The other side of the equation Systematic effects in the determination of dust masses Peter Scicluna ASIAA JCMT Users’ Meeting, Nanjing, 13th February 2017 Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 1 / 9 Done for LMC (Reibel+ 2012), SMC (Srinivasan+2016) Total dust mass ∼ integrated dust production over Hubble time What about dust destruction? Dust growth in the ISM? Are the masses really correct? The dust budget crisis: locally Measure dust masses in FIR/sub-mm Measure dust production in MIR Gordon et al., 2014 Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 2 / 9 What about dust destruction? Dust growth in the ISM? Are the masses really correct? The dust budget crisis: locally Measure dust masses in FIR/sub-mm Measure dust production in MIR Done for LMC (Reibel+ 2012), SMC (Srinivasan+2016) Total dust mass ∼ integrated dust production over Hubble time Gordon et al., 2014 Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 2 / 9 Dust growth in the ISM? Are the masses really correct? The dust budget crisis: locally Measure dust masses in FIR/sub-mm Measure dust production in MIR Done for LMC (Reibel+ 2012), SMC (Srinivasan+2016) Total dust mass ∼ integrated dust production over Hubble time What about dust destruction? Gordon et al., 2014 Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 2 / 9 How important are supernovae? Are there even enough metals yet? Dust growth in the ISM? Are the masses really correct? The dust budget crisis: at high redshift Rowlands et al., 2014 Measure dust masses in FIR/sub-mm Can’t measure production Large dust masses exist already Too soon for AGB stars Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 2 / 9 Dust growth in the ISM? Are the masses really correct? The dust budget crisis: at high redshift Rowlands et al., 2014 Measure dust masses in FIR/sub-mm Can’t measure production Large dust masses exist already Too soon for AGB stars How important are supernovae? Are there even enough metals yet? Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 2 / 9 The dust budget crisis: Rowlands et al., 2014 Measure dust masses in FIR/sub-mm Dust growth in the ISM? Are the masses really correct? Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 2 / 9 - free parameters: Md, Td, βFIR Fit (modified) black body Depends on measurements of optical properties of dust! (κ) Are extrapolations really okay? How are dust masses determined? Observe rest-frame FIR photometry 101 Flux 100 102 103 Wavelength [µm] Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 3 / 9 - free parameters: Md, Td, βFIR Depends on measurements of optical properties of dust! (κ) Are extrapolations really okay? How are dust masses determined? Observe rest-frame FIR photometry Fit (modified) black body 101 Flux 100 102 103 Wavelength [µm] Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 3 / 9 Depends on measurements of optical properties of dust! (κ) Are extrapolations really okay? How are dust masses determined? Observe rest-frame FIR photometry Fit (modified) black body d M , T - free parameters: Md, Td, βFIR T,(β) 101 Flux 100 β 102 103 Wavelength [µm] Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 3 / 9 Are extrapolations really okay? How are dust masses determined? Observe rest-frame FIR photometry Fit (modified) black body d M , T - free parameters: Md, Td, βFIR T,(β) 101 Depends on measurements of optical properties of dust! (κ) Flux 100 β 102 103 Wavelength [µm] Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 3 / 9 How are dust masses determined? Observe rest-frame FIR photometry Fit (modified) black body d M , T - free parameters: Md, Td, βFIR T,(β) 101 Depends on measurements of optical properties of dust! (κ) Are extrapolations really okay? Flux 100 β 102 103 Wavelength [µm] Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 3 / 9 Depends on temperature! Not just a single, simple power law Not accounted for in current dust models! New lab measurements New measurements in FIR/sub-mm Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 4 / 9 Not just a single, simple power law Not accounted for in current dust models! New lab measurements New measurements in FIR/sub-mm Depends on temperature! Mennella et al. (1998) Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 4 / 9 Not accounted for in current dust models! New lab measurements New measurements in FIR/sub-mm Depends on temperature! Not just a single, simple power law Coupeaud et al. (2011) Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 4 / 9 New lab measurements New measurements in FIR/sub-mm Depends on temperature! Not just a single, simple power law Not accounted for in current dust models! Coupeaud et al. (2011) Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 4 / 9 Flatter βFIR, not constant Fraction of amC crucial - Can only reproduce old values with no carbon! short wavelengths: converge to existing models long wavelengths: much stronger emission! - × ∼ 4 @ 230 µm - × ∼ 8 @ 850 µm - × ∼ 10 @ 1300 µm Dust models with the new lab measurements Temperature-dependent dust models Compact spheres 3 10 Hollow spheres Ellipsoids Typical values used in literature by observers 102 ] 1 − g 2 m c 1 [ 10 s b a 100 10-1 102 103 Wavelength [µm] Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 5 / 9 Fraction of amC crucial - Can only reproduce old values with no carbon! short wavelengths: converge to existing models long wavelengths: much stronger emission! - × ∼ 4 @ 230 µm - × ∼ 8 @ 850 µm - × ∼ 10 @ 1300 µm Dust models with the new lab measurements Temperature-dependent dust models Compact spheres Flatter βFIR, not constant 3 10 Hollow spheres Ellipsoids Typical values used in literature by observers 102 ] 1 − g 2 m c 1 [ 10 s b a 100 10-1 102 103 Wavelength [µm] Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 5 / 9 short wavelengths: converge to existing models long wavelengths: much stronger emission! - × ∼ 4 @ 230 µm - × ∼ 8 @ 850 µm - × ∼ 10 @ 1300 µm Dust models with the new lab measurements Temperature-dependent dust models Compact spheres Flatter βFIR, not constant 3 10 Hollow spheres Fraction of amC crucial Ellipsoids Typical values used in literature by observers Can only reproduce old values with no 2 - 10 ] 1 carbon! − g 2 m c 1 [ 10 s b a 100 10-1 102 103 Wavelength [µm] Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 5 / 9 Dust models with the new lab measurements Temperature-dependent dust models Compact spheres Flatter βFIR, not constant 3 10 Hollow spheres Fraction of amC crucial Ellipsoids Typical values used in literature by observers Can only reproduce old values with no 2 - 10 ] 1 carbon! − g 2 m short wavelengths: converge to existing c 1 [ 10 s b models a long wavelengths: much stronger 100 emission! - × ∼ 4 @ 230 µm 10-1 - × ∼ 8 @ 850 µm 102 103 Wavelength [µm] - × ∼ 10 @ 1300 µm Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 5 / 9 Slightly better at moderate redshift Implications for astronomy z=0 20 M fitted, T =50K, β = 1.5 d d − T & M fitted, β = 1.5 d d − Td, Md & β fitted 15 Dust masses overestimated by large 10 fraction Ratio Depends strongly on fitting method, temperature and wavelength (redshift!) 5 0 10-6 10-5 10-4 10-3 10-2 10-1 100 Fraction of hot dust Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 6 / 9 Implications for astronomy z=1 16 M fitted, T =50K, β = 1.5 d d − 14 T & M fitted, β = 1.5 d d − Td, Md & β fitted 12 Dust masses overestimated by large 10 8 fraction Ratio Depends strongly on fitting method, 6 temperature and wavelength (redshift!) 4 Slightly better at moderate redshift 2 0 10-6 10-5 10-4 10-3 10-2 10-1 100 Fraction of hot dust Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 6 / 9 Beelen et al. (2006): 6 quasars, from z=1.8–6.5 Using only single wavelengths, results more similar for shorter wavelengths Typically only ∼ 30% different for λ . 150µm up to 70% for longer wavelengths! Comparison to published dust masses Try computing masses from literature fluxes - Ignoring possible multiple T components! Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 7 / 9 Using only single wavelengths, results more similar for shorter wavelengths Typically only ∼ 30% different for λ . 150µm up to 70% for longer wavelengths! Comparison to published dust masses Try computing masses from literature fluxes - Ignoring possible multiple T components! Beelen et al. (2006): 6 quasars, from z=1.8–6.5 Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 7 / 9 Typically only ∼ 30% different for λ . 150µm up to 70% for longer wavelengths! Comparison to published dust masses Try computing masses from literature fluxes - Ignoring possible multiple T components! Beelen et al. (2006): 6 quasars, from z=1.8–6.5 Using only single wavelengths, results more similar for shorter wavelengths Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 7 / 9 Comparison to published dust masses Try computing masses from literature fluxes - Ignoring possible multiple T components! Beelen et al. (2006): 6 quasars, from z=1.8–6.5 Using only single wavelengths, results more similar for shorter wavelengths Typically only ∼ 30% different for λ . 150µm up to 70% for longer wavelengths! Peter Scicluna JCMT Users’ Meeting, Nanjing, 13th February 2017 7 / 9 Next steps Different/mixed compositions 5 10 Redshift evolution 104 /g) Build model galaxies