Can the CMIP5 models represent winter frontal precipitation?

Abstract

Much of the day-to-day variability of rainfall in the midlatitudes is controlled by the passage of extratropical cyclones and their related fronts. A good representation of fronts and their associated rainfall in climate models is essential to have confidence in future projections of midlatitude precipitation. An objective front identification method has been applied to the data from ERA-Interim and 18 Coupled Model Intercomparison Project, version 5 (CMIP5) models and the fronts linked with daily precipitation estimates to investigate how winter front-related precipitation is represented in the models. While the front frequency is well represented, the frequency of frontal precipitation is too high and the intensity is too low, thus adding little bias to the rainfall total. Although the intensity of the modeled frontal precipitation is too low, the intensity of other precipitation is even lower; thus, the ratio of frontal precipitation to total precipitation is higher in the models than in the reanalysis. Key Points Evaluation of the winter frontal precipitation in CMIP5 models using objective front identification Frontal precipitation well represented on average but due to complex compensating factors Amplifying impact of fronts on rainfall intensity too high in models because of low biased total