Fluctuations of net primary production along a standard transect in the Barents Sea and their relationships with environmental factors.

ABSTRACT Accurate mid-short term prediction of sea surface salinity (SSS) is essential for operational ocean monitoring, particularly for capturing short-term salinity fluctuations that affect regional ocean dynamics and weather conditions. However, existing models struggle to extract complex spatiotemporal dependencies and are often limited to local regions, reducing their global applicability. To address these challenges, we propose a Dynamic Adaptive Graph Convolutional Recurrent Network (DAGCRN) for global SSS prediction. The DAGCRN employs an encoder–decoder architecture, where both the encoder and decoder integrate adaptive graph convolutional recurrent units (AGCRUs) and gated recurrent units (GRUs). AGCRUs dynamically construct topological relationships via graph convolution to model spatial variations, while GRUs capture temporal dependencies. This enables DAGCRN to effectively model the nonlinear and dynamic nature of global SSS variations. We evaluate DAGCRN's performance on the ESA Sea Surface Salinity CCI v3.21 dataset, which provides global gridded SSS observations from February 2010 to September 2020. Forecasting lead times range from 1 to 12 days. DAGCRN consistently outperforms LSTM, BiLSTM, ConvLSTM, and TCN. For 12-day prediction, RMSE is reduced by 36.0%, 24.4%, 13.0%, and 5.5%, respectively, demonstrating its effectiveness in modeling spatiotemporal dependencies for global SSS forecasting.

Abstract. High-frequency variability of the partial pressure of CO2 (pCO2) in coastal environments reflects the complex interplay of physical, chemical and biological drivers. Multiscale statistical approaches provide a robust framework for understanding dynamics across timescales and for reliably assessing coastal carbon processes. In this study, pCO2 has been measured on the Astan cardinal buoy (Brittany, west coast of France) with at 30 min intervals by Gac et al. (2020), yielding a dataset of 32 582 data points collected over a period of nearly five years. These measurements were then coupled with others of sea surface temperature and salinity, chlorophyll a, oxygen saturation and atmospheric pressure. The aim of this study was to consider the statistical properties of the thermal and non-thermal component of pCO2, based on its relation with temperature established by Takahashi et al. (2009). Using Fourier spectral analysis, it was demonstrated that all marine scalars exhibited scaling properties with power-law slopes ranging from 1.73–1.85 for timescales spanning from 12 h to at least 80–100 d. The results obtained from this analysis indicate a turbulent and intermittent dynamics for all the considered scalars, including sea surface temperature and salinity, chlorophyll a, oxygen saturation, pCO2, and pCO2 thermal and non-thermal components. A time-reversibility analysis evidenced the irreversibility of the pCO2 components above 30 d. The irreversibility exhibited by the thermal component was found to be higher than that of the non-thermal component, with an average value of the associated irreversibility index that was approximately 3.5 times higher than that of the non-thermal component over the period of 50–70 d. Furthermore, a methodology known as the Probability Density Function quotient was employed, a method that has not been widely utilized. This approach enabled the identification of values for which there were statistical relationships between variables. This facilitated the quantification of the influence of primary production on the non-thermal pCO2, or the influence of periods of depression on supersaturation due to atmospheric or terrigenous inputs. This provided new insights into the stochastic coupling between biological and physical processes, when considering high-frequency pCO2 variability.

Robust daily satellite sea surface salinity reconstruction using deep learning in low-salinity coastal regions.

Given the active development of the Arctic with an extremely rare network of observations, there is a high demand for reliable hydrometeorological forecasts of the ice, marine and meteorological conditions in this region. For this purpose, a system has been implemented for hydrometeorological forecasting of atmospheric, ocean and sea ice circulation parameters for the White Sea region. The polar version of the WRF model was used for predicting atmospheric circulation, the ROMS model was used for predicting ocean (sea) circulation, and the parameters of the sea ice state were calculated using the CICE model. Early results of calculating hydrometeorological parameters have been obtained and an assessment of the quality of calculations has been carried out, which helped to identify the advantages and disadvantages of the system used. For the atmospheric calculations, the errors are at or below the published estimates from similar papers. The fields of sea surface temperature, surface salinity, and ocean level are in good agreement with the GOFS 3.1 analysis data and are at the level of other authors' quality assessments. Inaccuracies have been identified in the reproduction of the above characteristics at the ice /open water boundary. For the sea surface temperature, errors at the ice /open water boundary reached 0.4 °C, for salinity 0.4 ‰, for current velocity up to 0.18 m/s, and a level of 0.2 m. A comparative analysis was carried out for two schemes of parameterization of ice thermodynamics in the CICE — BL99 and Mushy models. It is shown that when both schemes are used, a systematic overestimation of the total volume of sea ice is observed. However, compared to the Mushy scheme, the simpler BL99 scheme had fewer errors.

Enhancing sea surface salinity short-term prediction using physically informed deep learning

Brightness temperature is operationally used to retrieve sea surface salinity (TB-SSS) over the global ocean, but is contaminated by land and sea ice in close proximity. Ocean color can be used to retrieve SSS (OC-SSS) via the relation between color and salinity, but this relation is only valid over the coastal ocean with terrestrial influence. Important ecological areas exist where both spectral domains can provide SSS estimates. Here we compare these estimates over the St. Lawrence Estuary and Gulf in Eastern Canada, where a large collection of near-surface in situ salinity measurements is available. While TB-SSS faces a significant limitation in undersampling spatial variability, OC-SSS is predominantly hindered by cloud cover. Offshore, TB-SSS data are considerably more abundant than OC-SSS data, the latter of which are available only about 30% as often as the former. However, OC-SSS estimates extend into more nearshore areas, such as the St. Lawrence Estuary. Additionally, OC-SSS estimates are more accurate, with a root mean square difference of 0.46 g kg −1 compared to 0.79 g kg −1 for TB-SSS. We employed each of these satellite-derived SSS products to compare the pronounced freshwater pulse of 2017 and post-tropical storm Dorian of fall 2019, finding that short-lived events were better captured by the OC-SSS product. In contrast, the TB-SSS product offered more extensive temporal coverage but smoothed out such events. Our analyses underscore the need for higher-resolution satellite salinity-sensors in coastal studies. In the meantime, ocean color data resolves submesoscale features and can help enhance our understanding of these dynamic environments.

Abstract The assimilation of satellite sea surface salinity (SSS) has been acknowledged to improve the upper-ocean stratification and the relevant oceanic processes. In this study, we focus on assessing the impact of SSS assimilation on predicting the Madden–Julian oscillation (MJO) propagation across the Maritime Continent for eight representative MJO cases using the National Aeronautics and Space Administration (NASA) Goddard Earth Observing System (GEOS) subseasonal to seasonal (S2S) prediction system, version 2 (GEOS-S2S-2). Two sets of forecast experiments are performed: one initialized from the operational ocean analysis without SSS assimilation, referred to as control (CTL), and the other initialized from the ocean analysis with SSS assimilation (SSS). Evaluated with the large-scale precipitation tracking method for these eight MJO events, the SSS forecasts outperform the CTL forecasts, showing better agreement with the observed eastward propagation over the Maritime Continent. In the SSS forecasts, the deeper mixed-layer depth (MLD) and greater upper-ocean heat content (OHC) result in reduced surface cooling during MJO convection. This leads to higher sea surface temperature (SST) compared to the CTL, which enhances latent heat flux anomalies via wind convergence and further supports MJO propagation. Based on the results of these representative MJO cases, this study suggests that improved initialization of ocean stratification and OHC in the upper ocean, enabled by SSS assimilation, can help partially overcome the MJO Maritime Continent prediction barrier and thus strengthen subseasonal forecast skills.

Abstract Benguela Niño and Niña events are episodes of extreme warming and cooling off Angola with impacts on fisheries, ecosystems, and rainfall in southwest Africa. They are typically forced remotely or locally by variations in equatorial or alongshore winds, respectively. We use an extensive in‐situ data set to show that sea surface salinity (SSS) changes can also act as a local forcing that amplifies these extreme warm and cold events by altering the water column stratification and consequently the impact of subsurface mixing. The mixed layer turbulent heat loss during an extreme warm episode with unusually low SSS in 1995 is nearly 3× lower than during a cold event with high SSS in 1997. We also demonstrate that interannual turbulent heat flux variability in early boreal spring off Angola is strongly impacted by salt advection fluctuations, and that this turbulent mixing is significant for altering mixed layer temperatures and restoring its salinities.

Temporal variability in environmental influences on silver croaker (Pennahia argentata) life-history traits in the East China Sea.

Abstract The partial pressure of carbon dioxide at the sea surface (pCO 2sea ) is a key component in the ocean carbon cycle, jointly influenced by the thermodynamic, dynamical, and biological processes. Among these, thermodynamic control generally induces a pronounced positive covariation between pCO 2sea and sea surface temperature (SST). However, using multiple observations, reanalysis data sets, and BIO‐ROMS model simulations, this study reveals an anomalous decoupling between pCO 2sea and SST in the northern Bay of Bengal (BoB) during Indian Ocean dipole (IOD) events, highlighting the importance of non‐thermal mechanisms in determining pCO 2sea interannual variability. During positive IOD events, anomalies in dissolved inorganic carbon (DIC) and sea surface salinity (SSS) jointly exert positive influences on pCO 2sea , though partially offset by the opposite SST effects. Specifically, IOD‐related negative SST anomalies in the equatorial eastern Indian Ocean trigger the Matsuno‐Gill atmospheric response, enhancing evaporation and suppressing precipitation over the northern BoB. This leads to positive freshwater flux anomalies that elevate both DIC and SSS, contributing to increased pCO 2sea . Simultaneously, positive wind stress curl anomalies in the northwestern BoB enhance cyclonic eddy and upwelling, bringing the colder, DIC‐rich subsurface water into the mixed layer. Overall, these processes result in surface cooling while further enriched DIC. Moreover, anomalous southerly wind in the southwestern BoB weakens the East India Coastal Current, facilitating anomalous transport of saline water that enhances positive SSS anomalies, thereby increasing pCO 2sea anomalies. Our findings underscore the complex interplay between thermodynamic and dynamical processes in shaping BoB carbon cycle variability under IOD influence.

Synoptic view of an unprecedented red tide in autumn/winter in the Bohai Sea, China, triggered by extreme rainfall events.

Climate-driven expansion of Takayama acrotrocha blooms: First outbreak along the north coast of Shandong, China linked to intense precipitation and warming.

Evaluation zooplankton community and energy transfer efficiency: A case in the coastal waters of Shandong, China.

Living on the edge: Temperature and salinity performance curves across levels of biological organization in a shallow water shrimp.

Impact of spatial and temporal resolution of satellite sea surface salinity measurements on ocean state prediction in the Tropical Indian Ocean; an OSSE framework using SMOS

A multi-scale spatiotemporal feature network for sea surface salinity forecast in the eastern tropical Pacific Ocean

Abstract Starting in late 2023, sea surface salinity estimates derived from measurements made by the SMAP L-band radiometer have been increasingly contaminated by radio frequency interference (RFI). We study various methods of identifying RFI-affected SMAP measurements over ocean during Level 2 processing. The most reliable RFI detection techniques are the chi-squared of the maximum likelihood estimator of the salinity retrieval algorithm, the difference in observed brightness temperature between forward and backward looking parts of the swaths, and the value of the observed 4 th Stokes parameter. These detection methods do not rely on any external ancillary salinity data. We demonstrate that the number of false alarms in the RFI detection can be reduced by checking for spatial clusters of cells that are affected. Additionally, the detection rate can be improved by flagging neighbors of a cell in which RFI has been detected. We test and evaluate the RFI detection algorithm in various case studies and assess missed detection and false alarm rates.

Simple SummaryCoastal China seas’ fish communities face threats like overfishing and climate change, but how these communities react to these threats is unclear. This study aimed to understand what shapes these fish communities using a method of community modelling. We analyzed data on 384 fish species (1980–2018) and environmental factors. The results showed that temperature and salinity mostly determine fish distribution, and fish prefer silt over fine sand habitats. Goby fish have more connections with other fish. The findings help predict coastal fish communities and guide efforts to protect their biodiversity, benefiting ocean health and related human activities.To address uncertainties in how threatened coastal China seas fish communities respond to stressors like overfishing and climate change, this study applied Hierarchical Modelling of Species Communities (HMSC) to disentangle the assembly rules shaping these communities, filling a critical gap in understanding their spatiotemporal dynamics. We analyzed data on 384 fish species (1980–2018) and key environmental factors, with variance partitioning revealing that environmental filtering dominated fish distributions (explaining over 99% of variance), far outweighing random effects (0.60%). Among environmental drivers, sea surface temperature (49.00%) and sea surface salinity (33.25%) were the most influential, while seafloor substrate and water depth played secondary roles; notably, fewer species occupied fine sand habitats, and more preferred silt habitats. Residual species associations—indicative of potential biotic interactions—were most frequent within Gobiidae, likely due to this highly diverse taxon’s specialized resource utilization and wide distribution, highlighting that biotic filtering is concentrated and ecologically relevant within this group. This work demonstrates HMSC’s utility in unraveling coastal fish community assembly, providing a robust basis for predicting community changes and guiding biodiversity conservation efforts that support ocean health and dependent human activities.

Abstract. The subpolar gyre is at risk of crossing a tipping point under future climate change associated with the collapse of deep convection. As such, tipping can have significant climate impacts; it is important to understand the mechanisms at play and how they are represented in modern climate models. In this study, we use causal inference to investigate the representation of several proposed mechanisms for subpolar gyre variability in CMIP6 models. As expected, an increase in sea surface salinity or a decrease in sea surface temperature leads to an increase in mixed layer depth in nearly all CMIP6 models due to an intensification of deep convection. However, the effect of convection on modifying sea surface temperature due to re-stratification is less clear. In most models, the deepening of the mixed layer caused by an increase in sea surface salinity does result in a cooling of the water at intermediate depths. The feedback from the subsurface temperature through density to the strength of the subpolar gyre circulation is more ambiguous, with fewer models indicating a significant link. Those that do show a significant link do not agree on its sign. The CMIP6 models that have the expected sign for the links from density to the subpolar gyre strength and from there to sea surface salinity are also the models in which abrupt shifts in the subpolar gyre region have been found in climate change scenario runs. One model (CESM2) contains all proposed mechanisms, with both a negative and a delayed positive feedback loop being significant.