Southern Ocean Monthly Climatology

Outline

This is the Southern Ocean Monthly Climatology of Yamazaki et al. "Unlocking Southern Ocean Under-ice Seasonality with a New Monthly Climatology". The interpolation method follows Barth et al. (2014) available via DIVAnd Julia package (https://github.com/gher-uliege/DIVAnd.jl). CTD data sourced from Argo, MEOP, and World Ocean Database (including low resolution ocean station data).

The dataset covers south of 40S and above 2000 dbar (above 1000 dbar for "_minimal"). The horizontal grid is 1/4 and 1/2 degs in latitude and longitude, and the vertical grid is the 66 WOA layers. Mixed layer depth, temperature, salinity, crudely derived from max("Δσθ_10m=0.03kg/m3", "Holte&Talley"), are also provided in "_MLD". 

The following variables are included (* are excluded in "_minimal"):

In-situ temperature (°C) in ITS-90

Practical salinity (psu)

*Standard deviation of temperature (°C), inferred by the spread of observations

*Standard deviation of practical salinity (psu), inferred by the spread of observations

*Interpolation error of temperature (°C), inferred by the sparsity of observations

*Interpolation error of practical salinity (psu), inferred by the sparsity of observations

*Cabbeling correction for temperature (°C)

*Cabbeling correction for practical salinity (psu)

*Density stabilization factor for temperature (°C)

*Density stabilization factor for practical salinity (psu)

Project Description

The advent of under-ice profiling float and biologging techniques has enabled year-round observation of the Southern Ocean and its Antarctic margin. These under-ice data are often overlooked in widely used oceanographic datasets, despite their importance in understanding seasonality and its role in sea ice changes, water mass formation, and glacial melt. We develop a monthly climatology of the Southern using Data Interpolating Variational Analysis, which excels in multi-dimensional interpolation and consistent handling of topography and horizontal advection. The dataset will be instrumental in investigating the seasonality and improving ocean models, thereby making valuable under-ice observations more accessible.

Methodology

A diverse set of observational data was collected from multiple platforms, including Argo profiling floats, biologging instruments attached to marine mammals, and Conductivity-Temperature-Depth (CTD) profiles obtained from ship-based measurements. This dataset spans more than a century, with a heavier reliance on recent data collected after 2005 by Argo float and biologging. The gridding covers the Southern Ocean from 75°S to 40°S with horizontal grid intervals of 1/2° longitude and 1/4° latitude, resulting in an approximate 25 km resolution at 60°S. The vertical grid consists of 66 WOA layers, ranging from 5 to 2,000 meters in depth, with monthly temporal resolution.

The initial climatology field was constructed using Data Interpolating Variational Analysis (DIVA), a sophisticated method designed to handle multidimensional data interpolation. This method efficiently integrates residual data from an existing climatology (WOA23) and applies topographic and advection constraints to improve the spatial correlation of the data, particularly in regions with complex topography, such as the Antarctic margin. 

The DIVA model underwent a cross-validation process to optimize interpolation parameters, including spatial and temporal correlation scales and the signal-to-noise ratio. Additionally, calibration tests of advection weighting and correlation time scales was carried out to further refine the model's physical consistency. Post-processing involved generating error maps to assess the accuracy of the interpolated fields. Corrections for non-linearities in seawater properties (cabbeling) were applied, along with density stabilization techniques to ensure that the final data output remained physically consistent. 

How to cite

Yamazaki, K., Bindoff, N. L., Phillips, H. E., Nikurashin, M., Herraiz-Borreguero, L., & Spence, P. (2025). Unlocking southern ocean under-ice seasonality with a new monthly climatology. Journal of Geophysical Research: Oceans, 130, e2024JC020920. https://doi.org/10.1029/2024JC020920