Title: Photometric Redshift Realism: A Technique for Reducing Catastrophic Outlier Redshift Estimates in Large-Scale Surveys

Abstract: We present results of using individual galaxies' effective redshift probability density information as a method of identifying potential catastrophic outliers in empirical photometric redshift estimation. In the course we develop a method of modification of the redshift distribution of training sets to improve both the baseline accuracy of high redshift (z>1.5) estimation as well as catastrophic outlier mitigation. We demonstrate these using two real test data sets and one simulated test data set spanning a wide redshift range (0<z<4). We present these results in the context of "photometric redshift realism" and aim to show that the methods and results presented here can inform a 'prescription' that can be applied as a realistic photometric redshift estimation scenario for a hypothetical large-scale survey. We find that with appropriate optimization, we can identify a large percentage (>30%) of catastrophic outlier galaxies while simultaneously incorrectly flagging only a small percentage (<7% and in many cases <3%) of non-outlier galaxies as catastrophic outliers. We find also that our training set redshift distribution modification results in a significant decrease (>10%) in the percentage of outlier galaxies greater than z=1.5 with only a small increase (<3%) in the percentage of outlier galaxies less than z=1.5 compared to the unmodified training set. In addition, we find that this modification can in some cases decrease the percentage of incorrectly identified non-outlier galaxies by almost 20%, while in other cases cause only a small (<1%) increase in this metric.