Thursday

Radio emissions trace star-formation rates within galaxies using the well established Radio-FIR correlation and obviates the problem of dust extinction that plague some other wavebands. Hence radio imaging of star forming galaxies can provide a tracer of global star-formation rates if emissions can be disentangled from the contributions of Active Galactic Nuclei (AGN). We present a probabilistic categorisation of these two broad emission classes of sources, ubiquitously observed in deep fields. Whereas traditional classification methods have relied almost exclusively upon human judgement of individual objects, leading to a simple binary designation (or perhaps none at all), this method generates a classification probability based on metrics derived from high angular resolution measurements of sources. Using recent data from e-MERLIN's Galaxy Evolution (eMERGE) survey which probes ~800 star-forming galaxies and AGN to z~5, classification probabilities are derived revealing disentangled differential source counts. The method can be extended to include deeper multi-wavelength data to further decrease uncertainties at even earlier epochs, revealing global star formation rates as a function of cosmic time.