Potential Vorticity as a Measure of Atmospheric Physical Consistency in Bias Corrected Fields

There is a growing use of regional climate model (RCM) lateral boundary condition bias correction as a means to reduce the impact of global climate model biases in regional models. The aim of these techniques is to improve climate simulations by attempting to impart more appropriate climatological fields into the regional model's boundary conditions. Current approaches perform bias correction on each input variable independently, with little consideration for variable correlations or the atmosphere's physical consistency. In this paper we present potential vorticity (PV) which we use to measure the creation of spurious anomalies in the mesoscale fields resulting from physically inconsistent bias correction. This study corrects global CSIRO Mk3.5 GCM simulations of zonal and meridional winds, geopotential height, and potential temperature of the current climate to ECMWF ERA-Interim reanalysis, on fifteen isentropic levels. We undertake two different split-sampled bias correction methods - a correction of the mean, and a correction of the mean and standard deviation - on each of the input variables. The calculated PV of the corrected fields is compared with a control bias corrected PV field. We show that applying bias correction parameters to each variable independently results in spurious PV anomalies throughout the atmospheric column, with particular irregularities centred on Rossby waves in the Tropopause. We speculate as to the effect of these inconsistencies if they were to be inputs into the RCM's lateral boundary conditions on the model simulation.