Assessing bias corrections of oceanic surface conditions for atmospheric models

Abstract

Future sea surface temperature and sea-ice concentration from coupled ocean-atmosphere general circulation models such as those from the CMIP5 experiment are often used as boundary forcings for the downscaling of future climate experiments. Yet, these models show some considerable biases when compared to the observations over present climate. In this paper, existing methods such as an absolute anomaly method and a quantile-quantile method for sea surface temperature (SST) as well as a look-up table and a relative anomaly method for sea-ice concentration (SIC) are presented. For SIC, we also propose a new analogue method. Each method is objectively evaluated with a perfect model test using CMIP5 model experiments and some real-case applications using observations. We find that with respect to other previously existing methods, the analogue method is a substantial improvement for the bias correction of future SIC. Consistency between the constructed SST and SIC fields is an important constraint to consider, as is consistency between the prescribed sea-ice concentration and thickness; we show that the latter can be ensured by using a simple parameterisation of sea-ice thickness as a function of instantaneous and annual minimum SIC.

Effect of prescribed sea surface conditions on the modern and future Antarctic surface climate simulated by the ARPEGE atmosphere general circulation model

“…Code and data availability. Climatological monthly means for each ARPEGE simulation presented in this study are available using https://doi.org/10.17605/OSF.IO/FGV64 (Beaumet et al, 2019b). Python scripts developed to produce the figures presented in this study can also be downloaded (https://doi.org/10.17605/OSF.IO/Q4YCT, Beaumet et al, 2019c).…”

“…Sea surface conditions have been identified as key drivers for the evolution of the climate of the Antarctic continent (Krinner et al, 2014;Agosta et al, 2015). In this study, SSCs obtained from CMIP5 projections are bias-corrected using recommendations from Beaumet et al (2019a) before being used as surface boundary conditions for the atmospheric model. The bias-correction methods used for SST and SIC mostly rely on anomaly methods, which consist in adding the anomalies coming from a coupled model projection to the observed SSCs while taking into account the quantile distribution of these anomalies.…”

“…Over the ocean, we use a 1-D version of sea ice model GELATO (Mélia, 2002), which means that no advection of sea ice is possible. The sea ice thickness is prescribed following the empirical parameterisation used in Krinner et al (1997Krinner et al ( , 2010 and described in Beaumet et al (2019a). The use of GELATO is therefore limited to the computation of heat and moisture fluxes in sea-ice-covered regions and also allows us to take into account the accumulation of snow on top of sea ice.…”

“…In this work, a bias correction of SSCs using a quantile mapping method for SST and an analogue method for SIC is achieved following the recommendations described in Beaumet et al (2019a). We drive the ARPEGE AGCM (Déqué et al, 1994) with both observed and simulated (from coupled models) SSCs for the recent past .…”

Responses to Reviewer #2

“…High-resolution polar-oriented RCMs provide more reliable estimates of the Antarctic SMB components, but they depend on their forcing boundary conditions, including SSCs. Using adequate SSCs in climate models could be as crucial as using a suitable downscaling model (Krinner et al, 2008;Beaumet et al, 2017). This is of particular importance since most general circulation models (GCMs) from the 5th phase of the Coupled Model Intercomparison Project (CMIP5; Taylor et al, 2012) have failed to reproduce the SSC temporal and spatial variability in the Southern Ocean area over the last decades (Mahlstein et al, 2013;Turner et al, 2013;Shu et al, 2015;Agosta et al, 2015;Roach et al, 2018).…”