Sea ice drift underpins air-sea ice coupling and model evaluation, but the Southern Ocean remains chronically under sampled relative to the Arctic. This work presents a cross-polar comparison of GNSS-tracked ice-drift time series using rotary spectral analysis, principal component analysis, inter-buoy coherence, complementary shape- and amplitude-sensitive spectral differences. Arctic data are aggregated into Beaufort Gyre and Transpolar Drift seasonal composites (2017-2024); Antarctic data comprise eight heterogeneous campaigns (2000-2022) with small buoy counts and short durations. After common quality control, the clearest cross-polar similarity is spectral organisation: both regions show strong low-frequency variance and a hemisphere-appropriate rotary enhancement near the Coriolis frequency, interpreted cautiously as a combined inertial-semidiurnal response. Coherence is highest at synoptic scales and declines towards higher frequencies. Southern Ocean campaigns occupy a higher-energy envelope, so similar band structure does not imply similar drift amplitude. Principal components capture much array-scale motion, but high cumulative variance does not substitute for dense sampling of smaller-scale processes. The results provide transferable frequency-resolved benchmarks (dominant variance below 0.5 cpd and in the near-inertial band, an expanded Arctic reference, and explicit separation of spectral shape from amplitude) to inform future observing-system design without assuming identical dynamical regimes.
We present a method for forecasting water levels at river open boundaries in coastal ocean models, where tidal signals are influenced by variable river discharge. The method uses non-stationary tidal analysis (NS_TIDE) to generate future water levels based on a deterministic tidal signal and extrapolations of upstream hydrological predictors. This allows the river boundary to be placed within the tidal propagation zone without modelling the full tidal extent or coupling a separate river model. Case studies for the Fraser and Saint John rivers demonstrate that the method reproduces water levels and river outflow skilfully under a range of seasonal flow conditions. Forecast accuracy is comparable to gauge-driven configurations in most downstream areas, with root mean square values of model data using gauge data or NS_TIDE for river water level differing by no more than five centimetres. Performance gradually declines upstream. This novel approach in modelling tidal estuaries reduces computational cost, avoids the complexity of coupled river modelling, and is well suited to operational forecasting systems requiring short-range river boundary forcing.
The presence of unusually low-salinity water from the Changjiang River (CR) poses a significant marine environmental threat in the East China Sea (ECS), disturbing aquatic habitats and ecosystems in surrounding countries' waters. Demand for early warning of such threats by the fishing and aquaculture industries has been increasing in recent decades. This article introduces the Data-assimilative Regional Ocean Prediction system (DROPS), an operational forecasting system designed to predict the spatiotemporal behavior of low-salinity plumes in the northern ECS. DROPS is based on a regional ocean model with a four-dimensional variational data assimilation, covering the Yellow Sea and ECS shelves. This operational system runs automatically once a day, including a 72-hour hindcast and a 120-hour forecast, using initial conditions from the previous assimilation run. Predictions were compared with real-time measurements from weather buoys, ocean research stations, and unmanned wave gliders. DROPS successfully captured the low-salinity and high-temperature water observed over the northern ECS during the summer of 2022, showing moderate skill in predicting salinity and temperature variations. Data assimilation decreased temperature and salinity errors by 27% and 22%, respectively, improving forecasting performance. Our prediction system provides essential information to support decision-making for issuing early warnings of low-salinity and high-temperature water events.
Astronomical tides are a primary driver of sea level variability in coastal areas, significantly contributing to high water extremes and coastal flooding. Therefore, accurate tide predictions are essential for analysing extreme sea levels and developing robust coastal sea level predictions. Classical harmonic tidal analyses rely on subjective user choices, such as training period length, temporal separation between training and prediction periods, selection of tidal constituents, and specific harmonic analysis methods (e.g. nodal correction factors). Here we show that these choices significantly impact tidal reconstructions and predictions. For example, the amplitude of the annual and semi-annual components can vary by up to 30% over the historical period when using 5-year moving windows, highlighting the importance of the training period selection. High-frequency tidal constituents' amplitude is less sensitive but still varies by up to 5% over the last 40 years. Furthermore, different choices in harmonic analysis result in differences exceeding 10 cm between tide predictions, a common accuracy benchmark in ports. These findings inform users of harmonic tidal analysis about the strengths and weaknesses of specific sets of parameters chosen for the tidal analysis. Based on our results, we propose a list of recommendations for tidal analysis tailored to different situations.
Many industries rely on wave data to understand the potential for wave energy extraction, or to understand the wave environment for the design of marine structures and to plan operations and maintenance. Three ocean reanalysis datasets, ERA5, WAVEWATCH III and Copernicus Global Ocean Waves Analysis and Forecast, are compared to in-situ wave buoy data collected along the north of Scotland. All reanalysis datasets correlated well with the wave buoy data, with the Copernicus Global Ocean Waves Analysis and Forecast dataset being statistically the closest to the buoy data. However, all three reanalysis datasets underpredict significant wave height during extreme wave events. From comparisons of the wave buoy data at one site, it was found that although extreme events are underpredicted, the WAVEWATCH III reanalysis data performed the best, although still under predicted extreme wave heights. Of the reanalysis models compared against wave buoy data here, it is suggested that for extreme wave analysis the WAVEWATCH III model is recommended, whilst for long term statistics and weather windowing the Copernicus Global Ocean Waves Analysis is a good option. Whilst reanalysis data sets are a valuable resource for marine renewable energy, developers should be aware of the limitations of these datasets, in particular for extreme wave conditions.
