Coastal and estuarine waters are among the most productive and valuable ecosystems on Earth, sustaining biodiversity, maintaining ecological balance, and supporting a wide range of socio-economic activities. However, their location at the land–sea interface also makes them particularly vulnerable to a wide array of natural and anthropogenic stressors.

Climate change has intensified this vulnerability through warming, sea-level rise, changes in freshwater inflows and salinity regimes, more frequent and intense extreme events, and growing pressures from pollution and coastal development. Together, these drivers pose significant risks to the structure, functioning, and resilience of these ecosystems. Understanding how physical, biogeochemical, and ecological processes interact across multiple scales is therefore essential. This understanding, together with the socio-economic drivers of vulnerability, enables assessing present conditions, anticipating future changes, conducting comprehensive risk assessments, and supporting effective management and adaptation strategies.

Numerical modeling has become an indispensable tool for addressing these challenges. By enabling the simulation of hydrodynamic, biogeochemical, and ecological processes under diverse conditions, models allow researchers to analyze historical events where observations are limited, assess current system dynamics, and explore potential future scenarios, supporting strategic planning and informed decision-making. Advances in computing, numerical methods, data assimilation, coupled physical-ecological modeling, and emerging artificial intelligence techniques have improved model resolution and reliability. This enables more accurate simulation and better prediction of stressor impacts and extremes and supports both ecosystem-based and engineering responses.

Accordingly, this Research Topic promotes innovation in coastal and estuarine modeling, highlighting five contributions that span uncertainty quantification, high-resolution ecosystem modeling, socio-economic risk assessment, habitat-mediated mitigation, and optimization of coastal engineering solutions. Together, they demonstrate how integrative approaches can deliver actionable insights for the science, management, and decision-making of complex coastal and estuarine systems.

Magel et al. use a coupled hydrodynamic–biogeochemical–seagrass model to test how eelgrass (Zostera marina) influences ocean acidification and hypoxia in Coos Bay (Oregon) under three habitat scenarios. Eelgrass increased variability in pH and dissolved oxygen, improving agreement with observations. Overall, it more consistently mitigated acidification (higher pH and aragonite saturation relevant to oysters) than hypoxia, with mixed oxygen responses that could slightly increase low-DO exposure relative to a salmon threshold. The model is proposed as a decision-support tool for targeting eelgrass restoration to maximize resilience.

Roh et al. use the non-hydrostatic NHWAVE model to investigate how submerged breakwaters affect coastal hydrodynamics and shoreline evolution. By simulating different wave conditions and structural configurations, including offshore position and vertical crest distance, the study quantifies wave attenuation, nearshore flow currents, and shoreline response, including erosion and accretion patterns. Results show that breaking-induced currents and vortex flow strongly influence shoreline response, and that breakwater dimensions critically determine wave dissipation and sediment transport. These findings provide valuable insights for optimizing coastal protection measures and illustrate how numerical modeling can guide engineering solutions to mitigate shoreline erosion while maintaining ecosystem functionality.

In the Elbe estuary (northern Germany), phytoplankton concentrations drop sharply by about 90% as river water enters the deep, highly turbid shipping channels around the Port of Hamburg, with major implications for the food web and carbon cycling, as the system can shift from net autotrophy to heterotrophy. While earlier work largely attributed this “phytoplankton collapse” to zooplankton grazing, Steidle et al. propose that coagulation/aggregation of phytoplankton with inorganic suspended sediments is a key alternative mechanism. They develop a novel individual-based, Lagrangian modeling framework that couples hydrodynamics, sediment transport, and biogeochemistry to represent aggregation-driven sinking and its ecological consequences. The model indicates that aggregation can push phytoplankton into deeper, darker waters where light limitation drives high mortality, suggesting that more than 80% of phytoplankton larger than 50 µm may be lost this way. The predicted mortality hotspots also coincide with zones of organic-matter remineralization, reinforcing the idea that physical–biogeochemical interactions in engineered, turbid estuaries can strongly control plankton survival and ecosystem metabolism.

Rosli et al. present a modeling-based assessment of climate change-driven storm surge hazards applied to Southeast Asia, particularly along the east coast of Peninsular Malaysia facing the South China Sea. Statistically calibrated data from the d4PDF climate dataset are selected to force the high-resolution hydrodynamic model MIKE 21-FM HD, enabling the simulation of storm surge heights under different global warming scenarios. The results reveal a progressive increase in extreme storm surge levels across key coastal locations as climate change scenarios become more severe. By coupling these projections with historical flood loss data and national budget allocations, the study identifies a critical mismatch between escalating storm surge risks and current mitigation investments. In this sense, the findings underscore the need for enhanced regional forecasting and response systems, standardized resilient infrastructure guidelines, and AI-supported community-based risk mapping as key components to strengthen coastal disaster management and resilience planning.

Fanous et al. explore how to perform fast, reliable uncertainty quantification for mangrove hydro-morphodynamic models where full physics simulations are computationally expensive. They built a scalable probabilistic surrogate based on Deep Gaussian Processes combined with Bayesian GPLVM dimensionality reduction and variational inference, and applying it to a high-resolution case study of the Sundarbans (Bay of Bengal) with water-surface elevation as the main quantity of interest. Using leave-one-out validation across time steps, the deep GP emulator reproduced the numerical model with low error and produced spatially explicit uncertainty estimates (highest near complex features like channel bifurcations/shore–vegetation interfaces), while outperforming a standard single-layer GP. Crucially, the surrogate reduced computation from multi-day runtimes for 24-hour simulations to about ~1 min 43 s end-to-end (reported as >3 orders of magnitude faster), enabling rapid scenario testing and uncertainty-aware decision support for nature-based coastal resilience planning in dynamic mangrove environments.

In summary, these studies apply innovative modeling approaches, including non-hydrostatic and high-resolution hydrodynamic models, Lagrangian methods, coupled hydrodynamic–biogeochemical–seagrass frameworks, and mangrove hydro-morphodynamic models to address key questions in coastal and estuarine science. Together, they advance our ability to understand, monitor, and predict coastal processes, while improving the assessment and mitigation of stressor impacts and extreme events in these ecosystems. As a result, their findings should be of broad interest to researchers and coastal managers worldwide.