The study of water masses is important as they transport water properties affecting the biosphere and ocean dynamics. In this study, we revisit water masses in the Caribbean Sea using climatology and 11 months of observations at different depths from 3 moorings placed in the Guajira upwelling region, providing some new findings. The Caribbean Surface Water (CSW) seasonal variability is studied at the mixed layer depth. Salinity differences between CSW and the saltier North Atlantic Subtropical Underwater (SUW) determine static stability spatial and temporal variations, with implications for regional ocean dynamics. Besides, we assess the climatologic distribution of water masses below the salinity maximum using the optimum multiparameter analysis and the Thermodynamic Equation of Seawater 2010, defining their source water indices when entering the Caribbean Sea. The SUW, with its core at ~ 150 m depth, occupies 16% of the Caribbean Sea volume, complemented by 38% of Antarctic Intermediate Water, with its core at ~ 700 m depth and North Atlantic Deep Water, which as bottom water occupies 46% of the volume. Hydrographic observations do not differ from climatology, regardless of their large sub-annual variations decreasing with depth. Daily time series of dominant water fractions at different depths correlate at each mooring, indicating a common forcing. Besides, rotated wind stress, which is an indicator of the Guajira upwelling, correlates regularly with water mass fractions down to 700 m depth. However, during strong wind shifts, upwelling seems to affect them down to 1450 m depth.
The Inter-Americas Seas (IAS), involving the Gulf of Mexico, the Caribbean and a section of the eastern tropical Pacific Ocean bordering Central America, Colombia and Ecuador, exhibits very active ocean-land-atmosphere interactions that impact socio-economic activities within and beyond the region, and that are still not well understood or represented in state-of-the-art models. On seasonal-to-interannual timescales, the main source of variability of this geographical area is related to interactions between the Pacific and the Atlantic oceans, involving to anomalous sea-surface temperature (SST) patterns like El Niño-Southern Oscillation (ENSO), and regional features in the Caribbean linked to the bi-modal seasonality of the Caribbean Low-Level Jet. This study investigates seasonal-to-interannual IAS surface-temperature anomalies in observations, and their representation in am eddy-permitting, 1/9o-resolution simulation using the Regional Ocean Modeling System (ROMS), interannually-forced by the Climate Forecast System Reanalysis. Here, rather than analyzing model biases locally (i.e., gridbox-by-gridbox), a non-local SST pattern-based diagnostic was conducted via a principal component analysis. The approach allowed to identify magnitude, variance and spatial systematic errors in SST patterns related to the Western Hemisphere Warm Pool, ENSO, the Inter-American Seas Dipole, and several other variability modes. These biases are mainly related to errors in surface heat fluxes, misrepresentation of air-sea interactions impacting surface latent cooling in the Caribbean, and too strong sub-surface thermal stratification, mostly off the coast of Ecuador and northern Peru.