Using Argo Measurements to understand model bias (2011)

Andreas Sterl (1)*   

(1) : Royal Netherlands Meteorological Institue (KNMI), De Bilt, Netherlands

* Corresponding author : Andreas Sterl

Argo data are used to understand the reasons behind a climate model's failure to reproduce observed SSTs in the Southern Ocean. EC-Earth is a newly developed climate model (Hazeleger et al. 2011; Sterl et al. 2011). It is based on the NEMO ocean model (Madec 2008) and ECMWF’s atmospheric forecast model. While the model represents many aspects of the climate very well (Hazeleger et al. 2011; Sterl et al. 2011), it has a SST bias reaching up to 5 K in the Southern Ocean (Figure 1).

One of the potential reasons for this bias are problems with mixed layer (ML) dynamics in the model. Based on four years of Argo data Dong et al. (2008) produced a climatology of mixed layer depths (MLD). Comparing their climatology with modeled MLD shows that the latter are much too shallow in the Southern Ocean. Typical differences are 50 m during austral summer and 200-300 m in winter (Sterl et al. 2011). In both cases this amounts to a factor of two in MLD!


Figure 1 : Annual-mean SST difference between a long EC-Earth control run and the WOA09 climatology (from Sterl et al., 2011).

To better understand the processes behind this lack of deep mixed layers, we compare individual Argo profiles with model results at or near the same position. Figure 2 shows one such comparison. The modelled profiles of T and S in year 1 (red) and year 20 (green) of a run starting from observed temperature and salinity (World Ocean Atlas 2009 (WOA09): Antonov et al. 2010; Locarnini et al. 2010) are compared with the three profiles measured by an Argo float in the respective months. From the figure the observed MLD is approximately 350 m in winter (lower row) and 100 m in summer (upper row). Below the ML model and observations agree very well, and the modeled values hardly change during the 20 years of integration. Large changes, however, occur in the ML. Right from the beginning of the integration (red curves) the ML is two shallow. Even in winter (lower row) it hardly reaches 100 m. The observed summer profiles show a characteristic sub-surface temperature minimum (upper-left panel). It is the remnant of the deep winter ML that is capped by warm water during summer. As the model does not produce a deep winter ML, sub-surface temperature increases and the summer minimum erodes away.


Figure 2 : Comparison of modeled and observed profiles of temperature (left) and salinity (right) in the Southern Ocean for january (upper) and July (lower). The thick red and green lines are the modeled profiles in year 1 and in year 20, respectively, and the thin black lines are the measured profiles of Argo float 6900733 in January and July 2010 (dates and positions plotted in legend).

Clearly the model is not capable of producing deep enough mixed layers in the Southern Ocean, and as a result the characteristic temperature profile with a summer sub-surface temperature minimum cannot be sustained. Work is underway to identify ML-deepening processes that are missing or under- represented in the model. Carefully comparing modeled with Argo profiles at different positions allows to identify the impact of different processes and their interplay.

  • Antonov JI, Seidov D, Boyer TP, Locarnini RA, Mishonov AV, Garcia HE, Baranova OK, Zweng MM, Johnson DR : World Ocean Atlas 2009, Volume 2: Salinity. S. Levitus, Ed., NOAA Atlas NESDIS 69, U.S. Government Printing Office, Washington, D.C., 184 pp, 2010.
  • Dong S, Sprintall J, Gille ST, and Talley L : Southern ocean mixed-layer depth from Argo float profiles, J. Geophys. Res., 113, C06013, doi:10.1029/2006JC004051, 2008.
  • Hazeleger W, Wang X, Severijns C, S ̧tefˇanescu S, Bintanja R, Sterl A, Wyser K, Semmler T, Yang S, Van den Hurk B, Van Noije T, Van der Linden E, Van den Wiel K : EC-Earth V2 : description and validation of a new seamless Earth system prediction model. Clim Dyn, online first, doi: 10.1007/s00382-011-1228- 5, 2011.
  • Locarnini RA, Mishonov AV, Antonov JI, Boyer TP, Garcia HE, Baranova OK, Zweng MM, Johnson DR : World Ocean Atlas 2009, Volume 1: Temperature. S. Levitus, Ed. NOAA Atlas NESDIS 68, U.S. Government Printing Office, Washington, D.C., 184 pp, 2010.
  • Madec G. : NEMO ocean engine, Note du Pole de modélisation, Institut Pierre- Simon Laplace (IPSL), France, No 27 ISSN No 1288-1619, 2008.