Title: The need for multicomponent dust attenuation in modeling nebular emission: Constraints from SDSS-IV MaNGA

Abstract: A fundamental assumption adopted in nearly every extragalactic emission-line study is that the attenuation of different emission lines can be described by a single attenuation curve. Here we show this assumption fails in many cases with important implications for derived results. We developed a new method to measure the differential nebular attenuation among three kinds of transitions: the Balmer lines of hydrogen, high-ionization transitions, and low-ionization transitions. This method bins the observed data in a multidimensional space spanned by attenuation-insensitive line ratios. Within each small bin, the variations in line ratios are mainly driven by the variations in the nebular attenuation. This allows us to measure the nebular attenuation using both forbidden lines and Balmer lines. We applied this method to a sample of 2.4 million star-forming spaxels from SDSS-IV MaNGA. We found that the attenuation of high ionization lines and Balmer lines can be well described by a single Fitzpatrick (1999) extinction curve with $R_V=3.1$. However, no single attenuation curve can simultaneously account for all three transitions. This strongly suggests that different lines have different effective attenuations, likely because spectroscopy at kiloparsec resolutions mixes multiple regions with different intrinsic line ratios and different levels of attenuation. As a result, the assumption that different lines follow the same attenuation curve breaks down. Using a single attenuation curve determined by Balmer lines to correct attenuation-sensitive forbidden line ratios could bias the nebular parameters derived by 0.06--0.25 dex at $A_V = 1$, depending on the details of the dust attenuation model. Observations of a statistically large sample of H II regions with high spatial resolutions and large spectral coverage are vital for improved modeling and deriving accurate corrections for this effect.