Despite the ever-growing number of ocean data, the interior of the ocean remains undersampled in regions of high variability such as the Gulf Stream. In this context, neural networks have been shown to be effective for interpolating properties and understanding ocean processes. We introduce OSnet (Ocean Stratification network), a new ocean reconstruction system aimed at providing a physically consistent analysis of the upper ocean stratification. The proposed scheme is a bootstrapped multilayer perceptron trained to predict simultaneously temperature and salinity (T−S) profiles down to 1000 m and the mixed-layer depth (MLD) from surface data covering 1993 to 2019. OSnet is trained to fit sea surface temperature and sea level anomalies onto all historical in situ profiles in the Gulf Stream region. To achieve vertical coherence of the profiles, the MLD prediction is used to adjust a posteriori the vertical gradients of predicted T−S profiles, thus increasing the accuracy of the solution and removing vertical density inversions. The prediction is generalized on a $\mathrm{1}/\mathrm{4}$ ^∘ daily grid, producing four-dimensional fields of temperature and salinity, with their associated confidence interval issued from the bootstrap. OSnet profiles have root mean square error comparable with the observation-based Armor3D weekly product and the physics-based ocean reanalysis Glorys12. The lowest confidence in the prediction is located north of the Gulf Stream, between the shelf and the current, where the thermohaline variability is large. The OSnet reconstructed field is coherent even in the pre-Argo years, demonstrating the good generalization properties of the network. It reproduces the warming trend of surface temperature, the seasonal cycle of surface salinity and mesoscale structures of temperature, salinity and MLD. While OSnet delivers an accurate interpolation of the ocean stratification, it is also a tool to study how the ocean stratification relates to surface data. We can compute the relative importance of each input for each T−S prediction and analyse how the network learns which surface feature influences most which property and at which depth. Our results demonstrate the potential of machine learning methods to improve predictions of ocean interior properties from observations of the ocean surface.

Through the Synthetic Aperture Radar (SAR) embarked on the satellites Sentinel-1A and Sentinel-1B of the Copernicus program, a large quantity of observations is routinely acquired over the oceans. A wide range of features from both oceanic (e.g., biological slicks, icebergs, etc.) and meteorologic origin (e.g., rain cells, wind streaks, etc.) are distinguishable on these acquisitions. This paper studies the semantic segmentation of ten metoceanic processes either in the context of a large quantity of image-level groundtruths (i.e., weakly-supervised framework) or of scarce pixel-level groundtruths (i.e., fully-supervised framework). Our main result is that a fully-supervised model outperforms any tested weakly-supervised algorithm. Adding more segmentation examples in the training set would further increase the precision of the predictions. Trained on 20 × 20 km imagettes acquired from the WV acquisition mode of the Sentinel-1 mission, the model is shown to generalize, under some assumptions, to wide-swath SAR data, which further extents its application domain to coastal areas.

Phytoplankton plays a key role in the carbon cycle and supports the oceanic food web. While its seasonal and interannual cycles are rather well characterized owing to the modern satellite ocean color era, its longer time variability remains largely unknown due to the short time-period covered by observations on a global scale. With the aim of reconstructing this longer-term phytoplankton variability, a support vector regression (SVR) approach was recently considered to derive surface Chlorophyll-a concentration (Chl, a proxy of phytoplankton biomass) from physical oceanic model outputs and atmospheric reanalysis. However, those early efforts relied on one particular algorithm, putting aside the question of whether different algorithms may have specific behaviors. Here, we show that this approach can also be applied on satellite observations and can even be further improved by testing performances of different machine learning algorithms, the SVR and a neural network with dense layers (a multi-layer perceptron, MLP). The MLP, thanks to its ability to capture complex non-linear relationships, outperforms the SVR to capture satellite Chl spatial patterns (correlation of 0.75 vs. 0.65 on a global scale, respectively) along with its interannual variability and trend, despite an underestimated amplitude. Among deep learning algorithms, neural network such as MLP models appear to be promising tools to investigate phytoplankton long-term time-series.

Over the core region of the Agulhas Current, new estimations of ocean surface velocities are reported using the increasing data set from the Automatic Identification System (AIS), initially designed to monitor vessel traffic. A two‐step strategy is suggested. A first guess is evaluated from the collective behavior of vessels for a given space‐time interval. Individual vessel trajectories are then re‐analyzed and interpolated. Applied during year 2016, these ocean surface current estimates are demonstrated to well determine the intensity of surface currents. The improved spatial resolution helps the decomposition of the optimally interpolated surface current vector field between irrotational ≈80%–90% and divergence‐free components ≈10%–20%, for example, Helmholtz‐Hodge decomposition. Comparisons are performed between in situ drifting‐buoys and represent up to 25% gain in respect of the altimetry gridded current (for the meridional component). Others comparisons with data collected during the Agulhas Current Time‐series experiment, as well as with the mean Doppler‐derived surface currents obtained from satellite synthetic aperture radar measurements also reveal a significant benefit of using the AIS derived estimates. Comparisons with the Sea Surface Temperature from MODIS sensors confirm the occurrence of meandering events for the current path. For the Agulhas Current region, the high density of vessel traffic can provide new means to study and monitor intense upper ocean currents with more detailed resolution and precision.

More than twelve satellite scatterometers have operated since 1992 through the present, providing the main source of surface wind vector observations over global oceans. In this study, these scatterometer winds are used in combination with radiometers and synthetic aperture radars (SAR) for the better determination and characterization of high spatial and temporal resolution of regional surface wind parameters, including wind speed and direction, wind stress components, wind stress curl, and divergence. In this paper, a 27-year-long (1992–2018) 6-h satellite wind analysis with a spatial resolution of 0.125° in latitude and longitude is calculated using spatial structure functions derived from high-resolution SAR data. The main objective is to improve regional winds over three major upwelling regions (the Canary, Benguela, and California regions) through the use of accurate and homogenized wind observations and region-specific spatial and temporal wind variation structure functions derived from buoy and SAR data. The long time series of satellite wind analysis over the California upwelling, where a significant number of moorings is available, are used for assessing the accuracy of the analysis. The latter is close to scatterometer wind retrieval accuracy. This assessment shows that the root mean square difference between collocated 6-h satellite wind analysis and buoys is lower than 1.50 and 1.80 m s⁻¹ for offshore and nearshore locations, respectively. The temporal correlation between buoy and satellite analysis winds exceeds 0.90. The analysis accuracy is lower for 1992–1999 when satellite winds were mostly retrieved from ERS-1 and/or ERS-2 scatterometers. To further assess the improvement brought by this new wind analysis, its data and data from three independent products (ERA5, CMEMS, and CCMP) are compared with purely scatterometer winds over the Canary and Benguela regions. Even though the four products are generally similar, the new satellite analysis shows significant improvements, particularly in the upwelling areas.