Earth

This study aimed to identify and characterise major wildfire events in South Africa (SA) between 2004 and 2023 by integrating long-term multi-source datasets with anomaly detection techniques. Biomass-burning emissions, such as black carbon from biomass burning (BCBB), organic carbon from biomass burning (OCBB), and carbon monoxide (CO), were retrieved from the Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2) reanalysis dataset, while burned area data were obtained from the Moderate Resolution Imaging Spectroradiometer (MODIS). Isolation Forest (IF; contamination = 0.05) and the Generalised Extreme Studentised Deviate (GESD) test were applied independently and integrated at the decision level: single-method detections were classified as candidate anomalies, while agreement between the methods identified high-confidence anomalies. Calendar-month standardisation, Spearman rank correlation, Benjamini–Hochberg false-discovery-rate correction, and calendar-month-matched event composites were used to assess meteorological relationships. IF-selected 12 candidate months were identified throughout, whereas GESD identified a smaller subset. July 2007 was the only high-confidence burned area anomaly. High-confidence emission anomalies occurred in 2010, 2018, 2021, and 2023, predominantly between September and November, while the only high-confidence precipitation anomaly occurred in August 2006. After false-discovery-rate correction, the burned area was weakly associated with higher wind speed and lower daytime relative humidity. CO, BCBB, and OCBB were weakly associated with lower precipitation, lower daytime relative humidity, and higher wind speed, while surface air temperature (SAT) showed no significant relationships. Event composites displayed similar patterns, but none of the 16 comparisons remained statistically significant after correction. Spatial analyses showed that the July 2007 burned area anomaly was concentrated in eastern SA; CO and BCBB anomalies were prominent across the northern, central, and eastern interior, and the 2018 OCBB anomalies were concentrated in the southern Western Cape. Percentile-selected spatial composites demonstrated additional regional heterogeneity but were distinct from the IF-GESD consensus anomalies and were interpreted descriptively. The framework, therefore, provides transparent confidence stratification rather than evidence of superior predictive accuracy, while highlighting limitations arising from national monthly aggregation and differences in product resolution. Full article

Groundwater quality in semi-arid regions is influenced by interacting geological, climatic, and human factors. This study integrated hydrochemical analysis, ionic relationships, multivariate statistics, Water Quality Index (WQI), and GIS-based spatial analysis to evaluate groundwater quality in Fars Province, southern Iran. A total of 171 groundwater wells were sampled during each of the 2020 and 2021 monitoring campaigns. Hydrochemical diagrams and ionic relationships indicated the predominance of Na–Cl facies and showed that groundwater chemistry is mainly controlled by carbonate and silicate weathering, evaporite dissolution, cation exchange, water–rock interaction, and evaporation–crystallization. Chloro-alkaline indices indicated a mixed cation-exchange system, with positive CAI values predominating regionally and negative values occurring locally. Principal Component Analysis (PCA) and Hierarchical Cluster Analysis (HCA) consistently identified groundwater mineralization as the dominant source of hydrochemical variability, characterized by EC, TDS, major ions, and hardness, while bicarbonate and nitrate represented a secondary source of variability reflecting both carbonate-related processes and localized nutrient inputs. Based on WQI, about 48% and 42.7% of the sampled wells were classified as excellent or good in 2020 and 2021, respectively, whereas 26% and about 30% were unsuitable for drinking. Spatial analysis revealed widespread mineralization and enrichment of Na⁺, Cl⁻, and SO₄²⁻, although interpolation results for EC and TDS should be interpreted cautiously and primarily for exploratory visualization of spatial patterns because of their lower predictive performance. In general, regional groundwater quality is governed primarily by natural hydrogeochemical evolution, while localized anthropogenic influences may contribute to nutrient variability. The results provide a basis for targeted monitoring and sustainable groundwater management in semi-arid aquifers. Full article

Metal contamination in karst aquifers is a growing concern in rapidly urbanizing coastal regions. The Ruta de los Cenotes in Quintana Roo comprises groundwater-fed cenotes of high hydrogeological and ecological relevance. This study characterized the hydrogeochemical setting of five cenotes along a coastal–inland gradient, quantified metals in groundwater and sediments, and assessed ecological risk. Groundwater was sampled at 1, 15, and 25 m during dry and rainy seasons, and sediments at 30 m depth. Metals (Al, Ba, B, Cd, Cr, Cu, Fe, Li, Ni, Pb, and Zn) were determined by ICP-OES. Coastal and transitional cenotes showed calcium-sulfate waters and EC > 1200 µS/cm, whereas inland cenotes showed calcium-bicarbonate waters and EC < 1000 µS/cm. Multivariate analysis identified mineralization gradients during the dry season and site-specific variability during the rainy season. In groundwater, Al reached 0.487 mg/L in A-Ha, exceeding the NOM-127-SSA1-2021 limit of 0.20 mg/L. Al, Fe, Li, Cu and Zn were the main contributors to high ecological risk, with Al (RA = 3043.75) and Fe (RA = 74.15) showing the highest values during the rainy season. Overall, risk rankings differed seasonally, with Li > Cu > Zn > Fe > B > Ba in the dry season and Al > Fe > Zn > Cu > Cr in the rainy season. Sediment-associated ecological risk was highest at one coastal cenote (RI = 1868.66), driven mainly by Cd (2.17 mg/kg). These results provide quantitative information that can support future assessments and monitoring of metal-related vulnerability in karst ecosystems. Full article

Accurate solar irradiance forecasting in data-sparse tropical regions remains challenging due to complex terrain, rapid convective cloud formation, and limited ground-based observations. This study introduces a novel hybrid forecasting framework that integrates the Weather Research and Forecasting (WRF) model (10 km) with station-based Long Short-Term Memory (LSTM) bias correction, complemented by three innovative spatial site-adaptation strategies to produce spatially coherent short-term irradiance fields. The hybrid system leverages hourly Global Horizontal Irradiance (GHI) data from eight BMKG stations (2023) alongside GK2A satellite cloud information to dynamically correct WRF forecast biases, capturing nonlinear cloud–irradiance interactions that standard Numerical Weather Prediction (NWP) models fail to resolve. Results indicate that the hybrid WRF–LSTM system reduces 1–3-day root mean square error (RMSE) by 120–127 W/m² and relative RMSE (rRMSE) by 26%, while lowering relative mean bias error (rMBE) from 31–36% (raw WRF) to 2.5–4.1%, with the largest improvements observed in regions exhibiting initially high WRF errors. Among the spatial adaptation methods, the average-based scheme minimizes RMSE but exhibits weak spatial coherence; the distance-weighted scheme achieves the strongest spatial consistency with regional reanalysis (R² = 0.37) with minimal bias; and the elevation-based scheme ensures full-domain coverage with moderate skill. This study demonstrates that the integration of dynamical NWP modeling with LSTM-based bias correction and tailored spatial transfer strategies provides a robust, scalable approach for short-term solar irradiance forecasting and resource mapping in tropical environments. The proposed framework offers practical implications for PV power forecasting, grid management, and renewable energy planning in regions where observational data are sparse and the terrain is highly heterogeneous. Full article

Conventional evaluations of rainfall prediction models rely on average accuracy, often masking the conditions, locations and causes of model failure and reduced reliability. This study proposes a paradigm shift from conventional average-accuracy benchmarking toward failure-aware forecast-error diagnosis in the Bengawan Solo River Basin, a tropical monsoon river basin in Indonesia with moderate topographic gradients (grid elevations span ≈ 300–650 m). Methodologically, forecasts from previously published models are treated as fixed inputs and their errors are modelled as the response variable, so the analysis diagnoses when and where models fail rather than retraining them. By treating forecast errors as response variables, rather than as random residuals, this study analyses 345,180 model–grid records–month records from ten individual models (RF, XGB, LGBM, SVR, MLP, LSTM, GRU, TCN, CNN, Transformer) and one best ensemble model (Ensemble_Q, a stacking of RF, XGB, SVR, MLP, LGBM, LSTM, GRU, TCN, CNN, Transformer) against observed CHIRPS (Climate Hazards Group InfraRed Precipitation with Station data) precipitation, seasonal phase, ENSO and IOD regimes (El Niño–Southern Oscillation and Indian Ocean Dipole, respectively), the MJO index (Madden–Julian Oscillation) as an additional analysis, and elevation as a topographic control, using log-error models, high-error logistic regression, interaction tests, and block bootstrap validation (N = 1000), false discovery rate, and spatial statistics. Results indicate that prediction errors are not random but are systematically controlled: the Transition II phase increases log-error by 245% (pooled log-error model) and raises the odds of a high-error event roughly 40-fold relative to the dry season; La Niña conditions amplify errors by 41% and the odds of a high-error event by 3.3 times (though this ENSO signal is largely entangled with co-occurring Negative-IOD months), and every 100 m increase in elevation increases errors by 26%, with errors forming distinct spatial clusters (Moran’s I = 0.78; p = 0.001). Ensemble_Q outperforms the baseline on an aggregate basis (mean absolute error, MAE = 54.10 mm) but still experiences error amplification under these conditions, while spatial deep-learning architectures (TCN, CNN, Transformer) prove most vulnerable to elevation gradients. All major patterns persisted across variations in thresholds, model subsets, ENSO definitions, multiplicity corrections, and bootstrapping. These findings confirm that superior mean accuracy does not guarantee operational reliability, and that conditional failure diagnosis is an essential complement to benchmarking rainfall predictions in tropical monsoon regions. Full article

Vegetation change is regarded as a key indicator of environmental change and ecosystem functional evolution. In mountainous regions, vegetation dynamics arise from complex and non-linear interactions among climate, topography, land use, and human activities, yet the mechanisms governing these interactions remain poorly understood. Clarifying long-term vegetation trajectories and their interacting controls is essential for understanding ecosystem structure and function. In this study, we integrated machine learning and causal modeling by combining random forest (RF) and structural equation modeling (SEM) to quantify both the relative importance and the direct and indirect pathways of natural and anthropogenic drivers of vegetation change in the Longnan region from 2000 to 2020. The RF results showed that the selected driving factors explained 82.62% and 72.39% of the spatial variation in vegetation cover in 2000 and 2020, respectively, with climatic and anthropogenic factors ranking as the most important drivers. Although ecological restoration activities contributed to an overall improvement in vegetation conditions, land-use heterogeneity and ecological constraints imposed by high elevation jointly produced contrasting local responses, resulting in vegetation degradation in surrounding areas. SEM revealed that the net influence of anthropogenic activities shifted from positive to negative over time, mainly due to land-use change, indicating a reorganization of human–vegetation interactions. Climate effects remained positive, with precipitation having a stronger influence than temperature. Topography moderated vegetation responses, as slopes below 40° favored vegetation growth. Soil effects shifted from positive to negative, likely associated with changes in soil organic matter. By jointly applying RF and SEM, this study captures both non-linear responses and causal pathways, providing a system-level perspective on the complex mechanisms underlying vegetation change. Full article

The spatial distributions of the potential ecological risk index (RI) and the geoaccumulation index (Igeo) in soils of the Zhetysu region (Kazakhstan) were examined. The enrichment factor ranged from 39.4 for cadmium to 21,481 for nickel. There are few non-ferrous metal enterprises and mineral deposits located in the region. Soil sampling was conducted at 54 locations across the region. The contamination indices were calculated based on the concentrations of four heavy metals in the soil in 2025: copper (Cu), lead (Pb), cadmium (Cd), and nickel (Ni). The RI index varied across the study area within range of 53–353, with an average value of 129. The main points of high environmental risk are the Balkhash Lake’s eastern coast (RI = 318) and industrial city Tekeli (RI = 353). The Igeo values were as follows: Cu −0.61–3.82 (mean: 2.07); Pb −0.58–3.35 (mean: 0.49); Cd −2.91–2.15 (mean: −0.11); and Ni −4.23–4.17 (mean: 0.43). The transport of contaminants occurs under the influence of the wind regime and water erosion processes. Lead, cadmium, and nickel have a variation coefficient greater than 50, indicating metal particle transport from adjacent areas. Wind patterns in certain areas of the region determine the distribution of foci of high heavy-metal concentrations in the soil. Full article

Sustainable marine economic development depends on the balanced integration of economic, social, and environmental dimensions. However, empirical evidence on the current degree of coordination among these dimensions remains limited, particularly in maritime countries. This study assesses the Coupling Coordination Degree (CCD) among these three dimensions of sustainable marine economic development in 15 Indonesian coastal provinces selected based on a proportion of coastal villages exceeding the national average and availability of complete 2024 data. Using subsystem performance scores derived from a previously published Grey Relational Analysis (GRA) index, the study evaluates how balanced the current coordination state is across provinces. The findings show that higher coordination is associated not only with higher subsystem performance but also with the relative balance among dimensions. Provinces with relatively balanced development tend to achieve higher coordination, whereas strong performance in a single dimension does not necessarily lead to a well-coordinated development system. Conversely, high coordination may also arise from uniformly low performance across all dimensions, indicating that coordination and development performance represent distinct characteristics of sustainability rather than interchangeable concepts. These findings demonstrate that CCD complements conventional performance-based assessments by capturing the degree of integration among multiple development dimensions. The study contributes to the sustainable marine development literature by providing a more comprehensive framework for evaluating multidimensional sustainability and offers insights for designing integrated, place-based blue economy policies that better align economic growth, social inclusion, and environmental conservation. Full article

The conventional linear “treat-and-discharge” model has been challenged due to rapid urbanisation and the increasing generation of complex industrial wastewater. To this end, circular economy (CE) approaches can be considered as a key strategy for resource efficiency, sustainability, and waste reduction. The current study aims to review CE principles in industrial wastewater management policies across Asian countries. Findings from our review indicate that countries such as Singapore, Japan, South Korea, and China have made significant progress through integrated institutional frameworks, standards of reclaimed water, eco-industrial park initiatives, advanced treatment technologies, and supportive policy mechanisms. However, some developing Asian economies still depend primarily on pollution-control regulations and effluent discharge compliance; have limited guidelines on application-specific water-reuse quality standards, limited financial and tax incentives for SMEs, and inadequate policy support for resource recovery and wastewater valorisation. Strengthening these policy and institutional frameworks can promote circular industrial wastewater management, thus enhancing water security, resource efficiency, and sustainable industrial development across Asia. Full article

Lenggong Valley is an important heritage landscape in Malaysia commonly recognized for its outstanding archaeological, cultural, and environmental significance. However, increasing land use pressure around heritage areas may affect landscape authenticity, environmental quality, and long-term conservation planning. This study aims to analyze historical land use change and predict future land use patterns within the Lenggong catchment, a sub-catchment of the Sungai Perak catchment. Land use data for 2000, 2010, and 2020 were obtained from PlanMalaysia and reclassified into five major classes: water bodies, agriculture, forest, built-up areas, and vacant land. The ANN model was calibrated using the 2000 and 2010 land use maps, while the simulated 2020 map was validated against the observed 2020 map using Kappa statistics. Following validation, the 2010–2020 transition pattern and cellular-automata neighborhood effects were used to predict land use for 2040 through two consecutive 10-year simulation iterations. The results showed that forest and agriculture remained the dominant land use classes in the catchment. However, forest decreased from 58.9% in 2000 to 54.4% in 2020, while agriculture increased from 33.3% to 35.5%. Built-up areas also increased from 1.8% to 3.5% and were predicted to reach 3.8%. The model also denoted acceptable performance, with an overall Kappa value of 0.71 and validation accuracy of 83.1%. Within the 1 km heritage buffer, built-up areas increased from 3.6% in 2000 to 10.5% in 2020, with a projected increase to 13.2%. Overall, the findings highlight increasing development pressure around the Lenggong heritage landscape and provide useful spatial evidence for heritage-sensitive planning and long-term conservation management. Full article

Water erosion is a core issue that threatens the ecological integrity of the highland ecosystems in the Andes Mountains and the agricultural sustainability of the Ramis River basin. This study uses the Revised Universal Soil Loss Equation (RUSLE), which integrates five factors, rainfall erosivity (R), soil erodibility (K), topography (LS), cover and management (C), and support practices (P), to estimate the spatial distribution of potential water erosion rates in this basin. The results show that the very low and low erosion classes together cover 73.21% of the basin, while the high, very high, and extreme erosion classes account for 17.29% of the total area. Among these, the extreme erosion class, with an annual erosion volume exceeding 250 tons per hectare, covers 8.07% of the basin, equivalent to 1190.13 square kilometers. This extreme erosion is concentrated in steep headwater areas and five sub-basins including Cuenca Grande. Comparative model verification shows that the Ordinary Least Squares (OLS) model only identifies a positive correlation between slope gradient and potential soil loss, with an extremely low explanatory power (R² = 0.045). Its residuals exhibit significant spatial autocorrelation (Moran’s I = 0.204, p < 0.001). In contrast, the Geographically Weighted Regression (GWR) model greatly improves the model fit (R² = 0.359, RMSE = 148.288) and eliminates the spatial autocorrelation of residuals, proving that the slope-erosion relationship has spatial non-stationarity. Sensitivity analysis shows that the C factor has the highest sensitivity (0.980), followed by the LS factor (0.626). Based on these findings, this study proposes that cover and management measures such as vegetation restoration should be prioritized in high-risk headwater sub-basins. It should be noted that the values estimated in this study are potential soil loss amounts, rather than actually measured erosion values. Full article

Reliable spatial evidence can support quality assurance and quality control for land use, land-use change and forestry (LULUCF), but land-cover and species-distribution layers do not by themselves identify IPCC Forest Land or species-pure stands. We estimated 2010 and 2020 above- and below-ground living biomass carbon within tree-covered European beech (Fagus sylvatica L.) distribution and occurrence footprints in Greece. Our operational hypothesis was that increasingly restrictive species masks would materially alter the mapped extent and carbon estimates. ESA Climate Change Initiative Biomass v6, ESA WorldCover 2021, European Forest Genetic Resources Programme (EUFORGEN) polygons, and Forest Information System for Europe (FISE) relative probability of presence layers were processed in Google Earth Engine. Biomass was converted with IPCC default carbon fractions and root:shoot ratios, and the results were summarized nationally and for GAUL Level-2 units. The broad EUFORGEN footprint covered 22,133 km², whereas the Combined overlap of EUFORGEN, FISE relative probability of presence ≥ 0.50, and tree cover covered 2742 km². Within the Combined footprint, the pixel mean living biomass carbon density was 60.33 Mg C ha⁻¹ in 2010 and 62.50 Mg C ha⁻¹ in 2020, and the area-integrated change was +0.58 Tg C; the area-normalized regional change was positive in 13 of 17 units and negative in 4. Across masks, the mean decadal change ranged from −0.50 to +3.37 Mg C ha⁻¹ and the approximate area-integrated totals from −1.10 to +0.58 Tg C. These scenario-conditioned estimates are neither official national greenhouse gas inventory estimates nor tests of statistical significance; instead, they provide reproducible spatial screening while making mask sensitivity and unpropagated uncertainty explicit. Full article

Flash floods are among the most damaging hydrometeorological hazards in the Western Balkans (WB), yet regionally consistent, cross-border susceptibility assessments remain scarce because of fragmented national datasets and differing methodological standards. This study develops a harmonized, cloud-based flash-flood susceptibility framework for the WB (208,052 km²) by implementing a physiography-based modified Flash-Flood Potential Index (FFPI) in Google Earth Engine (GEE) at 30 m resolution. The modified FFPI integrates slope, land cover, soil texture, vegetation exposure (Bare-Soil Index), and soil erodibility, and is aggregated across 9524 EU-Hydro sub-basins to produce an operational catchment-level ranking. Additionally, CHIRPS-derived maximum daily precipitation is used to derive a rainfall-triggered hotspot layer that highlights sub-basins where terrain-controlled susceptibility coincides with strong observed rainfall extremes over the 2001–2025 period. Enhanced susceptibility is concentrated in Adriatic and Aegean-facing mountain basins of Albania, Montenegro, and North Macedonia, with 44.2% of sub-basins classified as high or very-high susceptibility. Multi-source validation against inventoried torrential catchments, published GIS-based susceptibility maps, and flood records yielded moderate to very strong agreement (68.6–92.0%), together with an AUC-ROC of 0.79 and F1-score of 0.77 for the pooled orthophoto-based validation dataset (n = 336 sub-basins). The framework provides a reproducible transboundary tool for regional flood-risk screening and demonstrates the potential of cloud-based geospatial platforms to overcome cross-border data fragmentation in hazard assessment. Its main limitations are the static physiographic nature of the FFPI, the coarser resolution of CHIRPS and SoilGrids relative to small sub-basins, and possible overestimation in karst terrains where subsurface drainage reduces surface runoff. Full article

Population growth is increasing food demand, while watershed food systems face intense pressure from land-use/land-cover (LULC) change. Therefore, this study aimed to identify the critical threshold for food sufficiency in the Cimanuk Watershed. The methods used include dynamic analysis and prediction of LULC changes in line with four scenarios using Support Vector Machine (SVM)–Markov model, ecosystem service-based food production modeling through Crop Production module in Integrated Valuation of Ecosystem Services and Trade-offs (InVEST) model, food demand estimation based on population projections, and calculation of Food Sufficiency Index (FSI) to identify the critical threshold at which the food system shifts from surplus to deficit. The results showed that paddy rice production is projected to continue meeting food demand across all scenarios through 2042. However, food demand is expected to increase as the population grows, specifically in the Accelerated Population Growth (APGS) scenario. FSI analysis indicated that the Cimanuk Watershed remained in food surplus (FSI > 1) up to 2042. The watershed may transition to a food-deficit condition after 2062 in the absence of intervention measures. Therefore, this study recommends protecting productive paddy fields, conserving forests, and managing population growth to maintain long-term food sufficiency. Full article

Tropical mining landscapes require plot-based structural and carbon baselines to support restoration planning and environmental monitoring. This study estimated aboveground, root, and total tree carbon stocks across six mining concessions located in the Inambari River basin, Madre de Dios, southeastern Peruvian Amazon. Field inventories were conducted in 39 plots of 0.1 ha, where all trees with DBH ≥ 10 cm were measured. Aboveground biomass was estimated using a pantropical allometric equation based on wood density, diameter, and height; root biomass was estimated using a baseline root-to-shoot ratio of 0.24; and biomass was converted to carbon using a baseline carbon fraction of 0.47. Structural attributes, biomass stocks, carbon stocks, diameter-size profiles, basal-area contribution by diameter class, and multivariate structural-carbon patterns were evaluated among concessions. The concessions differed significantly in measured structural attributes, and these differences translated into contrasting derived biomass and tree carbon estimates. Edmilot I showed the highest basal area per hectare, total biomass, and total tree carbon, reaching 118.34 Mg C ha⁻¹, while Yesica recorded the lowest total tree carbon, with 58.54 Mg C ha⁻¹. Most trees were concentrated in the 10–20 cm and 20–30 cm DBH classes, but intermediate and larger trees contributed disproportionately to basal area. Principal component analysis summarized the concessions according to a reduced set of non-redundant structural and carbon variables, with the first two components explaining 99.21% of the total variation. These findings show that mining concessions retain contrasting forest conditions and should not be treated as homogeneous units. Structural and carbon baselines can help identify internal differences among concessions, prioritize restoration actions, and improve environmental monitoring in tropical mining landscapes. Full article

Urban and industrial activities have lasting effects on Earth ecosystems, impairing their functionality. Technosols offer a sustainable solution for restoring degraded urban and industrial ecosystems. In a 30-day microcosm experiment, a mixture of six pioneer plant species was sown in three substrates: a sewage sludge-based Technosol (GF), a zeolite-enriched Technosol (GZ), and a commercial potting mix control (GC). The plant population dynamics, final performance, and substrate biochemical properties were monitored. A strong environmental filter allowed only two species (Bromus inermis Leyss. and Lolium perenne L.) to establish. Technosols exerted a demographic bottleneck, delaying emergence and reducing the total biomass relative to the control. Zeolites in the GZ Technosol mitigated this delay, accelerating early establishment due to their microporous structure and high cation exchange capacity. However, GZ caused the greatest reduction in individual biomass and functional plant performance index, corresponding to a microbial shift toward oxidative activity at the expense of hydrolytic nutrient mineralization. These results show that sewage sludge Technosols can initiate functional ecological succession. While zeolites positively affect germination, their microbial interaction suggests a temporary decoupling between the establishment speed and final productivity. Integrated monitoring of demographic and biochemical dynamics is therefore essential to optimize Technosol-based environmental restoration. Full article

Reliable coastal wetland habitat mapping is often constrained by the scarcity and the cost of reliable reference data, especially in data-limited coastal environments. We propose a validation-controlled, label-efficient framework pairing multi-season Sentinel-1 and Sentinel-2 predictors with a CatBoost teacher and a lightweight MLP student. A candidate is pseudo-labeled only when both separately calibrated models agree and exceed class-specific thresholds; accepted labels are class-balanced and down-weighted. The framework was evaluated at the Sidi Moussa–Oualidia wetland complex and Merja Zerga lagoon in Morocco. At Sidi Moussa–Oualidia, 62 configurations were compared through nested polygon-grouped validation and then frozen before a five-seed held-out evaluation. The supervised MLP and Agreement-augmented MLP achieved mean Macro-F1 values of $0.9518\pm 0.0044$ and $0.9509\pm 0.0062$, indicating that augmentation did not materially change the already strong full-data baseline. Under a stricter budget of 30 training and 20 validation observations per class, Agreement yielded a mean Macro-F1 of $0.9092\pm 0.0102$ compared with $0.9023\pm 0.0093$ for the supervised baseline and produced pseudo-labels in all five seeds. A spatial-range sensitivity analysis further showed that both models retained Macro-F1 values of 0.9391 and 0.9403 for test observations located beyond the largest estimated within-class autocorrelation range. At Merja Zerga, the native six-class supervised MLP achieved $0.9456\pm 0.0050$, compared with $0.9401\pm 0.0047$ after Agreement augmentation. Spatially blocked four-class experiments nevertheless showed that 20 to 30 local training labels per class recovered approximately 96–98% of the corresponding full-data performance. The framework therefore supplies an operational criterion for using unlabeled observations: augmentation is adopted only where calibrated filtering yields adequate class coverage, and validation confirms a downstream effect; otherwise the supervised model is retained. For the strict Sidi Moussa–Oualidia reduced-label experiment, the reported development budgets count every site-specific label used for fitting, early stopping, and calibration. The Merja Zerga blocked experiments separately quantify training-label sensitivity while retaining their blocked validation resources. Full article

Mangroves are highly productive ecosystems due to their great capacity to store carbon, but they are also vulnerable to human activities and adverse environmental conditions. Their conservation is often constrained by the opportunity costs of shifting from traditional economic activities to blue carbon projects. This work analyzes the prices of the different carbon markets in ecosystems, compares them with the benefits obtained from other productive activities, and evaluates the viability of implementing carbon projects in mangroves. An exhaustive literature search is conducted to assess the carbon prices of ecosystems worldwide. The opportunity costs of mangrove carbon capture projects in Mexico are estimated from site-specific and regionally relevant economic data; additionally, broader national benchmarks are presented for contextual comparison but are not interpreted as direct opportunity costs where they do not represent realistic land-use alternatives for the mangrove sites analyzed. The price of carbon ranges from 4 to 86 USD per Mg CO₂e. The most studied natural ecosystems are forests. The highest gross annual profit (GAP) from carbon sales is observed in Tabasco and Campeche. GAP with mangrove wood harvesting ranges from 628.0 USD ha⁻¹ year⁻¹ to 3917.7 USD ha⁻¹ year⁻¹. The highest GAP for crops is obtained for white corn in the state of Hidalgo. GAP of the economic activity of livestock ranges from 3167.59 USD ha⁻¹ year⁻¹ to 3365.71 USD ha⁻¹ year⁻¹. The blue carbon projects are competitive with other productive activities at relatively high prices (86 USD per Mg CO₂e). In Tabasco, under certain high-price and high-sequestration scenarios, blue carbon projects can be competitive with local agricultural activities; however, this competitiveness is highly conditional on carbon price, sequestration rates, and local opportunity costs, and therefore cannot be generalized to all mangrove owners without site-specific appraisal. Fair carbon prices are required to make mangrove conservation projects attractive to producers. Full article

Per- and polyfluoroalkyl substances (PFASs) are persistent synthetic chemicals of global concern. While most research has focused on their occurrence and toxicity, far less attention has been paid to their impacts on the global carbon cycle. This review synthesizes current evidence on how PFASs influence carbon cycling across soils, aquatic systems, and the atmosphere. In soils, PFASs alter organic carbon inputs by affecting plant biomass and root exudates and shift microbial community composition and enzyme activities, thereby modulating organic matter decomposition. In aquatic ecosystems, PFASs biologically impair carbon sequestration by inhibiting plankton, and abiotically interact with extracellular polymeric substances to prolong the cycling of dissolved organic carbon. The atmosphere acts as a key mediator as follows: thermal treatment of PFASs generates perfluorocarbons, potent greenhouse gases that exacerbate global warming and further disturb carbon cycling. Despite clear disruptive effects, major knowledge gaps remain. Future research should use quantitative structure–property relationship modeling to assess PFAS alternatives (e.g., PFHxS), and employ advanced molecular tracking (e.g., isotopic labeling, NanoSIMS) and machine learning to unravel nonlinear PFAS–carbon dynamics. Improved detection technologies are needed to identify greenhouse gas byproducts from PFAS thermal treatment. Ultimately, deploying high-resolution flux observation networks and integrating PFAS dynamics into Earth system models and carbon-accounting frameworks are critical for predicting carbon–climate feedback and supporting global carbon neutrality goals. Full article