Investigating Predictability of DIC and SST in the Argentine Basin Through Wind Stress Perturbation Experiments. Issue 21 (9th November 2021)

• Record Type: Journal Article
• Authors: Swierczek, Stan; Mazloff, Matthew R.; Russell, Joellen L.
• Is Part Of: Geophysical research letters. Volume 48:Issue 21(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-11-09
• Subjects: predictability -- perturbation experiment -- Argentine basin -- biogeochemical (BGC) ocean modeling -- ocean forecasting; Geophysics -- Periodicals; Planets -- Periodicals; Lunar geology -- Periodicals; 550
• DOI: 10.1029/2021GL095504
• Languages: English
• ISSNs: 0094-8276

Abstract

Abstract: The confluence of the Malvinas and Brazil currents over the Argentine Basin give the region chaotic dynamics and severely limit potential predictability. To probe the forecast horizon for ocean surface quantities of temperature and carbon, we construct regional models of the Argentine Basin with biogeochemistry at 1/3° and 1/12° resolution and design a series of experiments. We add positive and negative zonal wind stress anomalies over small and large areas during a short period in different model runs. We calculate the response of the surface temperature and DIC. The 1/3° model maintains predictability for up to 45 days, while the 1/12° model has a shorter window of about two weeks. However, the 1/3° model response is only consistent with the 1/12° model for about 8 days calling into question the potential predictive skill of the coarser model at longer lead times. Plain Language Summary: The Argentine Basin is a turbulent region in the Southern Ocean that is not well understood. We want to determine how well we can make forecasts of the region given some level of error in our starting point, and additionally investigate the dependence of model resolution on our ability to forecast. We design a series of numerical experiments where we make a change in the winds at the start of the model run and quantify how the heat and carbon in the surface ocean change compared to model runs without changing the winds. These differences are measured as time progresses. We determine that in the coarser resolution models, the differences decay for up to 45 days before increasing, whereas in the finer resolution model the differences decay for two weeks before increasing. This gives us a time frame during which small errors in the forecast models do not compound. However, a caveat is that though the coarser model will be less sensitive to errors our results also suggest that it may be less accurate in forecasting the true ocean dynamics. Key Points: Wind stress perturbation experiments are conducted in two models of the Argentine Basin with different resolutions The predictability of SST and surface DIC is approximately one to two weeks for a 1/12° model This can be extended to 45 days in the 1/3° model with a corresponding loss of model fidelity