We study the radio–far infrared ( FIR ) correlation in “ blue cloud ” galaxies chosen from the PRism MUltiobject Survey ( PRIMUS ) up to redshift ( z ) of 1.2 in the XMM-LSS field .
We use rest-frame emission at 1.4 GHz in the radio and both monochromatic ( at 70 \mu m ) and bolometric ( between 8 - 1000 ~ { } \mu m ) emission in the FIR .
To probe the nature of the correlation up to z \sim 1.2 , where direct detection of blue star-forming galaxies is impossible with current technology , we employ the technique of image stacking at 0.325 and 1.4 GHz in the radio and in six infrared bands , viz .
24 , 70 , 160 , 250 , 350 and 500 ~ { } \mu m. For comparison , we also study the correlation for more luminous galaxies that are directly detected .
The stacking analysis allows us to probe the radio–FIR correlation for galaxies that are up to 2 orders of magnitude fainter than the ones detected directly .
The k - correction in the infrared wavebands is obtained by fitting the observed spectral energy distribution ( SED ) with a composite mid-IR power law and a single temperature greybody model .
We find that the radio luminosity at 1.4 GHz ( L _ { 1.4 GHz } ) is strongly correlated with monochromatic FIR luminosity at 70 \mu m ( L _ { 70 \mu m } ) having slope 1.09 \pm 0.05 and with bolometric luminosity ( L _ { TIR } ) having slope 1.11 \pm 0.04 .
The quantity q _ { TIR } ( = \log _ { 10 } [ L _ { TIR } / ( 3.75 \times 10 ^ { 12 } L _ { 1.4 GHz } ) ] ) is observed to decrease with redshift as q _ { TIR } \propto ( 1 + z ) ^ { -0.16 \pm 0.03 } probably caused due to the non-linear slope of the radio–FIR correlation .
Within the uncertainties of our measurement and the limitations of our flux-limited and color-selected sample , we do not find any evolution of the radio–FIR correlation with redshift .