Share
Embed
Why cosmology should care about the Milky Way
JPL Postdoc Seminar Series talk on cosmological foregrounds
Daniel
September 28, 2017
Transcript
81 slides
-
Why cosmology should care about the Milky Way in collaboration
with O. Doré, B. Hensley, P. Bull, G. Lagache, P. Serra Daniel Lenz JPL Postdoc Seminar Series Sept 28, 2017 © 2017 California Institute of Technology. Government sponsorship acknowledged. -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Why should cosmologists
care about the Milky Way? 2 'Galactic' astronomers… ❖ … want to understand how galaxies are formed/ evolve/merge etc. ❖ … observe and model our Galaxy in different wavelengths -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 3 Why should
cosmologists care about the Milky Way? 'Cosmologists' … ❖ … want to understand the history, evolution and fate of the entire Universe ❖ …study millions of galaxies and their statistical properties Hubble Ultra Deep Field -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 3 Why should
cosmologists care about the Milky Way? 'Cosmologists' … ❖ … want to understand the history, evolution and fate of the entire Universe ❖ …study millions of galaxies and their statistical properties Hubble Ultra Deep Field Warm dark matter Cold dark matter Bromm+ (2009) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 4 Why should
cosmologists care about the Milky Way? 'Cosmologists' … ❖ … want to understand the history, evolution and fate of the entire Universe ❖ …study cosmological backgrounds Cosmic Microwave Background (Planck) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 5 Why should
cosmologists care about the Milky Way? NASA/WMAP -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 6 Why should
cosmologists care about the Milky Way? -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Cosmology and the
Milky Way: Synergies 7 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Cosmology and the
Milky Way: Synergies ❖ Our Galaxy is an inconvenient foreground for Cosmology 7 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Cosmology and the
Milky Way: Synergies ❖ Our Galaxy is an inconvenient foreground for Cosmology ❖ Cosmology needs a model of the Milky Way to remove it 7 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Cosmology and the
Milky Way: Synergies ❖ Our Galaxy is an inconvenient foreground for Cosmology ❖ Cosmology needs a model of the Milky Way to remove it ❖ Galactic astronomers benefit from cosmology missions, giving them fantastic data to work with 7 -
Which foregrounds do we care about? "(…) the name of
the game is component separation, not noise reduction" H.K. Eriksen, 'Advances in Theoretical Cosmology in Light of Data 2017' ❖ Extinction for cosmological galaxy surveys ❖ Cosmic infrared background measurements ❖ De-lensing of CMB data for primordial gravitational waves Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 8 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology E(B-V) ❖ E(B-V)
= Extinction in B band - Extinction in V band ❖ More dust => larger E(B-V) ❖ E(B-V) maps essential for correcting observations for Galactic reddening 10 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Mapping E(B-V): Direct
approach 11 ❖ Find many sources with known spectrum (e.g. stars, passive galaxies) ❖ Measure spectra, attribute differences to dust ❖ E.g. Schlafly+ (2014), Green+ (2015) used 500 million stars from Pan- STARRS to measure reddening to 4.5 kpc -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Mapping E(B-V): Direct
approach 11 ❖ Find many sources with known spectrum (e.g. stars, passive galaxies) ❖ Measure spectra, attribute differences to dust ❖ E.g. Schlafly+ (2014), Green+ (2015) used 500 million stars from Pan- STARRS to measure reddening to 4.5 kpc ❖ Direct measurements are hard! ❖ Photometric/ spectroscopic errors ❖ Ensuring sources lie behind full dust column ❖ Ensuring adequate number of sources have been measured -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Mapping E(B-V): Direct
approach 11 ❖ Find many sources with known spectrum (e.g. stars, passive galaxies) ❖ Measure spectra, attribute differences to dust ❖ E.g. Schlafly+ (2014), Green+ (2015) used 500 million stars from Pan- STARRS to measure reddening to 4.5 kpc ❖ Direct measurements are hard! ❖ Photometric/ spectroscopic errors ❖ Ensuring sources lie behind full dust column ❖ Ensuring adequate number of sources have been measured Green+ (2015) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Dust emission as
measure of E(B-V) ❖ E(B-V) is proportional to the dust column, so can convert dust column tracer to E(B-V) ❖ SFD used dust emission from IRAS to derive a calibration factor from FIR emission to E(B-V) ❖ Full-sky, high sensitivity measurements -2 -0.3 log10 (E(B V )SFD [mag]) Reddening map of Schlegel, Finkbeiner, and Davis (1998) 12 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology ❖ Requires a
temperature correction to go from dust emission to a dust column density ❖ FIR emission may have contributions from zodiacal light and unresolved galaxies -2 -0.3 log10 (E(B V )SFD [mag]) Reddening map of Schlegel, Finkbeiner, and Davis (1998) 13 Dust emission as measure of E(B-V) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology ❖ Requires a
temperature correction to go from dust emission to a dust column density ❖ FIR emission may have contributions from zodiacal light and unresolved galaxies -2 -0.3 log10 (E(B V )SFD [mag]) Reddening map of Schlegel, Finkbeiner, and Davis (1998) 13 Dust emission as measure of E(B-V) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology HI emission as
basis for E(B-V) ❖ Neutral atomic hydrogen (HI), most abundant element ❖ Gas and dust are well-coupled in the interstellar medium ❖ Resulting maps free from errors due to dust temperature, zodi, and extragalactic emission ❖ Limited by non-HI gas along the line of sight 14 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Galactic HI Surveys
16 Credit: S. Janowiecki Southern hemisphere: Galactic All-Sky Survey (GASS) McClure-Griffiths+ (2009) Kalberla+ (2010, 2015) D. McClenaghan -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Galactic HI Surveys
16 B. Winkel Northern hemisphere: Effelsberg-Bonn HI Survey (EBHIS) Kerp+ (2011) Winkel, DL+ (2010, 2016) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology HI4PI Survey ❖
Merges data from Effelsberg and Parkes ❖ New state-of-the-art survey ❖ Higher sensitivity & resolution, fewer systematics, full sampling 20 21 22 log(NHI [cm 2]) 180 135 90 45 0 315 270 225 180 60 30 0 30 60 HI4PI collaboration (2017) 17 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology HI emission as
basis for E(B-V) ❖ Use HI4PI column density as new measure of dust reddening ❖ How well does it actually fit other reddening tracers? 19 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 10 1 100
(B V ) [mag] 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 NHI [cm 2] ⇥1020 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 E(B V ) [mag] E(B V ) [mag] = 1.216+0.009 0.009 ⇥ NHI [1022 cm 2] + 0.015+0.0002 0.0002 [mag] = 0.02406+0.00006 0.00006 102 103 104 # data points 0 10 20 30 40 50 60 70 80 # data points 10 1 100 (B V ) [mag] 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 NHI [cm 2] ⇥1020 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 E(B V ) [mag] E(B V ) [mag] = 1.113+0.002 0.002 ⇥ NHI [1022 cm 2] + 0.000+0.0001 0.0001 [mag] = 0.00570+0.00001 0.00001 102 103 104 # data points 0 100 200 300 400 # data points The E(B-V)/NHI ratio Pan-STARRS E(B-V), Schlafly+ (2014) SFD E(B-V) Star-based Dust-based 20 DL, Hensley, Doré (2017) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 10 1 100
(B V ) [mag] 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 NHI [cm 2] ⇥1020 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 E(B V ) [mag] E(B V ) [mag] = 1.216+0.009 0.009 ⇥ NHI [1022 cm 2] + 0.015+0.0002 0.0002 [mag] = 0.02406+0.00006 0.00006 102 103 104 # data points 0 10 20 30 40 50 60 70 80 # data points 10 1 100 (B V ) [mag] 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 NHI [cm 2] ⇥1020 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 E(B V ) [mag] E(B V ) [mag] = 1.113+0.002 0.002 ⇥ NHI [1022 cm 2] + 0.000+0.0001 0.0001 [mag] = 0.00570+0.00001 0.00001 102 103 104 # data points 0 100 200 300 400 # data points The E(B-V)/NHI ratio Pan-STARRS E(B-V), Schlafly+ (2014) SFD E(B-V) Star-based Dust-based 20 DL, Hensley, Doré (2017) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology HI-based reddening model
❖ For the full sky, allow different E(B-V)/NHI ratios for different velocities ❖ High-velocity gas has less dust, as expected 21 102 101 0 101 102 vLSR [km s 1] 0.000 0.005 0.010 0.015 0.020 0.025 ↵v = E(B V )/Nv HI [mag/1020 cm 2] ↵v DL, Hensley, Doré (2017) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 130 112 95
l [deg] 42 52 62 b [deg] SFD Modelsimple 130 112 95 l [deg] SFD Modeltophat 0.00 0.02 0.04 0.016 0.008 0.000 0.008 0.016 E(B V ) [mag] 102 103 104 105 # data points HI-based reddening model ❖ Black: HI high-velocity clouds ❖ Color: Reddening residuals 22 Full HI column density Only local HI (|v| < 90 km/s) DL, Hensley, Doré (2017) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 0 0.05 E(B
V )model [mag] The E(B-V) map 40% sky coverage, 16.1’ resolution 23 DL, Hensley, Doré (2017) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Dust systematics ❖
Peek & Graves (2010) used SDSS passively evolving galaxies as "standard crayons" ❖ Correction to the SFD map at 4.5 deg 24 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Dust systematics Based
on extragalactic sources Based on galactic HI 25 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology When and why
to use this extinction map ❖ New HI-based extinction map ❖ In line with independent corrections, but much higher resolution and better sky coverage ❖ Yahata+ (2007) find correlation of SFD with large-scale structure ❖ For high latitudes, our map overcomes many of the SFD problems and is much more sensitive than stellar data- based E(B-V) maps 26 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology The Cosmic Infrared
Background What is the CIB? 27 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology The Cosmic Infrared
Background What is the CIB? 27 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology The Cosmic Infrared
Background ❖ Unresolved background radiation ❖ Made up from dust in galaxies at z=1-3 Lagache+ (2002) 28 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology The Cosmic Infrared
Background ❖ Unresolved background radiation ❖ Made up from dust in galaxies at z=1-3 Lagache+ (2002) 28 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology The CIB as
cosmological probe … of star formation history Planck collaboration (2013 XXX) ❖ Strong constraints on SFH up to z=2.5 ❖ Probe dust temperature across cosmic times ❖ Understand star formation in DM halos 29 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology ❖ CMB lensing
and CIB match great in z and MHalo ❖ Ideal probe of relation between dark and luminous matter … of large scale structure to cross-correlate with lensing Planck collaboration (2014 XVIII) The CIB as cosmological probe 30 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology ❖ CMB lensing
and CIB match great in z and MHalo ❖ Ideal probe of relation between dark and luminous matter … of large scale structure to cross-correlate with lensing Planck collaboration (2014 XVIII) The CIB as cosmological probe 30 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology How to obtain
CIB maps? ❖ Galactic thermal dust and CIB dust dominate on large scales at ~200 to 1000 GHz ❖ How to disentangle them? 31 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology How to obtain
CIB maps? A. Fit different frequency channels with modified blackbody spectra B. Utilize the different angular power spectra of these components C. Use template maps of Galactic dust (e.g. HI-based) ❖ Galactic thermal dust and CIB dust dominate on large scales at ~200 to 1000 GHz ❖ How to disentangle them? 31 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Current CIB maps:
Planck (2013 XXX) ❖ Limited sky coverage, hard to access large scales 32 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Modeling dust foregrounds
• • Velocity separation difficult for complex structures and large scales Radial Velocity HVC IVC LVC I = ✏HVC NHVC + ✏IVC NIVC + ✏LVC NLVC 33 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Modeling dust foregrounds
• Generalised linear model (GLM) • Radial Velocity I = X i ✏iTi B 34 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Modeling dust foregrounds
• Generalised linear model (GLM) • • Regularised: • • Accounts for all features along line of sight I = X i ✏iTi B Radial Velocity | Datai Modeli |2 + ↵ · |✏i | 35 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology CIB: Galactic poles
Total FIR intensity 36 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology CIB: Challenges ❖
How to verify, especially on large scales? ❖ Simulations and re-sampling 37 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology CIB: Challenges ❖
How to verify, especially on large scales? ❖ Simulations and re-sampling ❖ How to avoid fine-tuning of the component separation? 37 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology CIB: Challenges ❖
How to verify, especially on large scales? ❖ Simulations and re-sampling ❖ How to avoid fine-tuning of the component separation? ❖ How to extend this larger areas, probing non-HI gas? 37 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology CIB: Challenges ❖
How to verify, especially on large scales? ❖ Simulations and re-sampling ❖ How to avoid fine-tuning of the component separation? ❖ How to extend this larger areas, probing non-HI gas? ❖ How to jointly use frequency information and angular power spectra? 37 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology De-lensing the cosmic
microwave background 38 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology De-lensing the CMB
39 ❖ Large-scale structure lenses the CMB, smoothes the power spectrum -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology CMB de-lensing: TT
❖ ❖ Ideally: Construct lensing potential from CMB itself ❖ Too noisy, but will be relevant for CMB-S3 and CMB-S4 ❖ Currently: Use external tracer of lensing potential such as the CIB 42 Te↵( x ) = Ttrue( x + r ( x )) -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology CMB de-lensing: TT
43 ❖ Using large-scale CIB maps to de-lens Planck CMB TT ❖ Characteristic sharpening of the peaks detected at high significance ❖ Sharpening can hardly be mimicked by other effects Larsen+ 2016 Large scales Small scales -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology De-lensing: TT ❖
Underlying CIB maps very simplistic: M545 - M857/77 ❖ Only 40% correlated (ideal CIB would give up to 80%) ❖ High resolution, high accuracy, large scale CIB maps needed 44 Larsen+ 2016 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology De-lensing: TT ❖
Underlying CIB maps very simplistic: M545 - M857/77 ❖ Only 40% correlated (ideal CIB would give up to 80%) ❖ High resolution, high accuracy, large scale CIB maps needed 44 Larsen+ 2016 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology CMB de-lensing of
BB: The challenge ❖ Lensing of CMB E-modes leads to apparent B-modes ❖ One of the major systematics in the search for primordial gravitational waves 45 Courtesy A. Challinor Large scales Small scales -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Manzotti+ (2017) 46
CMB de-lensing of BB ❖ Wiener-filter E-mode map and CIB combined in harmonic space to build model of lensing B-modes -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology ❖ Herschel 500
micron as CIB template ❖ "No lensing" excluded at 7 sigma Manzotti+ (2017) 47 CMB de-lensing of BB Large scales Small scales -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology 48 Courtesy A.
Manzotti CMB de-lensing of BB: Outlook Large scales Small scales Correlation with lensing -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Summary on the
CIB ❖ Template maps of Galactic dust are key to obtain large- scale CIB maps ❖ Useful by itself to constrain star formation history and connection of dark and luminous matter ❖ Best tool to-date for de-lensing CMB TT and BB, better CIB maps are urgently required 49 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Current CIB maps:
GNILC ❖ Planck collaboration (2016 XLVIII), focus on removing CIB from Galactic dust maps ❖ Using the angular power spectra of the two components ❖ Does not agree that well on a pixel-to-pixel basis 51 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Current CIB maps:
GNILC ❖ Planck collaboration (2016 XLVIII), focus on removing CIB from Galactic dust maps ❖ Using the angular power spectra of the two components ❖ Does not agree that well on a pixel-to-pixel basis 51 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Current CIB maps:
GNILC ❖ Planck collaboration (2016 XLVIII), focus on removing CIB from Galactic dust maps ❖ Using the angular power spectra of the two components ❖ Does not agree that well on a pixel-to-pixel basis 51 0.3 0.2 0.1 0.0 0.1 0.2 0.3 Planck XXX 0.3 0.2 0.1 0.0 0.1 0.2 0.3 GNILC 100 101 102 103 -
Daniel Lenz, Caltech/JPL Foregrounds in observational cosmology Current CIB maps:
GNILC ❖ Cross correlation with Planck CMB lensing ❖ Missing CIB power, especially on the largest scales 52 -
Residuals vs. dust temperature 16 18 20 22 24 26
28 Tdust [K] 0.04 0.02 0.00 0.02 0.04 (E(B V )) [mag] SFD Model 16 18 20 22 24 26 28 Tdust [K] MF Model 16 18 20 22 24 26 28 Tdust [K] MF SFD 100 101 102 103 104 105 # data points -
Residuals vs. ecliptic lat 0.050 0.025 0.000 0.025 0.050 SFD
Model [mag] 50 0 50 Ecliptic latitude [deg] 0.050 0.025 0.000 0.025 0.050 MF Model [mag] 100 101 102 103 104 105 # data points -
20 40 60 80 100 120 |vcut LSR | [km
s 1] 0.0 0.2 0.4 0.6 0.8 1.0 Normalized RSS Tophat search -
Residuals intern 0.004 0.002 0.000 0.002 0.004 µ ( E(B
V )) [mag] SFD - Model v-binned SFD - Model tophat SFD - Model simple 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 NHI [1020 cm 2] 0.000 0.002 0.004 0.006 0.008 0.010 ( E(B V )) [mag] SFD - Model v-binned SFD - Model tophat SFD - Model simple -
Residuals extern 0.005 0.000 0.005 0.010 0.015 0.020 µ (
E(B V )) [mag] SFD - Model MF - Model Planck - Model 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 NHI [1020 cm 2] 0.000 0.002 0.004 0.006 0.008 0.010 0.012 ( E(B V )) [mag] SFD - Model MF - Model Planck - Model -
1020 1021 1022 NHI [cm 2] 10 3 10 2
10 1 100 E(B V ) [mag] 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 NHI [cm 2] ⇥1020 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 E(B V ) [mag] E(B V ) [mag] = 1.216+0.009 0.009 ⇥ NHI [1022 cm 2] + 0.015+0.0002 0.0002 [mag] = 0.02406+0.00006 0.00006 100 101 102 103 104 # data points 0 10 20 30 40 50 60 70 80 # data points 1020 1021 1022 NHI [cm 2] 10 3 10 2 10 1 100 E(B V ) [mag] 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 NHI [cm 2] ⇥1020 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 E(B V ) [mag] E(B V ) [mag] = 1.113+0.002 0.002 ⇥ NHI [1022 cm 2] + 0.000+0.0001 0.0001 [mag] = 0.00570+0.00001 0.00001 100 101 102 103 104 # data points 0 100 200 300 400 # data points The E(B-V)/NHI ratio Pan-STARRS E(B-V), Schlafly+ (2014) SFD E(B-V) Star-based Dust-based 59