Earth and Environmental Sciences

Accurate prediction of river flows in data-scarce regions is essential for sustainable water resources management under a changing climate, yet limited observations constrain the performance and application of data-driven models. This study systematically evaluates the effectiveness of statistical and physics-based data augmentation strategies for improving the performance of Machine Learning (ML) models in river flow simulation. The evaluation was conducted in two data-scarce Sub-Saharan African catchments with contrasting climates. To address data scarcity, bootstrapping and physics-based data augmentation were applied to expand the training datasets of the Feed Forward Neural Network (FFNN) and a Long Short-Term Memory (LSTM) model. The resulting model performances were benchmarked against the physically based Hydrologic Engineering Center Hydrologic Modeling System (HEC-HMS) model. Comparison of standalone ML models with HEC-HMS reveals data-dependent performance, with HEC-HMS outperforming ML models on very limited datasets. In contrast, ML models achieve superior performance as data availability increases. Integration of ML models with data augmentation approaches indicated improved performance of FFNN and LSTM, particularly when limited training data was available. Limited or comparable performance was observed when longer training datasets were used. The results further indicated that applied data augmentation approaches were independent of model architecture and catchment hydroclimatic conditions. These findings provide practical guidance for water resource engineers and modellers on the model-specific and data-dependent applicability of data augmentation strategies for river flow modelling in data-scarce regions.

Wastewater treatment plants (WWTPs) are transitioning into water resource recovery facilities (WRRFs), where the recovery of resources from wastewater streams is increasingly prioritized. CH 4 , CO 2 , and volatile organic compounds (VOCs) represent a promising resource for generating value-added products. This study presents a comparative techno-economic review of gas valorization pathways in the transition from conventional WWTPs to WRRFs, synthesizing technical evidence and literature-reported production costs. While several biological, catalytic, electrochemical, and membrane-based routes demonstrate strong technical potential, their economic viability remains uncertain. Cost analysis shows that added-value products obtained from CH 4 and CO 2 have production costs broadly comparable to market prices, suggesting potential competitiveness, whereas those derived from VOC lack reliable cost data. The findings underscore the need for comprehensive economic evaluations that include non-market externalities and carbon benefits.

This study examines how generational differences in travel socialisation shape sustainable travel preferences for commuting by bicycle and bus and intentions to purchase a hybrid or electric car if such options were available and affordable. Drawing on survey data from 519 millennial and 870 Generation Z university students in Ghana, the findings show cohort differences in the socialisation pathways associated with sustainable travel preferences and intentions. Among Gen Z respondents, early exposure to car ownership was associated with stronger preferences for bicycles and bus use, while among millennials, growing up in carless households were linked to stronger intentions to adopt hybrid or electric cars. Several socialisation variables significantly predicted cycling, bus use, and hybrid/electric car adoption, underscoring the enduring influence of childhood and adolescent experiences in shaping sustainable travel preferences and intentions. These findings extend travel socialisation theory by demonstrating how transport mode norms and dispositions formed early in life shape sustainable travel preferences and intentions in young adulthood. Beyond its theoretical contributions, this study offers significant practical implications for a diverse range of stakeholders. Bus service providers could combine subsidised or fare-free schemes for students with improvements in service frequency and network connectivity, while car retailers may leverage cohort-specific strategies to accelerate electric car adoption. Furthermore, universities can encourage low-carbon commuting through targeted institutional practices. Ultimately, policymakers can better align attitudinal foundations with infrastructure development and transport innovation to advance sustainable urban mobility.

We examined dose–response relations between a Japan Postcode-level Walkability Index (JPWI) and total walking, walking for leisure; and walking for transport, among adults living in 95 cities across the country. This nationwide cross-sectional study included 40,286 adults (males 50.5%, 49.4 ± 15.6 years) residing in a wide range of urban and rural settings, who responded to a web-based survey conducted in 2021 using quota sampling. The JPWI was calculated by generating a 1000 m road network buffer from a representative point within each postcode boundary using the mean of z-scores summarizing the population density, intersection density and number of destinations within this buffer. JPWI scores ranged from −1.848 to 2.590. There was a non-linear positive association between JPWI and total walking, with an inflection point at a JPWI of 0.78 (95% CI: 0.69, 0.87), below which the association with walking time was minimal (B = 22 [11,32]), but above which walking time increased significantly (B = 118 [106, 129]). Similar patterns of associations were observed in both leisure walking and walking for transport, with inflection points at JPWI values of 0.78 (95% CI: 0.69, 0.86) and 0.64 (95% CI: 0.42, 0.86), respectively. Findings highlight the potential utility of considering how the JPWI (and potentially related measures for other countries) of local areas compare with such empirically-identified inflection points when developing strategies for improving the built environment to promote physical activity.

Abstract Seagrasses perform various critical functions within social–ecological systems, though populations have declined drastically. Restoring seagrass is crucial, requiring holistic consideration of ecological and human dimensions. To date, biophysical factors predominantly guide restoration and marine spatial planning (MSP). Sociocultural data underpin optimal and just restoration design, but remain a “missing layer” in MSP. To address this gap, we combined participatory mapping and ground-truthed remote sensing to map seagrass distribution, human activities, and place values in Sanday, UK. Local knowledge guided in situ surveys, co-producing Sanday’s first scientific seagrass records. We “found” two meadows absent from current seagrass modelling, mostly within protected areas, and highlighted corresponding human activities (notably shellfisheries) and values (humanistic and scientific). We propose three restoration priority sites, given seagrass extent, habitat protection, and compatible activities and values. This study demonstrates potential for transdisciplinary mapping to improve restoration planning, addressing the missing sociocultural layer and key gaps within MSP processes.

Anthropogenic factors have negatively impacted the health of freshwater ecosystems by increasing the occurrence of harmful cyanobacterial blooms and their associated toxins. Water column mixing and reworking of substrate by organisms allow for cyanotoxin deposition into sediments and resuspension into the water column. Crayfish rework benthic substrates and are exposed to toxins in both benthic and pelagic phases. Since crayfish are keystone species and ecosystem engineers, understanding the spatial dynamics and effects of aquatic toxins on crayfish is necessary. In this study, crayfish were exposed to low concentrations of microcystin-LR (MC-LR) present in the water column or sediment of a closed system for 108 h. After exposure, crayfish were placed in flow-through mesocosms to monitor foraging and locomotive behaviors and were subsequently dissected to assess membrane integrity of hepatopancreas cells. The presence of EtOH or EtOH/MC-LR in the water column decreased crayfish foraging compared to the control. Crayfish locomotive behaviors increased when the water column was dosed with EtOH or EtOH/MC-LR and when the sediment was dosed with EtOH/MC-LR compared to the control. Non-significant trends suggest that foraging was impaired when MC-LR was present in the water column vs the sediment and that locomotion was enhanced when MC-LR was present in the water column vs the sediment. Additionally, MC-LR present in the water column significantly lowered fluorescein diacetate (FDA) fluorescence per crayfish hepatopancreas cell. A non-significant trend also suggests that the amount of hepatopancreas cells detected was lower when MC-LR was present in the water column vs the sediment. MC-LR was observed to negatively impact crayfish hepatopancreases membrane integrity which could explain the resulting behavioral changes. These findings have implications for crayfish health and for overall ecosystem functioning in environments known for the occurrence of MC-LR.

Abstract Projections of future changes in tropical cyclone (TC) rainfall are critical for understanding evolving flood risks and infrastructure impacts under climate change. This study couples a physics-based Tropical Cyclone Rainfall (TCR) model with the Columbia HAZard (CHAZ) model, a statistical–dynamical TC downscaling framework, which generates synthetic storm tracks and intensity, to produce large synthetic ensembles of TC generated rainfall downscaled from 12 CMIP6 models. TC rainfall is simulated in the North Atlantic basin based on the CMIP6 historical and future simulations under three Shared Socioeconomic Pathways (SSP2-4.5, SSP3-7.0, SSP5-8.5). Projections from this integrated framework are evaluated against the Geophysical Fluid Dynamics Laboratory (GFDL) Rainfall Climatology and Persistence model (R-CLIPER), a statistical TC rainfall model that provides an alternative to the dynamical simulations. TCR projects widespread increases in TC rainfall, with the strongest changes along the US East and Gulf coasts and for major hurricanes (Categories 3–5). By late century, average rainfall increases reach 40–100% in TCR compared to 30–60% in R-CLIPER, with the largest increases in the Northeast regions. Extreme 24-hour rainfall increases by up to 55% in TCR, roughly twice the magnitude simulated by R-CLIPER. Key drivers include higher TC intensities, increased atmospheric moisture, and projected slower translation speeds, which together contribute to increasing flood risks, especially when combined with more frequent sequential TC events.

Abstract. Atmospheric blocking is a key driver of midlatitude weather extremes, including heatwaves and cold spells. Yet general circulation models (GCMs) struggle to capture the frequency, persistence, and spatial characteristics of blocking. Here, we evaluate atmospheric blocking in next-generation storm-resolving Earth system models from the nextGEMS, EERIE, and DestinE projects, focusing on ICON and IFS-FESOM with ∼ 10 km atmospheric and ∼ 5 km ocean grid spacing. We also provide first insights into the IFS-FESOM under SSP3-7.0 forcing. Blocking frequency, duration, and size are assessed in historical simulations spanning 30 years for IFS and 27 years for ICON, relative to ERA5 reanalysis and a CMIP6 multi-model ensemble of eight models. We further examine links between blocking biases and the background flow, sea surface temperatures (SSTs), and storm-track activity. In the CMIP6 ensemble, persistent biases in blocking frequency, duration, and spatial extent are evident, particularly over the Euro-Atlantic sector, consistent with previous studies. Several of these biases persist in the storm-resolving coupled simulations or are even amplified, indicating that increased horizontal resolution alone does not systematically improve blocking representation. Among the storm-resolving models, performance varies regionally and seasonally. ICON exhibits larger winter biases, including overly zonal jets and an underestimation of Euro-Atlantic blocking compared to IFS. The coupled IFS configuration shows intermediate performance, reproducing some aspects of blocking variability but retaining substantial biases associated with SST errors and jet structure. In contrast, the atmosphere-only IFS simulation (IFS AMIP), which is forced with observed SSTs, reproduces blocking frequency and jet structure more realistically over both the North Atlantic and North Pacific. This highlights the strong sensitivity of blocking to sea surface temperatures and ocean–atmosphere coupling, and underscores the importance of realistic SST boundary conditions for improving blocking representation. Under SSP3-7.0 forcing, IFS projects reduced winter blocking at high latitudes (e.g., northern Europe) and reduced summer blocking frequency over the North Atlantic, northern Europe, and Russia. Changes in magnitude, spatial pattern, and persistence are often of the same order as the model biases, indicating that projected blocking responses are difficult to disentangle from systematic errors related to jet structure, SST biases, and storm-track activity. Overall, storm-resolving models show local improvements in blocking representation, particularly when forced with realistic SSTs. However, coupled simulations still exhibit large biases, underlining the need for further development of ocean–atmosphere coupling representation. These findings highlight both the potential and the current limitations of storm-resolving models for simulating and projecting persistent weather extremes in a warming climate.