Many problems in time domain astronomy are best approached with continuous, high frequency (<1 minute) observations over a large area of the sky, because event rates are generally less than 0.01 detections per sq. degree in a year. For GRBs and the earliest stages of SNe, follow-up spectroscopy, multiband photometry and polarization measurements need to be triggered in seconds or minutes. Plateau sites in Antarctica are optimal for a new survey, Xian (eXtreme Imaging Antarctic Network), for which the scientific motivation is presented.