Molecular and atomic line surveys of galaxies II: unbiased estimates of their star formation mode

We make use of our 'minimal' cold interstellar medium (ISM) emission line model that predicts the molecular and atomic line emission per unit dense, star-forming gas mass (Geach & Papadopoulos 2012; Paper I) to examine the utility of key line ratios in surveys of the so-called star formation 'mode' as traced by xi_SF = M_dense(H_2)/M_total(H_2). We argue that xi_SF and its proxies provide very sensitive, extinction-free discriminators of rapid starburst/merger-driven versus secular quiescent/disk-like stellar mass assembly, with the most promising diagnostic to be applied in the near-future being CO(4-3)/[CI](1-0). These lines are accessible across nearly the full range 0<z<2 (thus covering the bulk of galaxy evolution) with the Atacama Large Millimeter Array. In addition to their diagnostic power, another advantage of this combination is the similar observed frequencies (Delta nu_0 ~ 30 GHz) of the lines, resulting in nearly spatially-matched beams for a fixed aperture, thus mitigating the effects of resolution/morphology bias in the interpretation of galaxy-averaged line ratios. Finally we discuss the capability of deep blind redshift surveys with the high frequency component of the Square Kilometer Arrray (SKA) in discovering H_2-rich galaxies with very low xi_SF values. These could be the progenitors of starburst galaxies seen prior to the onset of star formation; such galaxies could be a class of extreme outliers from local (gas surface density)-(star formation rate) scaling laws, which would exclude them from current star formatation or stellar mass selected samples. Our conservative model suggests that SKA could detect such systems residing at z~3 at a rate of 20-200 per hour.