This study investigates how pycnocline smoothing and subgrid-scale variability of density profiles influence the determination of the mixed layer depth (MLD) in the global ocean, and applies the results of analysis to assess the ability of ocean general circulation models (OGCM) to simulate the MLD. For this purpose, individual, monthly mean, and climatological profiles are analyzed over a horizontal resolution of 1° × 1° for both observation data (Argo) and eddy-resolving OGCM (OFES) results. It is found that the MLDs from averaged profiles are generally smaller than those from individual profiles because of pycnocline smoothing induced by the averaging process. A correlation is found between the decrease in MLD Δh and the increase in pycnocline thickness Δδ of averaged profiles, except during winter in the high-latitude ocean. The relation is estimated as Δh = -αΔδ - β, where a Δ 0.7 in all cases, but β increases with the subgrid-scale variability of density profiles. A correlation is also found between Δh and the standard deviation of the MLD within a grid. The results are applied to estimate how much of the MLD bias of OFES is due to prediction error and how much is due to profile error, induced by different pycnocline smoothing and the subgrid-scale variability of density profiles. The study also shows how profile error varies with the threshold density difference criterion.
Bibliographical noteFunding Information:
Acknowledgments. This work was supported by the National Research Foundation of Korea Grant funded by the Korean Government (MEST) (NRF-2009-C1AAA001-0093068). The Argo data were collected and made freely available by the International Argo Program and the national programs that contribute to it (http://www.argo.ucsd.edu; http://argo.jcommops.org). The Argo Program is part of the Global Ocean Observation System (doi:10.17882/42182). The OFES simulation was conducted on the Earth Simulator under the support of JAMSTEC.
All Science Journal Classification (ASJC) codes
- Ocean Engineering
- Atmospheric Science