Photometric redshift (photo-z) estimates are playing an increasinglyimportant role in extragalactic astronomy and cosmology. Crucial to manyphoto-z applications is the accurate quantification of photometric redshifterrors and their distributions, including identification of likely catastrophicfailures in photo-z estimates. We consider several methods of estimatingphoto-z errors and propose new training-set based error estimators based onspectroscopic training set data. Using data from the Sloan Digital Sky Surveyand simulations of the Dark Energy Survey as examples, we show that this methodprovides a robust, relatively unbiased estimate of photo-z errors. We show thatculling objects with large, accurately estimated photo-z errors from a samplecan reduce the incidence of catastrophic photo-z failures.
展开▼
