Ocean surface current observations are essential for monitoring upper-ocean transport of heat, nutrients, pollutants, validation of model forecasts, data assimilation, and marine operations. Existing surface current measurements from in situ drifters and moorings, shore-based radars, and satellite altimetry are either not available on a regular basis, cover only limited areas, and/or are not applicable near the coast. The Doppler shift acquired by satellite Synthetic Aperture Radars (SARs) over the ocean is a measure of the radial total surface motion induced by the near-surface wind, surface waves, and underlying surface currents. Given accurate removal of the sea state contributions, such data can be used to retrieve global surface current radial velocities (RVLs) with a high spatial resolution of 1 km. In this study, we developed an empirical Geophysical Model Function (GMF) for predicting the sea state contribution to the Doppler shift in order to improve the accuracy of the ocean surface current radial velocity retrievals in the coastal zone.

The Sentinel-1 mission is an operational constellation of two C-Band SAR instruments (A/B) launched in 2014/2016. Challenging calibration has prevented their usage for retrieving surface currents. Recently an experimental calibration emerged thanks to two months of telemetry from the gyroscope onboard Sentinel-1B, recalibrated on-ground, yielding promising improvement in the accuracy of the geometric Doppler shift component. Following this development, we generated experimental Sentinel-1B IW RVL products over the Norwegian coastal zone in December 2017 - January 2018 and collocated them with the wind and wave fields from the regional operational MEPS and MyWaveWAM models. We estimated that the Doppler shift from the experimental products has a better accuracy of about 3.8 Hz compared to the 6.8 Hz previously reported for standard products. We further trained the model function (CDOP3SiX) that predicts the sea state Doppler shift as a function of the range-directed wind, wind sea orbital velocity, swell orbital velocity, as well as incidence angle. As such, the CDOP3SiX accounts for a more realistic sea state representation compared to the previously used empirical models (e.g., CDOP) that relied only on access to the wind fields and assumption of fully developed wind sea in absence of swell. We found that the signal from the Norwegian Coastal Current of about 0.5 m/s can be systematically detected in the Sentinel-1 derived ocean surface current radial velocity fields with 1 km spatial resolution. Moreover, the Sentinel-1 derived surface currents also express the presence of meandering structures and boundaries consistent with the satellite-based sea surface temperature field. Comparison with the ocean model also reveals acceptable agreement, especially for the major surface current features.

Despite the study being constrained by only two months of data from the single Sentinel-1B satellite, the results are promising. Reprocessing of the full Sentinel-1 A/B dataset using novel attitude calibration is therefore essential for further improvement of the empirical algorithms and validation. The developed methodology can be applied for the observations from the operational Sentinel-1 mission (sustained operation until 2030) as well as adapted for candidate future satellite missions designed for monitoring of the upper ocean circulation (e.g., SEASTAR and Harmony). However, CDOP3SiX relies on the usage of collocated wind and wave forecasts that introduces corresponding errors and limits its application to the various regions. Therefore, the next generation of empirical models might explore the possibility to retrieve wind and wave fields directly from the SAR observations, thus avoiding the use of numerical model fields that cannot realistically represent the high spatial variability in the wind and wave conditions that are typical for given SAR acquisitions. In turn, a highly valuable dataset of Doppler-based radial velocities would stimulate more advanced studies of the upper ocean dynamics and comparison to numerical ocean model simulations and predictions, and, eventually, assimilation of the SAR-derived surface current retrievals.

The Doppler Centroid (DC) frequency shift recorded over ocean surfaces by Synthetic Aperture Radar (SAR) is a sum of contributions from satellite attitude/antenna and ocean surface motion induced by waves and underlying ocean currents. A precise calibration of the DC is needed in order to predict and subsequently remove contributions from attitude/antenna.
Recently, a novel data calibration technique based on combining gyroscope telemetry data and global Sentinel-1 WV OCN products (OceanData Lab Ltd, 2019), has demonstrated promising capabilities to quantify the Sentinel-1 (S1) attitude and hence provide calibrated estimates of the corresponding DC frequency shift. One year of S1 a and b WV OCN products, orbit data and gyroscope data are combined providing one year of restituted attitude data (AUX_ESTATT). For the same time period, mean DC bias versus elevation angle is computed on a daily basis from S1 IW land acquisitions (AUX_DCBIAS). The AUX_ESTATT and AUX_DCBIAS products are subsequently used to generate global data set of calibrated S1 WV OCN products as well as sub-sets of calibrate S1 IW OCN products from predefined super sites (Norwegian Coast, Agulhas, Mediterranean). In this paper we assess the accuracy and precision of the calibrated DC frequency of S1 WV and IW acquisitions acquired over both land and ocean areas. The DC standard deviation (STD) and bias show significant reduction for both satellites and for all swaths. Assessment of the performance of global WV data shows a STD around 6Hz, while the BIAS is less than 2 Hz. The performance is very similar for both satellites and for both swaths. For IW the STD is similar, but the bias is slightly higher and small DC bias between sub-swaths is sometimes observed.
The remaining errors are mainly due to change in antenna characteristics on a timescale not capture with the procedure used to generate the mean DC bias stored in the AUX_DCBIAS file. Such changes may come from thermo-elastic effects and/or temperature compensations applied to the antenna. This directly affects the IW mode DC, where it is also clearly visible in some scenes. For WV mode it mainly impacts the statistics.
We conclude that S1 WV mode has achieved a performance (i.e. accuracy and precision) within the requirement for climatology mapping of global ocean current features. For IW mode, we have achieved a precision within the requirement, but use of land areas within the scene is still required to achieve the required accuracy over all sub-swaths.

The ocean meso- and submesoscale dynamic processes are one of the gaps in our understanding of the ocean-atmosphere exchange of momentum and energy. Resolving the ocean currents and waves at submesoscale resolution is one of the missing pieces in the dynamic ocean-atmosphere processes puzzle. Harmony, an Earth Explorer 10 candidate currently in phase-A studies, proposes direct instantaneous Doppler frequency shift measurements of the ocean surface currents, surface wind stress, and wave spectra.

Harmony achieves the aforementioned measurements using two spacebourne synthetic-aperture radar (SAR) companion satellites flying together with Sentinel-1. Each of the two companions will form a bistatic SAR with Sentinel-1 as the illuminator, allowing for Doppler measurements of ocean surface velocities with two lines of sight and along-track interferometry (ATI). The novel acquisition geometry poses challenges in the modelling of measurement errors that are not addressed with methods for monostatic SAR systems.

In this study we present a quantitative assessment of the errors involved in the Doppler measurement of ocean surface vectors from bistatic, formation-flying SAR, particularly as it pertains to the Harmony mission. Specifically, we investigate the effects of two types of errors: random measurement noise and systematic errors related to the instrument and the satellite. Moreover, we present algorithms to correct for the systematic errors and evaluate calibration techniques for ocean Doppler measurements.

Measurement noise is typically the better understood type of error out of the two in the field of SAR Doppler measurements. It is modelled as thermal white noise driven by the ocean backscatter, the coherence of the measurement and the number of independent looks. The coherence is a function of the NESZ, the temporal baseline and the ambiguities. Mitigation thus can only be achieved by improving the instrument sensitivity and resolution during the design phase or by adjusting the baseline in the case of ATI. We present the impact of measurement noise on interferometric performance and the trade-offs in instrument design.

Systematic errors on the other hand, arise due to unknown perturbations in the output signal of the instrument driven by the electronics of the receiver and the processing chain, and by uncertainties in the position of the antenna phase centres and the position of the companions in the formation. The perturbations are a result of signal ambiguities and clock errors, while the position uncertainties can be due to attitude errors, structural vibrations and reference-height errors. All systematic errors materialise as an undesired phase offset in the Doppler measurement that needs to be estimated and calibrated for.

Calibration of a SAR instrument over the ocean surface is difficult due to the dynamic nature of the surface. Thus, calibration is in practice done over non-moving targets such as land. For such a technique to work, systematic errors must have a drift that stays as small as possible during acquisitions over the ocean. Understanding the spatial scale of the error variance and the rate at which decorrelation occurs is important in determining effective mitigation techniques for Harmony. Recognising the spatial dependence of the systematic errors, we propose adopting methods from the field of Spatial Statistics to better understand the uncertainties.

Direct measurement of the global ocean surface current is of great scientific interest and application value for understanding multiscale ocean dynamics, air-sea interaction, ocean mass and energy balance, and their variabilities under climate change. Presently, measurements of global ocean surface currents, which are mainly geostrophically derived from satellite altimeter data, are only available to resolve quasi-geostrophic current at large- to meso- scale in the off-equatorial open ocean. Ocean Surface Current multiscale Observation Mission (OSCOM) will launch a satellite equipped with a Doppler Scatterometer to directly measure ocean surface currents with a very high horizontal resolution of 5–10 km and a 3-day global coverage. OSCOM will provide an in-depth picture of non-equilibrium ocean state and air-sea interaction from mesoscale to submesoscale, and helps to construct the fine structure of deep ocean current through a combination with Argo profiling. Those direct measurements and derived dynamic parameters will further provide a novel and improved pathway to data assimilation and coupling of GCMs for ocean prediction and climate change.
OSCAR current and the surface drifters indicate that the non-geostrophic currents in the global ocean account for ~43% of the total current. Especially, the non-geostrophic currents determine the directions of the total currents in the near-equatorial trade winds and mid-latitude westerly winds prevailing regions, where the maximum non-geostrophic speed can reach twice the geostrophic speed and exceed 60% of the total current. The present current reanalysis cannot reveal the non-geostrophic processes in these regions and underestimate the weakening effect of the non-geostrophic process in the strong western boundary currents and the Antarctic Circumpolar Current. The influence of the non-geostrophic processes in the real world will be more significant, and the OSCOM is expected to reveal these processes in the future.

Ocean surface current is an essential ocean and climate variable, and it plays an important role in various scientific research and engineering applications. At present, there is only global geostrophic current derived from spaceborne altimetry, however, no direct global total ocean surface current vector observations from space are available. Geostrophic current is part of contribution to the total ocean surface current, and there are still ageostrophic current processes, however, the contribution of geostrophic current and ageostrophic current to the total ocean surface current is still unclear. Besides, retrieving the total ocean surface current is still challenging, as the total Doppler shift from ocean surface includes the contributions from sensor platform motions, wind-wave induced, and ocean surface current itself. The accurate wind-wave induced Doppler shift estimation is the major challenge for ocean surface current retrieval. All these issues need to be further investigated.
In the first half of this study, we use the most recent GDP (Global Drifter Program, GDP) drifter datasets from January 1st 2016 to December 31st 2020, which covers 6416 trajectories and more than 10 million observations, to investigate the contribution of geostrophic current to the total ocean surface current. The measured drifter velocity is the addition of the 15 meters depth large-scale current, the upper-ocean wind-driven current, the influence of tides and Stokes drifter and other forces on the drogue, and wind-induced slippage contribution. To account for the slippage contribution, a downwind velocity modeled as αW_s is subtracted from the drifter velocity, where W_s is 10-m wind speed, and α is the fraction of W_s converted to the slippage. After the slippage correction, we make a statistical comparison analysis between drifter velocity components, zonal (ut) and meridional (vt) velocity and the AVISO (Archiving Validation and Interpolation of Satellite Oceanographic Data, AVISO) geostrophic current, where the zonal (ugeo) and meridional (vgeo) velocities of geostrophic current are interpolated to the location of individual drifter observation by a three-dimensional time-space interpolation.
We carry out a direct comparison of the zonal velocity and meridional components between corrected drifter data and geostrophic current at a global scale, and calculate the probability in the cases of ugeo/ut < 0, vgeo/vt< 0. If ugeo/ut < 0 (vgeo/vt < 0), that indicates the ugeo (vgeo) is not the dominant contribution in the ut (vt), and the other ocean current processes dominate. Our preliminary results show that the portion of ugeo/ut < 0 reaches a value as large as 27.63% with a total number of effective observations is around 9.6 million, while the portion of vgeo/vt < 0 can be 31.59%. Additionally, while binning the corrected drifter data and the geostrophic current by a 0.5° × 0.5° bin, the portion of ugeo/ut < 0 and vgeo/vt < 0 could be 19.87% and 31.30%, respectively.
In the second half of this study, we investigate the effects of different wave directional spectra on the estimates of the wind-wave induced Doppler shift via a numerical Doppler model. By comparisons with some existing empirical or semi-empirical Doppler model, like CDOP model, KaDOP model, and KaDS model, our results show that the accurate estimation of wind-wave induced Doppler shift is highly dependent on the selection of wave spectra parameterization and directional spreading function. To accurately retrieve ocean current, a clear solution is to measure simultaneous and necessary ocean wave properties and wind vector to estimate the contribution that wind-wave induced.