The Baltic Sea is a brackish shallow sea, the state of which is determined by the mixing of freshwater from net precipitation and runoff with the salty water from the North Sea inflows. The freshwater content (FWC) of the Baltic Sea is calculated from the Copernicus regional reanalysis data covering the period 1993–2021. The FWC in the Baltic Sea has shown a steady decrease over the past 2 decades, with a linear trend of 23.9 km³ yr⁻¹; however, the trend has significant spatial variability. The Gulf of Bothnia has a positive FWC tendency, while the Baltic Proper has a negative FWC tendency. Temporal changes of FWC are opposite between the Bothnian Bay in the north and the southern Baltic Proper. In the Bothnian Bay, interannual changes of FWC are positively correlated with river runoff and net precipitation and negatively correlated with salt transport. In the southern Baltic Proper, the variations of FWC and salt transport through the Danish straits are negatively correlated from 1993 until 2010 but positively correlated thereafter. The seasonal freshwater content reflects the specific hydrophysical conditions of each sub-basin, with northern basins being influenced by seasonal river runoff and ice formation and melting, while the southern basins are more responsive to subsurface salinity changes due to salt transport through the Danish straits.

A method for reconstructing weekly Baltic gridded sea levels was developed and tested. This method uses input data from tide gauge and altimetry observations. The reconstruction is based on sea level empirical orthogonal function (EOF) modes, calculated as spatiotemporal statistics from daily model reanalysis results available from the Copernicus Marine Service for the 1993–2021 period. In the semi-enclosed, tideless Baltic Sea, the four leading EOF modes cover 99% of the sea level variance. Several experiments with different combinations of input data were carried out. This method was validated against coastal tide gauges and altimetry data. The best reconstruction was obtained when both the tide gauges and altimetry data were used as inputs. An assessment of the centered root-mean-square difference (cRMSD) of the reconstruction versus the tide gauges revealed a value of 0.05 m, and a result of 0.10 m was revealed versus altimetry. The average coefficient of determination (R²) was determined to be 0.93 for the tide gauges and 0.82 for the altimetry data. In the cases where only one type of input data was used, the reconstruction worsened with respect to other data sources. The reconstruction method demonstrated its usefulness for the reconstruction of coastal sea levels in unsampled locations and the calculation of changes in sea volume.

By decomposing the total sea surface current into its geostrophic and ageostrophic components, we examined the contribution of each to the long-term variability of the total sea surface current. Our findings demonstrate the importance of geostrophic currents in Baltic Sea gyre formations. Additionally, ageostrophic currents contribute significantly to the flow across the region. Quantifying the difference between total sea surface current fields has revealed two dominant general sea surface circulation patterns in the Baltic Sea, whose characteristics correspond to the monthly mean climatology of sea surface current fields in May and December. Subsequently, a machine learning technique was employed to effectively detect the types of sea surface circulation patterns using wind vectors and sea level anomaly fields. This underscored the combined influence of sea level anomaly-driven and wind-driven components in shaping surface current vectors in the Baltic Sea, consistent with geostrophic and ageostrophic decompositions.

Sea level predictions, typically derived from 3D hydrodynamic models, are computationally intensive and subject to uncertainties stemming from physical representation and inaccuracies in initial or boundary conditions. As a complementary alternative, data-driven machine learning models provide a computationally efficient solution with comparable accuracy. This study employs the deep-learning model HIDRA2 to forecast hourly sea levels at five coastal stations along the Estonian coastline of the Baltic Sea, evaluating its performance across various forecast lead times. Compared to the regional NEMO_BAL and subregional NEMO_EST hydrodynamic models, HIDRA2 frequently outperforms both, particularly in terms of overall forecast skill. While HIDRA2 shows limitations in resolving high-frequency sea level variability above (6 h)⁻¹, it effectively reproduces energy in lower-frequency bands below (18 h)⁻¹. Errors tend to average out over longer time windows encompassing multiple seiche periods, enabling HIDRA2 to surpass the overall performance of the NEMO models. These findings underscore HIDRA2's potential as a robust, efficient, and reliable tool for operational sea level forecasting and coastal management in the eastern Baltic Sea region.

Fuel efficiency optimization is of crucial importance in industries. Marine transportation industry is no exception. Multi-disciplinary optimization is a branch of engineering which uses optimization methods for solving problems in which the objective function is simultaneously affected by several different factors. As one of the tools for this type of optimization, genetic algorithm has a high quality and validity. The objective of the present study is to optimize fuel efficiency in tankers. All presented equations and conditions are valid for tankers. Fuel consumption efficiency of tankers is a function of various influential factors. Given the lack of equations for describing and modeling these factors and unavailability of valid performance database for inferring the equations as well as the lack of literature in this field, the preset study includes five optimizing factors affecting the fuel consumption efficiency of a tanker in genetic algorithm by using the genetic algorithm toolbox of MATLAB software package

The sea level predictions, which are done with state-of-the-art hydrodynamic models, often suffer from various uncertainties which are related to errors in atmospheric and open boundary forcing fields as well as in physical flux parametrizations. The errors in atmospheric prediction, and regional extreme phenomena like storm surges are due to models' approximated descriptions of the physical environments (e.g coupling with waves, atmosphere, ice, runoff), but also due to the stochastic nature of weather prediction, which is often treated as a single deterministic forecast in many applications. Moreover, the submission of correct warning alerts is subject to conversion to local reference level. The application of data-driven AI models can offer a promising addition to the classic hydrodynamic models in addressing these challenges. In the present study the sea level deep-learning forecasting model HIDRA2, demonstrates promising capabilities for forecasting storm surges in numerous coastal stations across the eastern coast of the Baltic Sea. Our comprehensive assessment of HIDRA2's includes intercomparison with the sea level forecasts predicted with the regional configuration of the NEMO 4.0 hydrodynamics model. Moreover the probabilistic storm surge forecast from the ensemble sea level predictions allows us to identify and refine the best sub-set of ensembles for accurately predicting storm surges. This case study will play an important role in guiding decision-making processes regarding the integration of deep-learning methodologies into the operational phase of sea level prediction, particularly in the Baltic Sea region.The EU funds this work under the agreement DE_330_MF between ECMWF and Météo-France. The on-demand capability proposed by the Météo-France led international partnership is a key component of the weather-induced extremes digital twin, which ECMWF will deliver through different phases of Destination Earth, launched by the European Commission.