Environmental and Earth Sciences

Abstract: Road traffic is the main source of exposure to environmental noise in Europe and Directive 2002/49/EC requires the competent authorities to evaluate and manage it. In port cities, a significant portion of this exposure is generated by road traffic induced by ferry embarkation and disembarkation operations: a pulsating and highly non-stationary source, which standard forecasting methods, designed to return long-term average indicators, describe only in aggregate form. This work compares two forecasting approaches applied to the same waterfront of Olbia (Sardinia, Italy) and addresses a question of an operational rather than metrological nature: which model provides the information that a specific noise management decision actually needs. A physical model based on ISO 9613 and the CNOSSOS-EU methodology was implemented in CadnaA and calibrated on continuous monitoring data conducted in three receptor locations during a low and a high season. A non-linear autoregressive artificial neural network, developed and experimentally validated previously on the same waterfront, was used to predict the equivalent sound pressure level starting from vehicle flow data alone. The Lday, Levening and Lnight indicators returned by the two models are compared, together with the required input data, the spatial and temporal resolution, the latency and the installation and update costs over the five-year cycle of the strategic acoustic mapping. The differences between the two models reach 14 dB in low-flow night periods, while the calculation times vary between 3 and 36 h for the physical model versus an almost real-time inference for the neural one. The two approaches are complementary and not alternative: the physical model remains necessary for planning and for scenarios not yet implemented, while the operational and intraday management of noise in a ferry port is better served by data-driven prediction. The two models agree within 1 dB on average during the day and evening periods, whereas at night the neural model returns systematically higher levels, by 5.4 dB on average and by up to 14 dB, at all receiver positions and in both seasons.

Abstract: The risk arising from the uncontrolled burning of waste is significant given the spatial conditions and the exposure to human health. The environmental risk associated with the emission of harmful gases and particulates into the air as a result of uncontrolled waste burning is the subject of an analysis of the legal consequences of the incident. In order to determine the risks involved, it is necessary to analyse the volume of emissions, the qualitative composition of these emissions, the range of pollutant dispersion during the event, and the level of human exposure in relation to the distribution of the pollutants present. As part of this study, the authors have undertaken work to address a research gap involving the development of a methodology for assessing environmental risk based on an analysis of the degree of exposure of people and the environment to air emissions from a waste storage facility fire. The resulting risk is assessed using analytical methods based on mathematical modelling of pollutant dispersion using the Gaussian “plume” model (Model Pasquilla) using a computer programme called Operat FB and analysing the reference levels set by the NDS (Maximum permissible concentration), EPA (The Environmental Protection Agency), OSHA (Occupational Safety and Health Administration) and under local Polish legislation. This paper presents a case study, namely the uncontrolled incineration of waste resulting from the treatment of municipal waste.

Abstract:

Two novel facultatively anaerobic thermophilic bacteria, strains 4302-co^T and VF2^T, were isolated from terrestrial hot springs of Geyser Valley and Lake Baikal region (Russian Federation), respectively. Cells were motile cocci, forming aggregates. Strain 4302-co^T grew optimally at 60–65 °C and pH 7.0; strain VF2^T – at 55 °C and pH 8.5. They were chemoorganoheterotrophs, utilizing mono-, di- and polysaccharides (cellulose, xylan, xanthan gum, arabinan, gluco- and galactomannan). Growth under anaerobic conditions was observed during fermentation or respiration with sulfur, nitrite, nitrate, ferric citrate or ferrihydrite. Strains 4302-co^T and VF2^T had genome sizes of 3.94 Mb and 4.32 Mb, with G+C contents of 65.8% and 65.2%, respectively. The genomes of both strains encoded a vast number of carbohydrate-active enzymes, including more than 100 glycoside hydrolases of diverse families and more than 15 polysaccharide lyases, along with enzymes involved in the catabolism of glucose, mannose, galactose, ribose, and rhamnose. Fermentation pathways as well as electron transfer chains of aerobic and anaerobic respiration were also reconstructed. Based on phylogenomic analysis and phenotypic features, the isolates 4302-co^T and VF2^T were assigned to novel species of the genus Limisphaera, for which the names Limisphaera vallis sp. nov. and Limisphaera sibirica sp. nov. are proposed.

Abstract: To investigate how the anatexis of accretionary complexes contributes to crustal differen-tiation in orogenic belts, we present an integrated analysis of whole-rock geochemistry, zircon U-Pb geochronology, and zircon Lu-Hf isotopes for Late Carboniferous granitoids from the Dunhuang orogenic belt. Zircon U-Pb dating indicates that the granitoids were emplaced between 323 and 312 Ma. Their zircon εHf(t) values (−39.3 to 10.7) are largely negative, suggesting derivation primarily from ancient crustal materials. Geochemical correlations and field relationships between the granitoids and their wall rocks suggest that the magmatic sources were dominated by meta-sedimentary rocks with minor me-ta-mafic rocks. The systematic whole-rock Rb-Sr-Ba variations are attributed to wa-ter-fluxed melting of muscovite in a metasedimentary source. Given the continued activity of the Dunhuang subduction zone during the Late Carboniferous, we infer that elevated temperature combined with fluid-rich conditions in the deeper accretionary wedge trig-gered anatexis of the accreted materials. The segregation and upward migration of granit-ic magmas leave a residue within the deeper accretionary wedge that resembles the lower continental crust in composition. This process enhances vertical differentiation within the orogenic belt crust and likely plays an important role in transforming loose accreted ma-terials into more mature continental crust.

Abstract: Primary teachers shape early environmental understanding, yet their formal preparation in waste management remains poorly documented. This study audited all nine Greek public undergraduate primary-teacher programmes. Official study guides and course descriptions available by 29 August 2026 were screened for explicit content on solid waste, wastewater, sludge, and biosolids. Generic environmental references and undifferentiated mentions of waste were excluded unless a target domain or qualifying management operation was explicit. Five programmes (55.6%) contained at least one relevant course. Solid waste appeared in four programmes (44.4%), wastewater in three (33.3%), and sludge or biosolids in none. Of the five retained courses, one was compulsory, one compulsory-choice, and three elective; only one programme (11.1%) guaranteed relevant exposure. Four courses provided substantive coverage and one incidental exposure. A full independent audit produced agreement on 40 of 45 programme-level binary decisions (88.9%; pooled Cohen's kappa = 0.72). Discrepancies were resolved by consensus using predefined rules and original sources. The documented curriculum therefore provides fragmented and uneven preparation, with compulsory exposure rare and no explicit coverage of treatment residuals. The findings support consideration of a minimum learning entitlement integrating the waste hierarchy, wastewater treatment, treatment residuals, resource recovery, risk communication, and primary-school pedagogy.

Abstract: The rubber antioxidant 6PPD and its toxic quinone transformation product 6PPD-Q are emerging aquatic contaminants with fragmented global occurrence data. This study constructed a full-process data mining framework integrating literature retrieval, lexical matching, attention-augmented neural network modelling and tabular extraction to compile 6PPD/Q records from 6360 Web of Science papers, achieving 85%-95% entity recognition accuracy. The compiled dataset revealed a pronounced mid-latitude enrichment pattern (30° N-45° N) correlated with population density and traffic intensity, with ultra-high hotspots in Los Angeles runoff (up to 6.10 μg/L) and Lake Sihwa sediment (330 ng/g). Toxicity assessment across 26 species confirmed 6PPD-Q is far more acutely toxic than 6PPD, especially to salmonids. Global risk quotient mapping consistently locked ecological hotspots to the Northern-Hemisphere mid-latitude belt, with the North-American west coast as the highest-risk zone. This work provides a reusable framework and the most comprehensive global 6PPD/Q dataset to support regulatory formulation and ecological risk mitigation.

Abstract: Low-light remote sensing images acquired by a TDI-ICMOS system under photon-limited conditions usually exhibit severe underexposure, amplified random noise, unstable local contrast, and unnatural highlight transitions. These challenges are more difficult in our target setting because the data are real 16-bit grayscale remote sensing images without paired clean references. We propose WISE-Net, a Wavelet-guided Illumination Self-supervised Enhancement framework for low-light remote sensing images. WISE-Net follows a two-stage fully self-supervised design. In Stage 1, the input image is decomposed by Haar wavelets, and only the low-frequency component is modulated by a learned brightness-adaptive gain while the high-frequency sub-bands are preserved for structure-consistent reconstruction. A composite self-supervised objective stabilizes enhancement, protects bright regions, and reduces gain artifacts. In Stage 2, the Stage-1 enhancement network is frozen and a self-supervised blind-spot refinement network suppresses enhancement-amplified noise and local artifacts. Experiments on three real low-light remote sensing test subsets show that WISE-Net achieves the lowest NIQE on all three subsets, the lowest PIQE on test1_2048×1440 and test2_4096×1440, and the highest entropy on all three subsets.

Abstract: A decline in CO₂ emissions is often treated as evidence of successful climate policy, although the same outcome may reflect technological modernization, economic contraction, demographic decline, deindustrialization, or external shocks. This study develops a three-stage diagnostic framework combining additive Logarithmic Mean Divisia Index (LMDI) decomposition, sectoral carbon cost exposure, and an investment-capacity proxy to distinguish managed decarbonization from carbon-intensive decoupling and structural or shock-driven emissions decline. The framework is applied to Germany, Poland, and Ukraine for 1990-2024, with subperiods reflecting major structural transformations and shocks. Germany reduced emissions by 482.48 Mt CO₂ despite rising population and GDP per capita, as lower energy- and carbon-intensity more than offset these upward pressures. Poland reduced emissions by 103.48 Mt CO₂, through energy-intensity improvements, while carbon-intensity gains were more limited. Ukraine recorded the largest decline (564.03 Mt CO₂), but substantial contributions came from lower GDP per capita and population, while carbon-intensity increased emissions over the period. At a common benchmark of EUR 75.28/tCO₂, Ukraine has the highest exposure among the integrated-steel cases, whereas Poland has the highest exposure among the electricity cases. The findings show that emissions outcomes should be interpreted jointly with their drivers, sectoral carbon-price transmission, and the capacity to finance technological response.

Abstract: High-density cities increasingly treat parks, greenways, waterways, streetscapes, and urban landscapes as a connected green infrastructure (UGI) system supporting climate adaptation, biodiversity, health, and recreation. Yet these functions depend on how land-use planning, project delivery, and long-term management are connected. This study uses a qualitative, mechanism-oriented comparison to examine how Singapore governs networked urban landscapes and how these functions relate to Shanghai's policy transition. A 2025 professional study visit provided contextual evidence alongside policy and planning documents, statistics, institutional materials, and peer-reviewed research. Five interrelated mechanisms were identified in Singapore: nested planning and implementation rules; networked multifunctional land use; project-level interagency co-production; professional stewardship linked to operational data; and sustained participation combined with differentiated access. Shanghai's recent plans address several corresponding functions, including life-cycle park management, integrated spatial data, ecological monitoring, and collaborative governance, but published evidence does not establish how consistently they operate across districts, projects, and asset owners. The comparison supports conditional rather than direct transfer: governance functions should be adapted to local authority, spatial and climatic conditions, operating capacity, and accountability. The resulting implementation-chain framework links spatial connectivity with institutional continuity and helps diagnose where urban landscape and greenway objectives may become disconnected from long-term operation.

Abstract: Bacterial cells are richly endowed with different surface functional groups that are known to be very useful in the sorption of heavy metals for environmental remediation. One example is Escherichia coli. But, these bacterial cells need to be inactivated in order to offer prolonged shelf-life for practical applications in wastewater treatment, and these inactivation methods should not damage or remove substantial amount of the cell surface functional groups. In this work, E. coli DH5α was used as model organism for examining the effectiveness of freeze drying, dry heat, and autoclave treatment in protecting the cell surface functional groups while inactivating the bacterial cells. Measurement of E. coli zeta potential-pH profile after cell inactivation treatments revealed that all three treatments significantly changed the point of zero charge of E. coli with the likely loss of substantial amount of amine functional group. However, there seemed to be much less removal of carboxylic acid functional groups due to the maintenance of the buffering range between pH 5 and 10 for E. coli with similar zeta potential values across E. coli treated with the three inactivation methods. Data obtained in this work illustrated that freeze drying, dry heat, and autoclave treatment did not significantly remove the main heavy metal binding carboxylic acid functional groups, and the uptake capacity for heavy metals likely remained high after cell inactivation treatment. On the downside, substantial removal of amine functional groups by these cell inactivation methods meant that inactivated E. coli cells have reduced uptake capacity for anionic heavy metal pollutants such as arsenic compounds.

Abstract: Avalanche companion rescue provides the only realistic opportunity to locate and extricate a buried victim during the critical 10–15 min survival interval before organized rescue teams can normally arrive. Motivated by this stringent time constraint, this paper presents a physics-based waveformdomain framework for three-dimensional avalanche victim localization using a helmet-mounted Sparse Uniform Circular Array (SUCA) receiver and a compact cooperative RF beacon. Unlike conventional avalanche transceivers, which require sequential signal-search, coarse-search, and finesearch procedures, the proposed approach determines the victim’s three-dimensional coordinates directly by correlating the received waveform with a precomputed electromagnetic dictionary, enabling immediate guidance toward the estimated burial location. A comprehensive electromagnetic model incorporates realistic snow dielectric properties, propagation through snow and air, snow–air refraction, LOG-based deterministic multifrequency waveform generation, and dictionary-based localization. Numerical simulations demonstrate unique waveform representation throughout the investigated search region, complete recovery of all candidate victim locations under noiseless conditions, practical localization accuracy using a 0.5 m spatial grid, and a calculated 38 dB link margin under conservative very-wet-snow conditions. The proposed framework also supports several promising extensions beyond victim localization, including respiration monitoring, approximate body-orientation estimation, and enhanced victim detectability using passive conductive textile threads integrated into avalanche garments. Although demonstrated for avalanche companion rescue, the proposed waveform-domain localization framework represents a new operational concept for time-critical search-and-rescue and may be extended to a broad class of cooperative and non-cooperative RF sensing applications.

Abstract: Urban park reviews provide a large, low-cost record of how visitors describe everyday green-space experiences, although such data require careful interpretation. This study analyzes 7,585 English-translated Google Maps reviews from 20 urban parks in Bangkok, Thailand, using three transparent lexicon-based signals: ecology, facilities, and social atmosphere. The analytical dataset was derived from an initial collection of 14,965 review records. Direct incidence of focal terms was calculated for individual reviews and then aggregated by park. Facility-related language appeared in 36.0% of reviews, social-atmosphere language in 29.5%, and ecology-related language in 25.8%; 59.1% of reviews contained at least one focal signal. Substantial park-level variation was observed. Lumphini Park had the highest ecology signal among parks with substantial review counts (45.0%), Benjakitti Park had the highest facility signal (50.1%), and Princess Mother Memorial Park had the highest social-atmosphere signal among parks with at least 100 reviews (40.3%). These findings demonstrate the value of online reviews for identifying recurring dimensions of park experience while also highlighting the need to account for self-selection, translation, and keyword ambiguity when interpreting user-generated content.

Abstract: The occurrence of anthropogenic materials in commercially important fish provides relevant information on environmental exposure in coastal urban systems. This study investigated materials detected in the gastrointestinal tracts of croaker (Micropogonias furnieri) and horse mackerel (Trachurus trachurus) marketed in Santos, Brazil, and Maputo, Mozambique, integrating µ-Raman spectroscopy, frequency analysis, and correspondence analysis with market categories and market-associated socioeconomic profiles. µ-Raman analysis yielded 65 material and compound spectral assignments, including carbonaceous materials, synthetic polymers, cellulosic materials, pigments, dyes, and other chemical signatures. Carbon and graphite were the most frequent components, while polyethylene terephthalate (PET) and polyester-related materials predominated among synthetic polymers. Both study areas showed comparable spectral-occurrence patterns, particularly for PET, graphite, and cellulosic materials. Global chi-square tests showed no statistically significant dependence between material occurrence and either market categories or socioeconomic profiles. However, adjusted standardized residuals and correspondence analysis revealed localized distributional patterns involving specific materials and market categories, which should be interpreted as exploratory rather than causal relationships. Integrating µ-Raman spectroscopy and multivariate analysis provides a useful framework for characterizing chemically diverse materials associated with gastrointestinal exposure in commercially important fish. Because we analyzed only gastrointestinal tissues, the findings do not demonstrate contamination of edible muscle tissues, human dietary exposure, or health risk.

Abstract: Wildfires increasingly affect ecological systems, rural livelihoods, infrastructure, and emergency-management capacity across Southeast Europe, yet national assessments often treat active-fire detections and mapped burn scars as separate evidence streams. This study examined the spatial distribution, intensity, and territorial consequences of open-space fires in Serbia during June-August 2025 by integrating 5,233 NASA Fire Information for Resource Management System (FIRMS) records with 269 burned-area features from the European Forest Fire Information System (EFFIS). MODIS and VIIRS active-fire detections were filtered to retain nominal- and high-confidence observations and were combined with administrative boundaries in ArcGIS Pro. Spatial Join/Intersect procedures assigned observations to municipalities, while Kernel Density Estimation with a 10-km search radius and Natural Breaks classification identified national and local concentration patterns. The results showed marked spatial heterogeneity. Smederevo, Bor, Kuršumlija, and Podujevo formed the highest-frequency municipal group, whereas the largest burned-area classes were concentrated primarily in the southern and southeastern municipalities of Žitorađa, Kuršumlija, Preševo, and Leskovac. Density modeling identified two principal cores: a southern Toplica-Pčinja concentration and an agricultural concentration around Smederevo, with secondary activity in eastern Serbia. Fire radiative power patterns distinguished a frequent, lower-intensity agricultural model from high-energy mountain-forest and mixed industrial-forest models. Integrating FIRMS detections with EFFIS polygons provided a more complete interpretation than either source alone by linking where intense thermal activity occurred with the spatial extent of subsequent land-cover damage. The findings support differentiated prevention, satellite-based early warning, accessibility modeling, and intermunicipal data exchange. Because no direct statistical model of meteorological variables was conducted, climatic explanations are interpreted as contextually plausible associations rather than tested causal effects.

Abstract: Multimodal remote sensing classification benefits from complementary spectral, structural, and elevation information. Convolutional neural networks (CNNs) have limited long-range modeling, whereas Transformers incur quadratic computational and memory costs. Although Mamba enables linear-complexity sequence modeling, multimodal Mamba methods usually fuse modalities before selective scanning, limiting the direct use of auxiliary information in parameter generation. We propose Cross-Modal Selective Scan (CM-S6), which uses lightweight projections to condition selective state-space parameters Δ, B, and C on auxiliary information. These parameters regulate state retention, information writing, and state readout, respectively. With input injection, output calibration, and five bounded learnable coefficients, CM-S6 forms an input-parameter-output fusion mechanism without increasing backbone width or depth. Across ten independent runs, CM-S6 achieves OA values of 96.16±0.18%, 91.91±0.49%, and 82.18±0.46% on Houston 2013, Augsburg, and MUUFL Gulfport, respectively, using fewer than 1 M parameters-1.8-2.9% of the strongest compared baselines. Paired tests detected no significant OA difference from the strongest baseline on any dataset, while CM-S6 showed significant OA gains over PICNet and MCAMamba/MTMixer on the respective datasets. CM-S6(L) retains parameter-level modulation, reduces model size to approximately 0.19 M parameters, and increases throughput by 3.86-5.82×, with a dataset-dependent accuracy trade-off. CM-S6 offers an alternative for multimodal fusion under resource constraints.

Abstract: In this research, a falling film reactor with UV light lamps as a photon emission source was designed and constructed for the degradation of the organic pollutant pyridine. Doping of the catalysts, zinc oxide with gold (ZnO/Au), and titanium dioxide with gold (TiO₂/Au) were performed by photodeposition technique. Subsequently, the doped photocatalysts were supported on ceramic clay sheets at 550 °C. The characterization of the photocatalysts was carried out by Scanning Electron Microscopy (SEM) and UV-vis DR spectroscopy. The following operating conditions were used for pyridine degradation: pH (acidic pH =5.5, natural pH =7.2 and basic pH =8.5) and hydraulic residence time (HRT) (0.2 L/min, 0.5 L/min and 1 L/min). Degradation monitoring was performed by UV-vis Spectroscopy, Total Organic Carbon (TOC), and High-Performance Liquid Chromatography (HPLC). A maximum pyridine removal of 97.64% was obtained using the ZnO/Ag photocatalyst, demonstrating the viability of this photocatalytic system for the degradation of organic pollutants such as pyridine.

Abstract: The Moving Bed Biofilm Reactor (MBBR) is commonly used for NH₄⁺-N removal from seawater recirculating aquaculture systems (RAS), while total nitrogen (TN) reduction is increasingly required. Aerobic denitrification offers a potential solution for simultaneous TN and NH₄⁺-N removal, but rapid reactor start-up remains challenging. This study investigated the start-up of aerobic denitrifying MBBRs treating mariculture wastewater. Eight MBBRs were operated at 25 ℃ with four C/N ratios (5.5, 4.5, 3.5, and 0) and two aeration rates (1.0 and 0.4 L/min). Only the CN4.5-0.4 MBBR met the predefined start-up criteria, achieving start-up in 26 days with an NH₄⁺-N removal efficiency of 93.8%. Effluent NH₄⁺-N and NO₂⁻-N remained below 0.5 mg/L without NO₃⁻-N accumulation. External carbon addition was essential, but excessive carbon did not accelerate start-up; CN3.5 MBBRs showed slower and less stable NH₄⁺-N removal, whereas CN5.5 MBBRs showed higher NO₃⁻-N removal but still accumulated NO₂⁻-N. Low aeration increased the average NO₃⁻-N removal efficiency at C/N = 5.5 from 39.0% to 66.2%. Microbial community analysis identified Pseudomonas, Vitellibacter, and Paracoccus as the dominant nitrogen-removing genera, with a combined relative abundance of 41.0% in the CN4.5-0.4 MBBR. Metagenomic sequencing identified genes associated with multiple nitrogen-transformation pathways and supported the potential for aerobic denitrification.

Abstract: The era of high-quality urban development demands refined, digitalized, and visualized scientific research, and 3D point cloud acquisition technology has emerged as a key tool for tree information collection, surveys, assessment, monitoring, and management, which facilitates the innovative development of the landscape architecture industry. While existing applications primarily focus on batch acquisition of point clouds for tree communities, scattered trees in fields such as ancient and famous tree conservation still rely on backpack/handheld LiDAR devices. These devices often require multiple circular scans around individual trees to ensure point cloud completeness, but motion distortion resulting from such multi-circular mobile acquisition has been generally overlooked. Based on Popper’s falsificationism, this study employs three methods: tree parameter comparison, optical porosity analysis, and CloudCompare-based point cloud distance comparison. These methods are used to observe that motion distortion significantly impairs the accuracy of individual tree point cloud data. Results indicate that motion distortion increases point cloud volume and Euclidean distance errors. Notably, a greater number of collection circles induces random variations in tree morphological parameters and a reduction in point cloud optical porosity. Consequently, tripod-mounted LiDAR is recommended over handheld/backpack devices. Significantly, this study is the first to highlight the potential substantial impact of accumulated point cloud motion distortion on landscape tree point cloud data acquisition. It provides a theoretical foundation for enhancing the precision of dynamically acquired tree point cloud data, which in turn improves the accuracy of tree analysis, monitoring, and assessment studies based on tree point cloud data.

Abstract: Continuous monitoring of permafrost dynamics remains challenging in remote high-mountain environments due to logistical constraints, harsh climatic conditions, and the limited availability of spatially distributed observations. In addition to direct temperature observations in boreholes, Autonomous Electrical Resistivity Tomography (A-ERT) offers significant potential for long-term monitoring by providing high temporal resolution observations of subsurface electrical properties, which is highly sensitive to freeze/thaw processes. This study presents the field validation of a low-power A-ERT system designed for long-term autonomous operation in extreme environments. The system was deployed at a high-altitude permafrost site near the Kumtor gold mine in the Central Tien Shan, Kyrgyzstan, representing the first application of continuous A-ERT monitoring in the Central Asian mountain ranges. The system operated continuously under harsh environmental conditions with air temperatures as low as −30 °C. Data quality remained consistently high throughout the monitoring period, with less than 1% of measurements removed during filtering, and inversion results with root-mean-square errors generally ranging between 3% and 4%. Time-lapse resistivity observations revealed strong seasonal freeze–thaw dynamics within the active layer and continued seasonal resistivity variations within the underlying permafrost despite permanently frozen conditions. Analysis of depth-dependent resistivity–temperature relationships revealed increasingly pronounced hysteresis behavior below the active layer, indicating that subsurface electrical properties were not controlled solely by temperature. This behavior likely reflects variations in unfrozen water content and pore connectivity within the fine-grained permafrost, where liquid water can persist at sub-zero temperatures. The results demonstrate the capability of the A-ERT system for reliable long-term autonomous monitoring in remote permafrost environments and investigation of coupled thermal and hydrological processes in permafrost systems.

Abstract:

High-altitude lakes are increasingly exposed to nutrient enrichment, organic loading and wastewater-derived contamination, while predictive performance can be overstated when temporally ordered observations are randomly partitioned or target-defining measurements are reused as predictors. We analyzed 177 consecutive monthly water-quality records from the Inner Bay of Lake Titicaca, Peru (January 2011–September 2025), integrating Carlson’s composite trophic state index (CTSI), Hamed–Rao modified Mann–Kendall tests, block-bootstrap Sen slopes, Pettitt change-point detection with 12-month block permutation, nutrient stoichiometry, correlation, principal component analysis, leakage-aware contemporaneous classification and one-month-ahead forecasting. Total phosphorus was treated as elemental P and phosphate as PO4. The bay remained chronically hypereutrophic (mean CTSI 74.08 ± 4.07; 83.1% of months). BOD5 increased by 0.548 mg L⁻¹ yr⁻¹, chlorophyll a by 3.254 mg m⁻³ yr⁻¹, total suspended solids by 0.752 mg L⁻¹ yr⁻¹ and conductivity by 13.13 µS cm⁻¹ yr⁻¹, whereas total phosphorus declined by 0.085 mg P L⁻¹ yr⁻¹. Block-supported shifts occurred in chlorophyll-a in November 2016 (21.81 to 54.18 mg m⁻³) and BOD5 in June 2018 (6.54 to 11.77 mg L⁻¹). After removing target-defining variables and preserving temporal order, the best trophic-state classifier had balanced accuracy 0.575 and MCC 0.257. Organic-pollution classification had balanced accuracy 0.624 but sensitivity only 0.271, whereas the fecal-indicator model was unstable (MCC 0.130; ROC-AUC 0.460). Persistence was the best BOD5 forecast (RMSE 4.193 mg L⁻¹; R² 0.390). The best CTSI model explained only 4.8% of future variance, and all thermotolerant-coliform forecasts had negative out-of-time R². Persistent ecological degradation was therefore evident, but monthly observations alone were insufficient for deployment-ready early warning. Higher-frequency sensing, hydrometeorological and wastewater-load covariates, spatial replication, direct microbiological measurements and prospective validation are required.