Adolf Stips

While marine policies increasingly demand functional ecosystem metrics like the Trophic Transfer Index (TTI) to assess eutrophication, basin-wide biological monitoring remains logistically prohibitive. This study applies artificial neural networks (ANNs) to deduce continuous TTI and Good Environmental Status (GES) for the Baltic Sea directly from standard environmental variables. Using 70 years of data, we compared feedforward (MLP) and recurrent (LSTM) architectures. The LSTM excelled at spatial extrapolation (accuracy: 0.89) by acting as a non-linear spatial regulariser. Conversely, the MLP proved superior for temporal forecasting (accuracy: 0.8), adapting faster to recent ecological recoveries where the LSTM struggled with historical inertia. Explainable AI (SHAP) revealed that bathymetry and water transparency dominated predictive logic, outweighing raw ambient nutrients. Finally, a simulated Remote Sensing scenario using satellite-retrievable proxies achieved high spatial accuracy (0.85), establishing ANNs as a scalable framework for cost-effective, spaceborne eutrophication monitoring.

The ecological status of the two nautical mile (2 nm) coastal waters of Bulgaria (BG) and Romania (RO) is evaluated in terms of eutrophication, focusing on various environmental and chemical indicators related to nutrient enrichment and its effects. Applying the Blue2 Modelling Framework (Blue2MF) Black Sea model, we simulate diverse combinations of nitrogen (N) and phosphorus (P) reduction. By integrating atmospheric and marine conditions, hydrology and environmental dynamics, the model provides a comprehensive framework for analysing the impact of external pressures on the marine environment and can be used to set regional sustainability goals. Each scenario is assessed with respect to environmental impact, such as enhanced water quality and potential alterations in phytoplankton communities. Furthermore, we establish integrated nutrient criteria for eutrophication control, concentrating on the management of both N and P inputs while maintaining the current ratio between them. The model results indicate that a 20% reduction in both N and P from European Union (EU) rivers would result in about a 24% decrease in N within the BG and RO 2 nm coastal waters, whilst sustaining the N:P ratio across various spatial scales—from river inputs to coastal and offshore zones. This approach is valuable for assessing the potential impacts of different nutrient reduction strategies, aiding the effective management of marine ecosystems to address eutrophication challenges.

Coastal ecosystems worldwide suffer from severe eutrophication, yet many fail to recover despite stringent nutrient reduction policies. A fundamental challenge in marine management is distinguishing whether this delayed recovery is a simple temporal lag or the lock-in of an alternative stable state driven by ecological memory. Using the Baltic Sea as a multi-decadal prototype, we apply a 69-year (1954–2022) coupled biogeochemical dataset to isolate the mechanisms decoupling external nutrient forcing from internal ecosystem responses. Structural analysis reveals highly asynchronous regime shifts: five external loading shifts triggered only three biological shifts, demonstrating initial ecological buffering and subsequent lock-in. Causally constrained correlation analysis identifies that deep-basin biogeochemistry lags external load reductions by 14 to 19 years. While shallow coastal margins respond reversibly, phase-space trajectories and multidimensional Information Flow Rate analyses prove that deep basins (≥ 50 m) are trapped in an alternative stable state. Here, the benthic legacy dissolved inorganic phosphorus reservoir supersedes contemporary external nutrient forcing as the primary causal hub controlling phytoplankton biomass and primary production. These findings mathematically demonstrate that achieving systemic recovery in stratified coastal seas remains bottlenecked by deep-water legacy effects, necessitating spatially differentiated environmental targets and the recalibration of global regulatory timelines.

Stips et al (2016) demonstrated the existing causal relationship between Green House Gases (GHG) concentrations and Global Mean Surface Temperature (GMTA) based on the Information Flow (IF) methodology. Critics on the application of the Information Flow concept as developed by Liang (2008, 2016) has focused on the underlying assumption of uncorrelated residuals (noise) between the time series. However, this assumption can only make sense for a system with two components, as for a multi-dimensional system unobserved noise may well exist. Fundamentally, there can be no such thing like correlated noise at all. It can seemingly only appear because of some hidden process(es). For investigating this in detail a multivariate information flow analysis has been developed. We will show that in our tests using processes with correlated noises, the preset causalities can be well reproduced. Further, it will be demonstrated that reducing autocorrelation within the time series by pre-whitening, confirms the achieved causality directions. Finally, we question the validity of the proposed alternative measure using forecast error variance decomposition based on vector autoregression by Goulet and Goebel (2021), because in their method causal directions can be simply reversed by reordering. A physically faithful causal measure should be generally independent of ordering. Coulombe, P. G. and Goebel, M. 2021. On Spurious Causality, CO2, and Global Temperature. Econometrics, 9(3), 33.Liang, X. S. 2008. Information Flow within Stochastic Dynamical System. Phys. Rev. E 78: 031113.Liang, X. S. 2016. Information Flow and Causality as rigorous Notions ab initio. Physical Review E 94: 05220.Stips, A., D. Macias, C. Coughlan, E. Garcia-Gorriz, and X. S. Liang. 2016. On the Causal Structure between CO2 and Global Temperature. Scientific Reports 6: 21691.

This study investigates the extension of the potential habitat of Harmful Algal Blooms (HABs) in the North Sea using observational data and model experiments under current and future climate scenarios. We assess the combined effects of temperature, salinity, and nutrient availability, particularly the nitrogen-to-phosphorus ratio, on HABs in the region. Climate change projections indicate a decrease in salinity concomitant with an increase in surface temperature, potentially leading to an offshore extension of HAB habitat. Reducing nitrogen and phosphorus loads in rivers differentially affects dissolved inorganic nitrogen (DIN) and phosphorus (DIP) levels, with DIN being more sensitive to load reduction, thereby constraining HAB habitat extension. We underscore the importance of considering both physical and biogeochemical factors in assessing HAB habitat dynamics and the potential impacts of climate change and nutrient reduction measures on HAB expansion in the North Sea. These findings have significant implications for environmental policy and management.

Marine plastic floating on the sea surface is a major environmental problem. The present studyinvestigated the potential transport patterns of floating marine litter and areas of itsaccumulation in the Baltic Sea by using a hydrodynamic model coupled with a particle-trackingmodel (Parcels). Mapping of marine plastic debris distribution in 2017–2018 revealed that the largestplastic accumulation area is between latitude 59° and 61°, which includes the Northern BalticProper, Archipelago Sea, and the Gulf of Finland. The floating plastic spreads from the largestplastic pollution site River Vistula to the waters of all the countries around the Baltic Sea.

We explore the patterns of Black Sea phytoplankton growth as driven by the thermohaline structure and circulation system and the freshwater nutrient loads. Seasonal and inter-annual variability of the phytoplankton blooms is examined using hydrodynamic simulations that resolve mesoscale eddies and online coupled bio-geochemical model. This study suggests that the bloom seasonality is homogeneous across geographic locations of the Black Sea inner basin, with the strongest bloom occurring in winter (February–March), followed by weaker bloom in spring (April–May), summer deep biomass maximum (DBM) (June–September) and a final bloom in autumn (October–November). The winter phytoplankton bloom relies on vertical mixing of nitrate from the intermediate layers, where nitrate is abundant. The winter bloom is highly dependent on the strength of the cold intermediate layers (CIL), while spring/summer blooms take advantage of the CIL weakness. The maximum phytoplankton transport across the North Western Shelf (NWS) break occurs in September, prior to the basin interior autumn bloom. Bloom initiation in early autumn is associated with the spreading of NWS waters, which in turn is caused by an increase in mesoscale eddy activity in late summer months. In summary, the intrusion of low salinity and nitrate-rich water into the basin interior triggers erosion of the thermocline, resulting in vertical nitrate uplifting. The seasonal phytoplankton succession is strongly influenced by the recent CIL disintegration and amplification of the Black Sea circulation, which may alter the natural Black Sea nitrate dynamics, with subsequent effects on phytoplankton and in turn on all marine life.