Atmospheric and Oceanic Sciences

Abstract The spatial distribution of suspended sediment load is a key indicator for watershed management by identifying primary sediment source areas and elucidating transport‐deposition patterns. However, clarifying such spatial patterns in alpine regions is challenging due to strong sediment heterogeneity and sparse hydrological monitoring data. Here, we focus on the Yarlung Tsangpo River and apply a global sediment simulator model to estimate mean annual suspended sediment loads at 17 cross‐sections along the mainstream and its tributaries. The model demonstrates good applicability in this region, with a high coefficient of determination ( R 2 = 0.97) between predicted and observed values. The sediment load exhibits significant spatial heterogeneity with a coefficient of variation (CV) of 1.31. A generally downstream increasing trend is observed, whereas the lower reach functions as a ‘sediment factory’ of the basin, contributing 92.9% of the sediment load (SL) at the basin outlet, with an average specific sediment yield (SSY) of 4495 t/km 2 and a local maximum of 18 629 t/km 2 . In contrast, the middle reach contributes only 5.9% of the sediment load and exhibits a high modern deposition rate of 81 Mt/a, far exceeding the sediment export of 14.36 Mt/a. Transport and deposition zones alternate spatially, with deposition predominantly occurring at tributary confluences and upstream of riverbed knickpoints. These insights advance the understanding of sediment production and transport mechanisms in complex alpine river systems and support informed water resource development.

The availability of reliable ground-truth data is one of the main bottlenecks for improving high-resolution forest attribute maps from Earth observation data. This is underpinned by the European Union (EU) Forest Strategy for 2030 that underscores the need for harmonized, cross-border forest resource assessments that integrate both remote sensing and field-based National Forest Inventory (NFI) data. However, confidentiality constraints on NFI plot coordinates present a significant barrier to aligning these datasets, thereby limiting the development of unified forest monitoring systems that can fully leverage the potential of Earth Observation data. To overcome these data-sharing limitations we explored the effectiveness of a privacy-enhancing technique, known as Federated Learning (FL), that is a form of distributed computing aimed at preserving the privacy and confidentiality of data owned by different organizations. This methodology has been tested for the collaborative modelling and mapping of forest timber volume across four European countries: Norway, Sweden, Finland, and Italy. We employed a time-series convolutional neural network (CNN) architecture tailored to integrate 40 years of Landsat or 7 years of Sentinel imagery and terrain variables with harmonized NFI data from more than 85,000 sample plots. This model architecture was used for the FL approach and compared to traditional country-specific and centralized modelling strategies. FL models achieved predictive performances comparable to the traditional models, which proofs the effectiveness of the proposed approach. Centralized or global models showed slightly reduced performance compared to the national models, highlighting the value of fine-tuning with local ground-truth data. By aligning with the EU’s forest monitoring objectives, FL facilitates the generation of harmonized models and maps of forest features, like timber volume and biomass, that are critical to support evidence-based forest policy and management. The findings underscore the potential of FL to transform collaborative environmental monitoring, particularly in domains where data confidentiality and interoperability are critical.

Urban rivers in megacities serve as major sinks for diverse emerging contaminants (ECs). However, a comprehensive understanding of their sources and dynamics remains limited. This study employs an integrated targeted and nontargeted analysis to investigate ECs in the Wenyu River basin of Beijing and its associated sewage treatment plant (STP) effluents. We identified/tentatively identified and quantified 480 ECs, with household chemicals dominating the contaminant profile. Hierarchical cluster analysis delineated five distinct anthropogenic source types: influenza prevention compounds, daily lifestyle chemicals, sunscreen/insecticide residues, rural diffuse inputs, and traffic/waterfront landscape emissions. Results indicate that STP effluents constitute the primary intermediary for the first three categories, exerting a dominant influence on riverine EC levels, demonstrated by path coefficients of 0.736 in spring and 0.585 in summer. Seasonal variations revealed clear patterns: influenza- and sunscreen-related sources showed pronounced fluctuations corresponding to flu season and sunlight exposure periods, whereas traffic-landscape sources peaked during summer rainfall events. Ecological risk assessment indicated that 91.4% of the total environmental risk originated from STP-derived ECs, with telmisartan and fipronil identified as the compounds posing the highest ecological risk. This study provides new insights into the source-specific dynamics and seasonal drivers of ECs in urban rivers, offering a scientific basis for targeted pollution control strategies.

Anaerobic digestion effluent (ADE) is an attractive feedstock for resource recovery due to its high concentrations of ammonia (NH 3 ) and phosphate (PO 4 3– ). However, the presence of elevated bicarbonate (HCO 3 – ) imparts strong buffering capacity, which inhibits PO 4 3– precipitation and increases chemical demand for NH 3 recovery. To overcome HCO 3 – interference, this study proposes an enzyme-based pretreatment that converts HCO 3 – into PO 4 3– using phosphoenolpyruvate carboxylase (PEPC). PEPC was immobilized onto carboxylated polystyrene beads via EDC/NHS coupling (PB@PEPC) to enhance stability and reusability. When applied to real ADE, 25 PB@PEPC beads achieved a 2.94-fold reduction in HCO 3 – concentration (from 6013.8 ± 339.8 mg·L –1 to 2045.5 ± 283.5 mg·L –1 ) within 12 h, while simultaneously increasing PO 4 3– concentration 10.79-fold (from 590.8 ± 148.3 mg·L –1 to 6363.3 ± 383.5 mg·L –1 ). The enriched PO 4 3– was recovered as calcium phosphate with efficiencies exceeding 95% using Ca(OH) 2 at a Ca/P molar ratio of 2.5. The elevated pH during PO 4 3– precipitation promoted NH 4 + conversion to volatile NH 3, enabling up to 82.4 ± 6.1% NH 3 recovery using a membrane contactor without additional alkaline reagents. PB@PEPC retained 86.1 ± 4.2% of its initial activity after five reuse cycles. Overall, this integrated process offers a sustainable pathway for simultaneous nutrient recovery from complex wastewater.

Sulfachloropyridazine (SCP) is a sulfonamide antibiotic widely detected in aquatic environments. Its terminal pyridazine ring contains vicinal pyridinic nitrogen atoms (=N–N=) with strong metal-complexing ability, which may cause distinct interactions with metal-oxide filter media. Here, we systematically investigated the removal of sulfonamide antibiotics by natural manganese sand (NMS), a representative filtration medium in water treatment. NMS exhibited pronounced selectivity toward SCP, while showing negligible removal of sulfadiazine (SD) and sulfamethoxazole (SMX). Under optimal conditions (NMS = 5 g·L−1, pH 3), 99.38% of SCP (5 mg·L−1) was removed within 6 h; the TOC reduction only reached 42.65%, suggesting the partial transformation of SCP. Mechanistic evidence suggests that the vicinal pyridinic N–N motif of SCP provides dual electron-donating sites, enabling inner-sphere Mn–N complexation on NMS. This coordination-driven adsorption is strongly pH-dependent and is inhibited under neutral to alkaline conditions (pH > 5) due to electrostatic repulsion. After selective binding, surface Mn(III)/Mn(IV) species can act as electron acceptors, driving the surface oxidation of SCP. Although NMS induced mild oxidative transformation of SCP, some degradation products still exhibited potential ecotoxicity and therefore require further attention in practical water treatment applications. These findings link terminal functional-group structure to selective abiotic removal on Mn-oxide media and inform targeted control of sulfonamide micropollutants in filtration-based water treatment.

This study numerically investigated the global sloshing flows in a partially filled horizontal circular tank subjected to harmonic excitations. The computational model adopted the geometry specified in the fifth Collaborative Computational Project in Wave Structure Interaction blind test, considering horizontal and vertical excitations. Commercial computational fluid dynamics software employing the volume-of-fluid method was used to resolve complex free-surface deformations. Furthermore, the grid uncertainty was assessed to ensure numerical reliability. The numerical results were validated against experimental data by comparing the wave elevations measured at the left and right sidewalls. Comprehensive analyses were conducted at various excitation frequencies and amplitudes using time-series data, phase-space trajectories, Poincaré sections, and a fast Fourier transform to characterize the dynamical responses of the sloshing flows. Under horizontal excitations, distinct nonlinear characteristics, including softening and jump phenomena, were observed. The phase-space analysis and frequency-sweep simulations revealed path-dependent hysteresis and coexistence of multiple stable states, thereby confirming the presence of subcritical bifurcations. Under vertical excitations, the flows exhibited typical Faraday wave patterns, thereby showing nonlinear characteristics driven by parametric resonance. The instabilities were identified using the Mathieu equation and Strutt stability diagrams, which established the boundaries between the stable and unstable sloshing motions.

Savitzky-Golay (SG) filtering can mitigate the ‘salt-and-pepper’ effect in pixel-wise classifications of hyperspectral imagery, but its optimal parametrization across diverse forest types remains unclear. We evaluated 81 SG parameter combinations (Window: 3–19, Order: 1–17) on species-level classifications across three structurally distinct Hungarian forests (Gemenc, Mecsek, Ságvár) using a Bayesian hierarchical model to quantify the effects of forest site, tree species, their interaction, and SG parameters on overall accuracy. Our analysis revealed three principal findings. First, forest site was the primary driver of SG performance, while individual species showed no significant effect; only one forest-species interaction ( Mecsek::pedunculate oak ) proved significant. Second, 72 of 81 parameter combinations were statistically equivalent, indicating substantial parameter flexibility. Third, the combination Window 5 and Order 1 (b_5::1) emerged as the most robust performer, while excessive variance reduction systematically degraded accuracy, particularly in the most diverse forest (Mecsek). We conclude that SG filter performance is governed by forest-level structural and community characteristics rather than species identity, and the wide equivalence of most parameters offers practical flexibility, with b_5::1 recommended as a robust default. Beyond these specific findings, this study demonstrates the value of adopting a hierarchical modelling framework-here, the Bayesian Hierarchical Model-for evaluating pre-processing techniques, offering an adaptable analytical template for future research across diverse remote sensing applications.

The seafloor is recognized as a major sink for marine debris, while deep-sea litter remains poorly investigated due to the logistical and economic constraints associated with seabed exploration. This study reports for the first time the occurrence of marine debris found on the Argentine deep seafloor using the SOI’s Remotely Operated Vehicle (ROV), providing the first documented evidence for this region. A total of 29 litter items were recorded across 55.6 km of surveyed seafloor. The highest debris abundance was recorded in submarine canyon areas (1.4 items/km), whereas the Malvinas Basin showed the lowest levels (0.1 items/km), likely associated with hydrodynamic conditions. In addition, plastic items, fishing ropes, and nets were the most frequently recorded types of debris. These findings establish a baseline of the current status of deep-sea marine debris and propose future directions for the region.

Abstract Urban weather station networks (WSNs) are widely used to monitor urban weather and climate patterns and aid urban planning. However, maintaining WSNs is expensive and labor-intensive. Here, we present a step-wise station removal procedure to thin an existing WSN in Freiburg, Germany, and analyze the ability of WSN subsets to reproduce air temperature and humidity patterns of the entire original WSN for a year following a simulated reduction of WSN density. We found that substantial reductions in station numbers after one year of full deployment are possible while retaining high predictive accuracy. A reduction from 42 to 4 stations, for instance, increased mean prediction RMSEs from 0.69 K to 0.83 K for air temperature and from 3.8% to 4.4% for relative humidity, corresponding to RMSE increases of only 20% and 16%, respectively. Predictive accuracy is worse for remote stations in forests than for stations in built-up or open settings, but better than a state-of-the-art numerical urban land-surface model (Surface Urban Energy and Water Balance Scheme). Stations located at the edges between built-up and rural areas are most valuable when reconstructing city-wide climate characteristics. Our study demonstrates the potential of thinning WSNs to maximize the efficient allocation of financial and personnel-related resources in urban climate research.

Abstract Urbanization and climate change have increased the frequency and severity of stormwater-related challenges, necessitating more effective approaches for selecting green infrastructure (GI) during the planning and design process. Although numerous studies have evaluated the environmental performance of individual GI practices, comparatively limited attention has been devoted to translating performance assessments into practical guidance for selecting and prioritizing GI practices before implementation. To address this gap, this research developed a pre-design GI prioritization framework by integrating local sensitivity analyses with three deterministic performance models. The framework quantified runoff mitigation, ecosystem-service benefits, and 20-year life-cycle costs to establish hierarchical rankings of GI practices under alternative decision-making scenarios. The framework was applied to a 266-ha redevelopment scenario in downtown Lansing, Michigan, USA, and subsequently evaluated through post-design performance prediction. Results consistently identified rain gardens, street planters, infiltration basins, wetlands, and vegetation filter strips as the highest-priority practices under the balanced environmental-economic scenario. Implementation of the resulting GI plan reduced annual runoff volume by 52%, decreased pollutant loading by 9%–15%, and generated substantial carbon sequestration benefits. The study demonstrates how sensitivity-analysis results derived from multiple performance models can be synthesized into a practical decision-support framework for pre-design GI prioritization, providing a transferable method for informing GI planning and resource-allocation decisions.

Abstract Exposure to ionizing radiation from galactic cosmic rays and solar energetic particles at aviation flight altitudes can have an adverse effect on human health. Although airline crews are classified as radiation workers by the International Commission on Radiological Protection (ICRP) , in most countries, their level of exposure is unquantified and undocumented throughout the duration of their career. As such, there is a need to assess aviation crew ionizing radiation exposure. The Nowcast of Aerospace Ionizing RAdiation System (NAIRAS) is a real‐time, global, physics‐based model currently used to assess such exposure. To evaluate recent model updates, radiation measurements from the Automated Radiation Measurements for Aerospace Safety (ARMAS) flight inventory are utilized. The inventory contains 1,324 flights at typical commercial aviation cruising altitudes during the most recent solar cycle, and covers latitudes 85°N to 85°S. Overall, while NAIRAS dose rates are uniformly biased high relative to the ARMAS dosimeter measurements, there is no bias relative to solar activity. Inspection of the model differences suggests that the TS05 magnetospheric magnetic field model may produce cutoff rigidities that are too low during quiet geomagnetic periods, particularly at mid‐to‐low latitudes, resulting in higher dose rates in NAIRAS. The difference in the altitude unit is also a source of error; the ARMAS flight database reports aircraft altitudes in GPS coordinates, while barometric pressure altitudes are required for NAIRAS. Future aircraft flights are being designed to constrain the uncertainties in the dosimetric measurements and model calculations.

Abstract El Niño–Southern Oscillation (ENSO) precipitation diversity exhibits pronounced phase-dependent asymmetry over the tropical Pacific. Observations indicate enhanced precipitation diversity over the western North Pacific during La Niña and over the equatorial central Pacific during El Niño. This asymmetry is associated with the positively skewed distribution and intensified extremes of tropical precipitation, implying that diversity tends to increase during ENSO phases with enhanced mean precipitation. Atmosphere-only simulations reasonably reproduce the observed asymmetric features of ENSO-related precipitation diversity, whereas fully coupled simulations underestimate its amplitude and shift the center westward. Inter-model comparisons further demonstrate weak consistency between the two simulations, with atmosphere-only experiments generally reproducing stronger precipitation diversity, suggesting an important role of SST forcing in determining the asymmetric ENSO precipitation diversity across tropical Pacific regions.

Abstract COP28 marked the first time that climate negotiations at the highest level, i.e. an official Conference of the Parties’ (COP) decision, indicated clear energy-related milestones for urgent climate action by adopting specific renewable energy and energy efficiency targets for 2030, and phasing down fossil fuels by 2050. These milestones are meant to serve as yardsticks to guide the path towards mitigating climate change and meeting the Paris Agreement (PA) goals. We assess the alignment of COP28 energy outcomes with the PA goals using several leading Integrated Assessment Models (IAMs) featuring high technological, sectoral, and regional detail. This detailed assessment could inform country- and technology-specific target design, which is becoming increasingly important in a fragmented world and under growing uncertainty regarding the prioritization of climate protection in governmental agendas. We additionally use and quantify novel transformation narratives such as “electrification” versus “combustion”, which could help develop a new generation of demand-side COP mitigation targets.

Accurate long-term global urban monitoring is essential for understanding socio-economic dynamics. To overcome limitations of single-source data at macroscopic scales, this study proposes an automated framework that fuses nighttime light (VIIRS-DNB) and vegetation indices (MODIS NDVI). We develop the Nightlight-Vegetation Rectified Index (NVRI) to suppress blooming effects in vegetated regions and amplify urban–rural contrasts. An adaptive arid grid identification mechanism further mitigates the desert blooming issue. A local adaptive thresholding strategy based on multi-scale grids and the Kneedle algorithm replaces conventional global thresholds. With spatiotemporal post-processing, we generate an annual global urban extent product (NTL Urban, 2012–2024). National-scale validation shows high agreement with MGUP and MODIS urban products. Urban extents also exhibit stronger linear correlations with GDP and population than traditional optical products, highlighting their value for macro socio-economic analysis. Despite residual limitations from blooming effects, lighting preferences, and sensor resolution, our framework offers a reproducible, scalable approach for global urban monitoring and NTL-based socio-economic research.

Urban canopy schemes are essential for urban climate modeling, yet their performance depends on the level of detail in Urban Canopy Parameters (UCPs). In this study, we evaluate the ICON TERRA_URB urban scheme against dense urban observations (near-surface sites, higher-level sites, and eddy-covariance towers) and satellite surface-temperature products, focusing on air and surface temperature, sensible and latent heat fluxes, and wind speed over Zurich and Basel during summer 2023 using four experimental configurations at 500 m resolution. These include simulations without an urban canopy scheme (No TU), TERRA_URB with spatially uniform UCPs (Constant TU), local-climate-zone-based spatially varying parameters (LCZ TU), and city-specific urban canopy parameters (Real TU). Results show that activating TERRA_URB, regardless of parameter detail, provides the largest improvement by reducing the nocturnal cold bias, increasing urban surface temperatures, reproducing slightly higher nocturnal wind speeds, and improving surface energy flux partitioning. Spatially varying UCPs further refine the simulations. In particular, Real TU best captures the nocturnal urban signal, improving agreement with observed nighttime temperatures and sensible heat fluxes, although some biases remain in city centers. Their added benefit is modest compared to the main gain from activating the urban scheme itself. A trade-off emerges aloft, as TERRA_URB tends to overestimate warming, with No TU often performing better at most higher-level sites, although this may also reflect coupled boundary-layer processes beyond the urban scheme itself. These results suggest that while bulk urban canopy schemes effectively capture surface exchanges, improved boundary-layer representation likely requires multi-layer canopy approaches.

Marginal and degraded lands in coastal areas have essential ecosystem functions and thus have experienced intensive exploitation over the past half-century. However, the responses of the agriculture-ecology-economy nexus (AEEN) to natural and human disturbances and their drivers and interactions remain insufficiently understood, thereby constraining coordinated environmental governance. Thus, this study performed a case study of Yellow River Delta City, which contains China’s highest delta wetland cover and has a long history of energy development, to examine the regime shifts of AEEN sectors and their interactions within a social-ecological system framework by a combination of biophysical models and socioeconomic statistics. Coastal development promoted built-up area expansion and substantial cropland occupation, while regime shifts in water-sediment dynamics increased the complexity of the delta’s morphology and geometric characteristics. Climate warming and fluctuating precipitation promoted vegetation greening and enhanced biomass accumulation, whereas soil loss remained stable. Economic growth, particularly due to the increasing importance of the tertiary industry, created sufficient employment opportunities. However, the inappropriate urban-rural industrial structure (i.e., over-reliance on energy industry) and low employment and livelihood diversity in rural areas led to population loss and industry transfer. Planting structure adjustments characterized by the expansion of economic crop (i.e., cotton) sown area from 2000 to 2013 temporarily reduced grain productivity. This reversal of the co-benefits between pairwise AEEN sectors during 1978–1999 and 2014–2023 resulted in trade-offs for the period 2000–2013. The findings regarding regime shifts of nexus sectors and their drivers and interactions offer valuable lessons for balancing the multiple dimensions of AEEN sectors and enhancing the adaptability of deltas to changing social-ecological contexts. Such insights may help optimize the efficacy of catchment-adaptive governance and sustain co-benefits for AEEN sectors worldwide.

Abstract Seismic source parameters, such as seismic moment, stress drop, and radiated energy, are key to understanding earthquake dynamics and rupture processes. In this study, we apply a non‐parametric Generalized Inversion Technique (GIT) to characterize the source properties of 735 earthquakes (local magnitudes 1.65 ≤ ML ≤ 4.55) recorded between 2016 and 2024 in the area of the European Southeastern Alps characterized by the presence of the Adriatic indenter. The high‐density seismic network coverage in the region allows for a robust, data‐driven inversion of Fourier amplitude spectra to separate source, path, and site effects. Our results highlight significant spatial variations in seismic source parameters, revealing distinct mechanical properties of fault systems. We find a general increase in stress drop with depth and observe deviations from self‐similar scaling, with apparent stress increasing with event size. The spatial distribution of the Savage‐Wood efficiency suggests differences in dynamic fault strength, with high‐efficiency events concentrated in the northeastern Italian Alps and Snežnik Mt. seismic area, respectively located in the Southeastern Alps Western sector and in the Dinarides seismotectonic domains, while lower‐efficiency events are more common along the Tolmin‐Idrija and Gemona‐Tarcento fault systems and in the Schio‐Vicenza‐Lessini seismotectonic domain. These findings provide new constraints on the mechanical behavior of faults in the northeastern Adriatic indenter section and have implications for seismic hazard assessment and ground motion modeling. The study underscores the importance of accurate source parameter estimation in improving our understanding of regional seismotectonic and earthquake generation processes.

Abstract Supershear ruptures propagate with speeds exceeding the bulk shear wave speed. The traditional Burridge‐Andrews mechanism of supershear transition requires faults to have a low seismic ratio. Liu and Lapusta (2008) used a single‐rupture model of a 1D fault to show that supershear transition can be triggered by favorable heterogeneity even when shear stresses are lower than required by the Burridge‐Andrews mechanism. Our study extends their work to rupture sequences on 2D faults governed by rate‐and‐state friction. We identify a parameter regime in which ruptures are all sub‐Rayleigh and demonstrate two mechanisms for this model to achieve supershear transition. First, introducing a patch of lower direct effect lowers the peak stress and decreases the local seismic ratio, allowing ruptures to become supershear. Then, modifying the parameters of the velocity‐strengthening loading region increases the stressing on the seismogenic zone, promoting supershear transition. In both cases, the supershear transition first occurs within the higher‐stressed bands at the boundary of the seismogenic region. We also investigate the role of inertial effects and show that supershear transition cannot occur in quasi‐dynamic sequences. We find that incorporating even limited stress‐concentrating dynamic effects ensures realistic rupture speeds. Finally, our simulations show that insufficient spatial resolution can change the response from supershear to sub‐Rayleigh. Our findings highlight the importance of studying 2D faults in long‐term sequence simulations, incorporating at least partial effects of dynamic stress concentrations, and adding expected long‐term stresses to single‐rupture dynamic simulations.

Abstract δ‐AlOOH can remain stable down to lower mantle depth, making it a potential carrier for water transport into the deep mantle and a plausible candidate for influencing deep‐seated velocity structures. To elucidate its impacts on the velocity structure of the deep Earth and the global water cycle, we conducted synchrotron radiation ultrasonic experiments to measure the sound velocities of δ‐AlOOH under conditions up to 18.5 GPa and 873 K, and derived its elastic parameters, yielding: K S 0 = 231.6 (2) GPa , K S ' = 4.1 (1) , ∂K S /∂T = −0.025 (1) GPa/K , G 0 = 159.2 (3) GPa , G' = 1.7 (1) , ∂G/∂T = −0.031 (1) GPa/K. By integrating the elastic parameters of major minerals within the stability field of subducting sediments, we modeled the velocity structure of δ‐AlOOH‐bearing sedimentary layers. Combined with the thermal stability of δ‐AlOOH and coexisting minerals, we propose that δ‐AlOOH may contribute to elevated velocity gradients of subducting sediments in the transition zone. Furthermore, using first‐principles calculations, we have determined the seismic velocities of δ‐AlOOH up to 130 GPa. This helps to evaluate its contribution to the seismic properties of localized, slab‐derived hydrous lithologies. Meanwhile, this phase plays a significant role in shaping the heterogeneity along the periphery of Large Low‐Shear‐Velocity provinces and in the generation of Ultra‐Low‐Velocity zones.

Introduction Engineering professionals are central to climate action because their decisions shape infrastructure, energy, water, transport, industrial, and built-environment systems that influence greenhouse gas emissions and climate resilience. This study assessed the awareness, training, engagement, and perceived responsibility of South African engineering professionals registered with the Engineering Council of South Africa (ECSA) in relation to climate change mitigation and adaptation. Methods An exploratory mixed-methods cross-sectional survey was conducted, combining quantitative analysis of closed-ended responses with thematic analysis of open-ended responses. The final dataset comprised 783 valid responses. Results Although 72.1% ( n = 545) of respondents had not received formal climate-related training, trained respondents were more likely to report participation in climate-related projects than untrained respondents. Among trained respondents, 67.8% ( n = 143) reported participation, compared with 34.5% ( n = 188) of untrained respondents. A Chi-square test confirmed a statistically significant association between training and participation [χ 2 (1, N = 756) = 68.43, p < 0.001]. The Mann-Whitney U-test also showed that trained respondents had stronger perceptions of professional responsibility in climate action ( U = 48,985, p < 0.001). Discussion Survey and thematic findings showed widespread but not unconditional support for ECSA to strengthen climate-related training, collaboration, guidance, accreditation, and professional conduct frameworks while avoiding unnecessary regulatory burdens. The study highlights the need to expand climate-related professional development and regulatory guidance to enhance engineering professionals' engagement in mitigation and adaptation and support South Africa's climate resilience objectives.