Calibration of star formation rate tracers for short- and long-lived star formation episodes

Context. To derive the history of star formation in the Universe a set of calibrated star formation rate tracers at different wavelengths is required. The calibration has to consistently take into account the effects of extinction, star formation regime (short or long-lived) and the evolutionary state to avoid biases at different redshift ranges. Aims. We use evolutionary synthesis models optimized for intense episodes of star formation to compute a consistent calibration of the most usual star formation rate tracers at different energy ranges, from X-ray to radio luminosities. Methods. We have computed the predicted evolution of the different estimators taking into account nearly-instantaneous and continuous star formation regimes and the effect of interstellar extinction (attenuation at high energies, thermal reradiation in the far infrared). We have also considered the effect of metallicity on the calibration of the different estimators. Results. A consistent calibration of a complete set of star formation rate tracers is presented, computed for the most usual star-forming regions conditions in terms of evolutionary state, star formation regime, interstellar extinction and initial mass function. We discuss the validity of the different tracers in different star formation scenarios and compare our predictions with previous calibrations of general use. Conclusions. In order to measure the intensity of star formation episodes we should distinguish between nearly-instantaneous and continuous star formation regimes. While the star formation strength (M(circle dot)) should be used for the former, the more common star formation rate (M(circle dot) yr(-1)) is only valid for episodes forming stars at a constant rate during extended periods of time. Moreover, even for extended star formation episodes, the evolutionary state should be taken into account, since most SFR tracers stabilize only after 100 Myr of evolution.