# The infrared-radio correlation of star-forming galaxies is strongly M$_{\star}$-dependent but nearly redshift-invariant since z$\sim$4

**Authors:** I. Delvecchio, E. Daddi, M. T. Sargent, M. J. Jarvis, D. Elbaz, S. Jin, D. Liu, I. H. Whittam, H. Algera, R. Carraro, C. D'Eugenio, J. Delhaize, B. Kalita, S. Leslie, D. Cs. Molnar, M. Novak, I. Prandoni, V. Smolcic, Y. Ao, M. Aravena  
**arXiv:** [2010.05510](https://arxiv.org/abs/2010.05510)  
**Date:** 2020-10-12  
**Summary page:** https://www.summarizepaper.com/en/paper/2010.05510v1/

## Key points

- The study focuses on the infrared-radio correlation (IRRC) in star-forming galaxies
- IRRC is used as a calibration tool for estimating star formation rates (SFR) based on radio emission
- The authors re-calibrate the IRRC by considering both stellar mass and redshift
- They analyze a sample of over 400,000 star-forming galaxies in the COSMOS field
- Stacked deep infrared/sub-mm and radio images are used within each bin defined by stellar mass and redshift
- The results show that the IRRC primarily depends on stellar mass, with more massive galaxies exhibiting lower q$_{TIR}$ values (the logarithm of the IRRC)
- There is also a weaker secondary dependence on redshift
- An analytical expression is provided to describe this relationship: q$_{TIR}$(M$_{*}$,z)=(2.646±0.024)×(1+z)^(-0.023±0.008)-(0.148±0.013)×(log M_{*}/M_{\odot}-10)
- When accounting for the UV dust-uncorrected contribution to total SFR, there is a further steepening of the q$_{TIR}$ dependence on stellar mass
- These findings highlight the need for novel M$_{*}$ dependent recipes when using radio emission as a proxy for SFR

## Summary

This study focuses on the infrared-radio correlation (IRRC) in star-forming galaxies, which is the ratio between total infrared and radio luminosity. The IRRC has been used as a calibration tool for estimating star formation rates (SFR) based on radio emission. Previous research has shown that the IRRC evolves with redshift, but the underlying reasons for this evolution are not well understood. In this study, the authors re-calibrate the IRRC by considering both stellar mass (M$_{*}$) and redshift. They analyze a sample of over 400,000 star-forming galaxies in the COSMOS field, selected based on their (NUV-r)/(r-J) colors and spanning redshifts from 0.1 to 4.5. Within each bin defined by M$_{*}$ and redshift, they stack deep infrared/sub-mm and radio images. The stacked infrared spectral energy distributions are fitted using templates for typical star-forming galaxies and infrared AGN sources. To ensure accuracy, they employ a recursive approach to remove potential radio AGN candidates. The results show that the IRRC primarily depends on stellar mass, with more massive galaxies exhibiting lower q$_{TIR}$ values (the logarithm of the IRRC). There is also a weaker secondary dependence on redshift; an analytical expression is provided to describe this relationship: q$_{TIR}$(M$_{*}$,z)=(2.646±0.024)×(1+z)^(-0.023±0.008)-(0.148±0.013)×(log M_{*}/M_{\odot}-10). Additionally, when accounting for the UV dust-uncorrected contribution to the total SFR, there is a further steepening of the q$_{TIR}$ dependence on stellar mass; this decrease in q$_{TIR}$ for more massive galaxies may be linked to higher star formation rate surface density which leads to larger cosmic ray scale heights. These findings highlight the need for novel M$_{*}$ dependent recipes when using radio emission as a proxy for SFR; such recipes will be crucial for accurately measuring SFR in future ultra deep radio surveys, particularly for low SFR and low M${_*} $galaxies.