Measuring Star Formation Rates in Dwarf Galaxies A

Measuring Star Formation Rates in Dwarf Galaxies “A Comparison of H and UV Star Formation Rates in the Local Volume: Systematic Discrepancies in Dwarf Galaxies” Lee, Gil de Paz, Tremonti, Kennicutt, Salim, Bothwell, Calzetti, Dalcanton, Dale, Engelbracht, Funes, Johnson, Sakai, Skillman, van. Zee, Walter, Weisz, 2009, Ap. J, in press Janice C. Lee Carnegie Observatories Galaxies 09, NRAO September 21, 2009

? SFR(UV) SFR(H ) = Janice C. Lee Carnegie Observatories Galaxies 09, NRAO September 21, 2009

SFR(H ) SFR(UV ) c 11 Mp

a galaxy SED… UV [OII] H PAHs MIR FIR

UV H

UV Photospheric emission from O & B stars; traces SF over last ~100 Myr blue- FUV; yellow- NUV Source: GALEX NGS H Recombination emission of gas ionized by O stars; traces SF over last ~5 Myr blue- R; red- H-alpha Source: 11 HUGS

H and UV SFR Comparisons: Previous Work Buat+02 filled points: Buat SF cluster gals SCAP, FOCA, FAUST open points: Bell & Kennicutt 00, UIT crosses: Sullivan+00 FOCA, UV-selected Salim+07 “UV” SFR: SDSS & GALEX broadband SED modeling “H ” SFR: Brinchmann+04 SDSS spectral modeling

H and UV SFR Comparisons: Previous Work Tentative evidence for generally bursty Small, non-statistical (biased? ) SFHs in dwarfs galaxies (e. g. , samples Sullivan+00; Bell & Kennicutt 01)? Variable IMF (e. g. Meurer+09)? Buat+02 filled points: Buat SF cluster gals SCAP, FOCA, FAUST open points: Bell & Kennicutt 00, UIT crosses: Sullivan+00 FOCA, UV-selected

The 11 Mpc H Imaging Survey GALEX�� The 11 Mpc H & UV Galaxy Survey (11 HUGS) The Local Volume Legacy Survey (LVL) Spitzer�� c 11 Mp

A Census of Star Formation and Dust in Local Galaxies… 11 Mpc H Imaging Survey Lee+07, Kennicutt+08, Lee+09 a R. Kennicutt (Cambridge), J. C. Lee (Carnegie), J. Funes (Vatican Obs. ), S. Sakai (UCLA), S. Akiyama (U. AZ) 11 Mpc H and Ultraviolet Galaxy Survey (11 HUGS) Lee+09 b, Lee & Gilde. Paz+09 (in prep. ) J. C. Lee (Carnegie), Kennicutt (Cambridge), A. UV, Gil de. H , Paz. IR (Madrid), LVL data. R. deliveries (Ngal=258) of data C. Tremonti (U. AZ), to. S. IRSA Sakai(http: //irsa. ipac. caltech. edu) (UCLA), J. Funes (Vatican Obs. ), L. van. Zee (U. April Indiana) began 2008 with anticipated completion by end of 2009. The Spitzer Local Volume Legacy Survey (LVL) Dale+09 J. C. Lee (Carnegie), R. Kennicutt (Cambridge), C. Engelbracht (U. AZ), D. Calzetti (U. Mass. ), D. Dale (U. Wyoming), A. Gil de Paz (Madrid), J. Dalcanton (U. Wash. ), E. Skillman (U. Minn. ), J. Funes (Vatican Obs. ), K. Gordon (STSci), B. Johnson (Cambridge), S. Sakai (UCLA), L. van. Zee (U. Indiana), F. Walter (MPIA), D. Weisz (U. Minn. ), B. Williams (U. Wash. ), Y. Wu (Caltech), M. Block (U. AZ)

80% of sample less luminous, active than ~LMC Image credit: S. Staudaher (U. Wyoming)

-6 -4 -2 0 Results: H and UV SFR Comparisons from 11 HUGS/LVL -4 Buat+02 filled points: Buat SF cluster gals SCAP, FOCA, FAUST open points: Bell & Kennicutt 00, UIT crosses: Sullivan+00 FOCA, UV-selected 11 HUGS/LVL -2 0

Results: H and UV SFR Comparisons from 11 HUGS/LVL <SFR(H )/SFR(UV)> systematically decreases below unity with decreasing SFR, M(B). Check assumptions underlying SFR/L=C (e. g. Kennicutt 98): - universal IMF (e. g. , Salpeter, Kroupa) - fully populated IMF - stellar evolution tracks and model atmospheres (Z=Z ) - continuous star formation over tequi - case B recombination (for H ); no leakage of ionizing photons observed relation, prior to corrections for internal dust attenuation - L represents intrinstic, dust corrected values

Impact of dust? observed relation, after corrections for internal dust attenuation

Discontinuous (Bursty) Star Formation? Iglesias-Paramo+04 (also Sullivan+00, Meurer+09)

Discontinuous (Bursty) Star Formation? Burst amplitudes of 10 -100 required to produce factor of 2 offsets from expected value. But required amplitudes too large: From Lee+09, starbust number fraction is ~6%, SF in dwarfs is ~23%, so average amplitude of burst is ~4. SFH reconstructed from resolved stellar pops show amplitudes < 5 (e. g. , Weisz+08) Iglesias-Paramo et al. (2004) see also Sullivan et al. (2000)

IMF: Not Fully Populated? <SFR(H )/SFR(UV)> systematically decreases below unity with decreasing SFR, M(B). Check assumptions underlying SFR/L=C (e. g. Kennicutt 98): - universal IMF (e. g. , Salpeter, Kroupa) - fully populated IMF - stellar evolution tracks and model atmospheres (Z=Z ) - continuous star formation over tequi - case B recombination (for H ); no leakage of ionizing photons - L represents intrinstic, dust corrected values

IMF: Not Fully Populated? (m)=m-2. 35 back of the envelope calculation: stochasticity should begin to affect H /UV for SFR<0. 001.

IMF: Not Fully Populated? (m)=m-2. 35 monte carlo simulations from Tremonti+ in prep.

IMF: Variations? <SFR(H )/SFR(UV)> systematically decreases below unity with decreasing SFR, M(B). Check assumptions underlying SFR/L=C (e. g. Kennicutt 98): - universal IMF (e. g. , Salpeter, Kroupa) - fully populated IMF - stellar evolution tracks and model atmospheres (Z=Z ) - continuous star formation over tequi - case B recombination (for H ); no leakage of ionizing photons - L represents intrinstic, dust corrected values

IMF: Variations? “Integrated Galaxial IMF” cluster=m-2. 35 ma x(m _st ar) = m_ clu ste r Weidner&Kroupa 06 cluster= m-2 for 50<m[Msun]<mmax m-1 for 5< m[Msun] <50

IMF: Variations? cluster=m-2. 35 cluster= m-2 for 50<m[Msun]<mmax m-1 for 5< m[Msun] <50

Closing Thoughts Consistency between data and IGIMF <L(H )/L(UV)> systematically predictions intriguing (note other decreases with decreasing SFR, M(B). recent claims in lit), but much more work is needed. - universal IMF (e. g. , Salpeter, Kroupa) top-lite? - Some factors difficult to rule out - Combination of effects? - fully populated IMF - Unexplored causes? - stellar evolution tracks and model - Investigate differences in H and atmospheres (Z=Z )? UV spatial distributions. - continuous star formation over tequi Empirically with current data: SFR C L(H ) assuming UV is robust tracer. See paper for more details on empirical recalibration of L(H ). - case B recombination (for H ); no leakage of ionizing photons? - L represents intrinstic, dust corrected values