard LCDM model, this cosmology does not provide a good fit to the CMB power spectrum at low multipoles, as noted previously by the WMAP team. While not of decisive significance, this is an anomaly in an otherwise selfconsistent analysis of the Planck temperature data.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: The Planck High Frequency Instrument (HFI) spectral response was determined through a series of ground based tests conducted with the HFI focal plane in a cryogenic environment prior to launch. The main goal of the spectral transmission tests was to measure the relative spectral response (including outofband signal rejection) of all HFI detectors. This was determined by measuring the output of a continuously scanned Fourier transform spectrometer coupled with all HFI detectors. As there is no onboard spectrometer within HFI, the groundbased spectral response experiments provide the definitive data set for the relative spectral calibration of the HFI. The spectral response of the HFI is used in Planck data analysis and component separation, this includes extraction of CO emission observed within Planck bands, dust emission, SunyaevZeldovich sources, and intensity to polarization leakage. The HFI spectral response data have also been used to provide unit conversion and colour correction analysis tools. Verifications of the HFI spectral response data are provided through comparisons with photometric HFI flight data. This validation includes use of HFI zodiacal emission observations to demonstrate outofband spectral signal rejection better than 10^8. The accuracy of the HFI relative spectral response data is verified through comparison with complementary flightdata based unit conversion coefficients and colour correction coefficients. These coefficients include those based upon HFI observations of CO, dust, and SunyaevZeldovich emission. General agreement is observed between the groundbased spectral characterization of HFI and corresponding inflight observations, within the quoted uncertainty of each; explanations are provided for any discrepancies.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: Based on CMB maps from the 2013 Planck Mission data release, this paper presents the detection of the ISW effect, i.e., the correlation between the CMB and largescale evolving gravitational potentials. The significance of detection ranges from 2 to 4 sigma, depending on which method is used. We investigate three separate approaches, which cover essentially all previous studies, as well as breaking new ground. (i) Correlation of the CMB with the Planck reconstructed gravitational lensing potential (for the first time). This detection is made using the lensinginduced bispectrum; the correlation between lensing and the ISW effect has a significance close to 2.5 sigma. (ii) Crosscorrelation with tracers of LSS, yielding around 3 sigma significance, based on a combination of radio (NVSS) and optical (SDSS) data. (iii) Aperture photometry on stacked CMB fields at the locations of known largescale structures, which yields a 4 sigma signal when using a previously explored catalogue, but shows strong discrepancies in amplitude and scale compared to expectations. More recent catalogues give more moderate results, ranging from negligible to 2.5 sigma at most, but with a more consistent scale and amplitude, the latter being still slightly above what is expected from numerical simulations within LCMD. Where they can be compared, these measurements are compatible with previous work using data from WMAP, which had already mapped these scales to the limits of cosmic variance. Planck's broader frequency coverage confirms that the signal is achromatic, bolstering the case for ISW detection. As a final step we use tracers of largescale structure to filter the CMB data, presenting maps of the ISW temperature perturbation. These results provide complementary and independent evidence for the existence of a dark energy component that governs the current accelerated expansion of the Universe.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: The Planck Catalogue of Compact Sources (PCCS) is the catalogue of sources detected in the first 15 months of Planck operations, the "nominal" mission. It consists of nine singlefrequency catalogues of compact sources, both Galactic and extragalactic, detected over the entire sky. The PCCS covers the frequency range 30857\,GHz with higher sensitivity (it is 90% complete at 180 mJy in the best channel) and better angular resolution (from ~33' to ~5') than previous allsky surveys in this frequency band. By construction its reliability is >80% and more than 65% of the sources have been detected at least in two contiguous Planck channels. In this paper we present the construction and validation of the PCCS, its contents and its statistical characterization.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: The two fundamental assumptions of the standard cosmological model  that the initial fluctuations are statistically isotropic and Gaussian  are rigorously tested using maps of the cosmic microwave background (CMB) anisotropy from the Planck satellite. Deviations from isotropy have been found and demonstrated to be robust against component separation algorithm, mask choice and frequency dependence. Many of these anomalies were previously observed in the WMAP data, and are now confirmed at similar levels of significance (about 3 sigma). However, we find little evidence for nonGaussianity, with the exception of a few statistical signatures that seem to be associated with specific anomalies. In particular, we find that the quadrupoleoctopole alignment is also connected to a low observed variance of the CMB signal. A power asymmetry is now found to persist to scales corresponding to about l=600, and can be described in the lowl regime by a phenomenological dipole modulation model. However, any primordial power asymmetry is strongly scaledependent and does not extend to arbitrarily small angular scales. Finally, it is plausible that some of these features may be reflected in the angular power spectrum of the data, which shows a deficit of power on similar scales. Indeed, when the power spectra of two hemispheres defined by a preferred direction are considered separately, one shows evidence for a deficit in power, while its opposite contains oscillations between odd and even modes that may be related to the parity violation and phase correlations also detected in the data. Although these analyses represent a step forward in building an understanding of the anomalies, a satisfactory explanation based on physically motivated models is still lacking.Astronomy and Astrophysics 03/2013; · 4.48 Impact Factor  [Show abstract] [Hide abstract]
ABSTRACT: We describe the data processing pipeline of the Planck Low Frequency Instrument (LFI) data processing centre (DPC) to create and characterize fullsky maps based on the first 15.5 months of operations at 30, 44 and 70 GHz. In particular, we discuss the various steps involved in reducing the data, starting from telemetry packets through to the production of cleaned, calibrated timelines and calibrated frequency maps. Data are continuously calibrated using the modulation induced on the mean temperature of the cosmic microwave background radiation by the proper motion of the spacecraft. Sky signals other than the dipole are removed by an iterative procedure based on simultaneous fitting of calibration parameters and sky maps. Noise properties are estimated from timeordered data after the sky signal has been removed, using a generalized least square mapmaking algorithm. A destriping code (Madam) is employed to combine radiometric data and pointing information into sky maps, minimizing the variance of correlated noise. Noise covariance matrices, required to compute statistical uncertainties on LFI and Planck products, are also produced. Main beams are estimated down to the 20 dB level using Jupiter transits, which are also used for the geometrical calibration of the focal plane.Astronomy and Astrophysics (Submitted). 03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: This paper presents the characterization of the inflight beams, the beam window functions and the associated uncertainties for the Planck Low Frequency Instrument (LFI). Knowledge of the beam profiles is necessary for determining the transfer function to go from the observed to the actual sky anisotropy power spectrum. The main beam distortions affect the beam window function, complicating the reconstruction of the anisotropy power spectrum at high multipoles, whereas the sidelobes affect the low and intermediate multipoles. The inflight assessment of the LFI main beams relies on the measurements performed during Jupiter observations. By stacking the data from multiple Jupiter transits, the main beam profiles are measured down to 20 dB at 30 and 44 GHz, and down to 25 dB at 70 GHz. The main beam solid angles are determined to better than 0.2% at each LFI frequency band. The Planck prelaunch optical model is conveniently tuned to characterize the main beams independently of any noise effects. This approach provides an optical model whose beams fully reproduce the measurements in the main beam region, but also allows a description of the beams at power levels lower than can be achieved by the Jupiter measurements themselves. The agreement between the simulated beams and the measured beams is better than 1% at each LFI frequency band. The simulated beams are used for the computation of the window functions for the effective beams. The error budget for the window functions is estimated from both main beam and sidelobe contributions, and accounts for the radiometer bandshapes. The total uncertainties in the effective beam window functions are: 2% and 1.2% at 30 and 44 GHz, respectively (at $\ell \approx 600$), and 0.7% at 70 GHz (at $\ell \approx 1000$).03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: The Planck nominal mission cosmic microwave background (CMB) maps yield unprecedented constraints on primordial nonGaussianity (NG). Using three optimal bispectrum estimators, separable templatefitting (KSW), binned, and modal, we obtain consistent values for the primordial local, equilateral, and orthogonal bispectrum amplitudes, quoting as our final result fNL^local= 2.7+/5.8, fNL^equil= 42+/75, and fNL^ortho= 25+\39 (68% CL statistical). NG is detected in the data; using skewC_l statistics we find a nonzero bispectrum from residual point sources, and the ISWlensing bispectrum at a level expected in the LambdaCDM scenario. The results are based on comprehensive crossvalidation of these estimators on Gaussian and nonGaussian simulations, are stable across component separation techniques, pass an extensive suite of tests, and are confirmed by skewC_l, wavelet bispectrum and Minkowski functional estimators. Beyond estimates of individual shape amplitudes, we present modelindependent, 3dimensional reconstructions of the Planck CMB bispectrum and thus derive constraints on earlyUniverse scenarios that generate primordial NG, including general singlefield models of inflation, excited initial states (nonBunchDavies vacua), and directionallydependent vector models. We provide an initial survey of scaledependent feature and resonance models. These results bound both general singlefield and multifield model parameter ranges, such as the speed of sound, c_s \geq 0.02 (95% CL), in an effective field theory parametrization, and the curvaton decay fraction r_D \geq 0.15 (95% CL). The Planck data significantly limit the viable parameter space of the ekpyrotic/cyclic scenarios. The amplitude of the 4point function in the local model tauNL < 2800 (95% CL). These constraints represent the highest precision tests to date of physical mechanisms for the origin of cosmic structure.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: We analyse the implications of the Planck data for cosmic inflation. The Planck nominal mission temperature anisotropy measurements, combined with the WMAP largeangle polarization, constrain the scalar spectral index to $n_s = 0.9603 \pm 0.0073$, ruling out exact scale invariance at over 5 $\sigma$. Planck establishes an upper bound on the tensortoscalar ratio of r < 0.11 (95% CL). The Planck data thus shrink the space of allowed standard inflationary models, preferring potentials with V'' < 0. Exponential potential models, the simplest hybrid inflationary models, and monomial potential models of degree n > 2 do not provide a good fit to the data. Planck does not find statistically significant running of the scalar spectral index, obtaining $d n_s/d ln k = 0.0134 \pm 0.0090$. Several analyses dropping the slowroll approximation are carried out, including detailed model comparison and inflationary potential reconstruction. We also investigate whether the primordial power spectrum contains any features. We find that models with a parameterized oscillatory feature improve the fit $\chi^2$ by ~ 10; however, Bayesian evidence does not prefer these models. We constrain several singlefield inflation models with generalized Lagrangians by combining power spectrum data with bounds on $f_\mathrm{NL}$ measured by Planck. The fractional primordial contribution of CDM isocurvature modes in the curvaton and axion scenarios has upper bounds of 0.25% or 3.9% (95% CL), respectively. In models with arbitrarily correlated CDM or neutrino isocurvature modes, an anticorrelation can improve $\chi^2$ by approximatively 4 as a result of slightly lowering the theoretical prediction for the $\ell<40$ multipoles relative to the higher multipoles. Nonetheless, the data are consistent with adiabatic initial conditions.03/2013;  [Show abstract] [Hide abstract]
ABSTRACT: This paper characterizes the effective beams,the effective beam window functions and the associated errors for the Planck HFI detectors. The effective beam is the angular response including the effect of the optics,detectors,data processing and the scan strategy. The window function is the representation of this beam in the harmonic domain which is required to recover an unbiased measurement of the CMB angular power spectrum. The HFI is a scanning instrument and its effective beams are the convolution of: (a) the optical response of the telescope and feeds;(b)the processing of the timeordered data and deconvolution of the bolometric and electronic time response; and (c) the merging of several surveys to produce maps. The time response functions are measured using observations of Jupiter and Saturn and by minimizing survey difference residuals. The scanning beam is the postdeconvolution angular response of the instrument, and is characterized with observations of Mars. The main beam solid angles are determined to better than 0.5% at each HFI frequency band. Observations of Jupiter and Saturn limit near sidelobes (within 5deg) to about 0.1% of the total solid angle. Time response residuals remain as long tails in the scanning beams, but contribute less than 0.1% of the total. The bias and uncertainty in the beam products are estimated using ensembles of simulated planet observations that include the impact of instrumental noise and known systematic effects.The correlation structure of these ensembles is welldescribed by five error eigenmodes that are subdominant to sample variance and instrumental noise in the harmonic domain. A suite of consistency tests provide confidence that the error model represents a sufficient description of the data. The total error in the effective beam window functions is below 1% at 100GHz up to ell~1500$,and below 0.5% at 143 and 217GHz up to ~2000.03/2013;
Publication Stats
9k  Citations  
1,048.72  Total Impact Points  
Top Journals
Institutions

2014

HarvardSmithsonian Center for Astrophysics
 Smithsonian Astrophysical Observatory
Cambridge, Massachusetts, United States


1997–2014

California Institute of Technology
 • Jet Propulsion Laboratory
 • Department of Physics
 • Division of Physics, Mathematics, and Astronomy
Pasadena, California, United States


2013

European Southern Observatory
Arching, Bavaria, Germany 
McGill University
 Department of Physics
Montréal, Quebec, Canada


2012

University of California, Irvine
 Department of Physics and Astronomy
Irvine, California, United States


2003–2011

University of Colorado at Boulder
 Center for Astrophysics and Space Astronomy
Boulder, Colorado, United States 
University of Chicago
 Department of Astronomy and Astrophysics
Chicago, IL, United States


2009

Cardiff University
 School of Physics and Astronomy
Cardiff, Wales, United Kingdom


2008–2009

University of Pennsylvania
 Department of Physics and Astronomy
Philadelphia, PA, United States


2006

NASA
 Jet Propulsion Laboratory
Washington, WV, United States


2004

University of Wales
Cardiff, Wales, United Kingdom


2002

Carnegie Mellon University
 Department of Physics
Pittsburgh, PA, United States


2000

Institute for Advanced Study
Princeton Junction, New Jersey, United States


1995

University of California, Berkeley
 Department of Physics
Berkeley, California, United States
