Verification of the mixed layer depth in the OceanMAPS operational forecast model for Austral autumn

The ocean mixed layer depth is an important parameter describing the exchange of fluxes between the atmosphere and ocean. In ocean modelling a key factor in the accurate representation of the mixed layer is the parameterization of vertical mixing. An ideal opportunity to investigate the impact of di...

Full description

Bibliographic Details
Published in:Geoscientific Model Development
Main Authors: D. Boettger, R. Robertson, G. B. Brassington
Format: Article in Journal/Newspaper
Language:English
Published: Copernicus Publications 2018
Subjects:
Online Access:https://doi.org/10.5194/gmd-11-3795-2018
https://doaj.org/article/1a0da8bfadbf4a9f8ca6df12d776d57c
Description
Summary:The ocean mixed layer depth is an important parameter describing the exchange of fluxes between the atmosphere and ocean. In ocean modelling a key factor in the accurate representation of the mixed layer is the parameterization of vertical mixing. An ideal opportunity to investigate the impact of different mixing schemes was provided when the Australian Bureau of Meteorology upgraded its operational ocean forecasting model, OceanMAPS to version 3.0. In terms of the mixed layer, the main difference between the old and new model versions was a change of vertical mixing scheme from that of Chen et al. (1994) to the General Ocean Turbulence Model. The model estimates of the mixed layer depth were compared with those derived from Argo observations. Both versions of the model exhibited a deep bias in tropical latitudes and a shallow bias in the Southern Ocean, consistent with previous studies. The bias, however, was greatly reduced in version 3.0, and variance between model runs decreased. Additionally, model skill against climatology also improved significantly. Further analysis discounted changes to model resolution outside of the Australian region having a significant impact on these results, leaving the change in vertical mixing scheme as the main factor in the assessed improvements to mixed layer depth representation.