Surface ocean dynamics play a key role in the Earth system, contributing to regulate its climate and affecting the marine ecosystem functioning. Dynamical processes occur and interact in the upper ocean at multiple scales, down to, or even less than, few kilometres. These scales are not adequately resolved by present observing systems, and, in the last decades, global monitoring of surface currents has been based on the application of geostrophic balance to absolute dynamic topography maps obtained through the statistical interpolation of along-track satellite altimeter data. Due to the cross-track distance and repetitiveness of satellite acquisitions, the effective resolution of interpolated data is limited to several tens of kilometres. At the kilometre scale, sea surface temperature pattern evolution is dominated by advection, providing indirect information on upper ocean currents. Computer vision techniques are perfect candidates to infer this dynamical information from the combination of altimeter data, surface temperature images and observing-system geometry. Here, we exploit one class of image processing techniques, super-resolution, to develop an original neural-network architecture specifically designed to improve absolute dynamic topography reconstruction. Our model is first trained on synthetic observations built from a numerical general-circulation model and then tested on real satellite products. Provided concurrent clear-sky thermal observations are available, it proves able to compensate for altimeter sampling/interpolation limitations by learning from primitive equation data. The algorithm can be adapted to learn directly from future surface topography, and eventual surface currents, high-resolution satellite observations.

The Arabian Sea and Sea of Oman circulation and water masses, subject to monsoon forcing, reveal a strong seasonal variability and intense mesoscale features. We describe and analyze this variability and these features, using both meteorological data (from ECMWF reanalyses), in situ observations (from the ARGO float program and the GDEM – Generalized Digital Environmental mode – climatology), satellite altimetry (from AVISO) and a regional simulation with a primitive equation model (HYCOM – the Hybrid Coordinate Ocean Model). The model and observations display comparable variability, and the model is then used to analyze the three-dimensional structure of eddies and water masses with higher temporal and spatial resolutions than the available observations. The mesoscale features are highly seasonal, with the formation of coastal currents, destabilizing into eddies, or the radiation of Rossby waves from the Indian coast. The mesoscale eddies have a deep dynamical influence and strongly drive the water masses at depth. In particular, in the Sea of Oman, the Persian Gulf Water presents several offshore ejection sites and a complex recirculation, depending on the mesoscale eddies. The associated mechanisms range from coastal ejection via dipoles, alongshore pulses due to a cyclonic eddy, to the formation of lee eddies downstream of Ra's Al Hamra. This water mass is also captured inside the eddies via several mechanisms, keeping high thermohaline characteristics in the Arabian Sea. The variations of the outflow characteristics near the Strait of Hormuz are compared with variations downstream.

Measurements of ocean surface topography collected by satellite altimeters provide geostrophic estimates of the sea surface currents at relatively low resolution. The effective spatial and temporal resolution of these velocity estimates can be improved by optimally combining altimeter data with sequences of high resolution interpolated (Level 4) Sea Surface Temperature (SST) data, improving upon present-day values of approximately 100 km and 15 days at mid-latitudes. However, the combined altimeter/SST currents accuracy depends on the area and input SST data considered. Here, we present a comparative study based on three satellite-derived daily SST products: the Remote Sensing Systems (REMSS, 1/10 ${}^{\circ }$ resolution), the UK Met Office OSTIA (1/20 ${}^{\circ }$ resolution), and the Multiscale Ultra-High resolution SST (1/100 ${}^{\circ }$ resolution). The accuracy of the marine currents computed with our synergistic approach is assessed by comparisons with in-situ estimated currents derived from a global network of drifting buoys. Using REMSS SST, the meridional currents improve up to more than 20% compared to simple altimeter estimates. The maximum global improvements for the zonal currents are obtained using OSTIA SST, and reach 6%. Using the OSTIA SST also results in slight improvements (≃1.3%) in the zonal flow estimated in the Southern Ocean (45 ${}^{\circ }$ S to 70 ${}^{\circ }$ S). The homogeneity of the input SST effective spatial resolution is identified as a crucial requirement for an accurate surface current reconstruction. In our analyses, this condition was best satisfied by the lower resolution SST products considered.

Sea surface temperature (SST) is one of the essential variables of the Earth's climate system. Being at the air–sea interface, SST modulates heat fluxes in and out of the ocean, provides insight into several upper and interior ocean dynamical processes, and is a fundamental indicator of climate variability potentially impacting the health of marine ecosystems. Its accurate estimation and regular monitoring from space is therefore crucial. However, even if satellite infrared/microwave measurements provide much better coverage than what is achievable from in situ platforms, they cannot sense the sea surface under cloudy and rainy conditions. Large gaps are present even in merged multi-sensor satellite products, and different statistical strategies, mostly based on optimal interpolation (OI) algorithms, have thus been proposed to obtain gap-free (L4) images. These techniques, however, filter out the signals below the space–time decorrelation scales considered, significantly smoothing most of the small mesoscale and submesoscale features. Here, deep learning models, originally designed for single-image super resolution (SR), are applied to enhance the effective resolution of SST products and the accuracy of SST gradients. SR schemes include a set of computer vision techniques leveraging convolutional neural networks to retrieve high-resolution data from low-resolution images. A dilated convolutional multi-scale learning network, which includes an adaptive residual strategy and implements a channel attention mechanism, is used to reconstruct features in SST data at $\mathrm{1}/\mathrm{100}$ ° spatial resolution starting from $\mathrm{1}/\mathrm{16}$ ° data over the Mediterranean Sea. The application of this technique shows an improvement in the high-resolution reconstruction, capturing small-scale features and providing a root-mean-squared-difference improvement of 0.02 °C with respect to the L3 ground-truth data.

The dynamics of the Sicily Channel and the southern Tyrrhenian Sea are highly influenced by the seasonal variability of the Mediterranean basin-wide circulation, by the interannual variability of the numerous mesoscale structures present in the Channel, and by the decadal variability of the adjacent Ionian Sea. In the present study, all these aspects are investigated using in-situ (Lagrangian drifter trajectories and Argo float profiles) and satellite data (Absolute Dynamic Topography, Sea Level Anomaly, Sea Surface Temperature, wind products) over the period from 1993 to 2018. The availability of long time series of data and high-resolution multi-sensor surface currents allow us to add new details on the circulation features and on their driving mechanisms and to detect new permanent eddies not yet described in literature. The structures prevailing in winter are mainly driven by wind, whereas those prevailing in summer are regulated by topographical forcing on surface currents. The strength of the surface structures located at the western entrance of the Ionian Sea and of the mesoscale activity along the northern Sicily coast is modulated by the large-scale internal variability. The vertical hydrological characteristics of these mesoscale eddies are delineated using the Argo float profiles inside these structures.