Environmental and Earth Sciences

Abstract: Retrieval of soil organic carbon (SOC) from uncrewed aerial vehicle (UAV) hyperspectral imagery depends on how laboratory measurements are linked to pixels at centimetre resolution. This study presents a quality-controlled workflow in which radiometric calibration, harmonisation against proximal spectroradiometer measurements, pixel-level screening, spectral support definition and model validation are treated as interdependent components of a single procedure. Imagery was acquired with a tunable Fabry–Pérot interferometer camera in six bands between 670 and 1100~nm over eight bare soil fields in Flanders, Belgium (85 topsoil samples) and five fields in Central Macedonia, Greece (42 samples). Single pixel spectra, neighbourhood aggregation and local pixel augmentation were compared at radii of 0.25, 0.50 and 1.00~m using Random Forest, XGBoost, a 1D convolutional neural network and TabPFN, with all pixels from a given sampling location assigned to the same cross-validation fold. In Flanders, 23 of 24 neighbourhood configurations reduced RMSE relative to single pixel extraction, the best reaching \(R^2 = 0.784\) and RMSE =0.155% SOC. In Central Macedonia, adverse acquisition conditions and a narrow target range prevented a usable SOC model, although electrical conductivity, CaCO₃ and pH reached \(R^2\) between 0.61 and 0.69. These results indicate that the spectral support should be defined to match the field sampling support, and that this correspondence belongs alongside radiometric calibration, pixel screening and validation in a reproducible workflow for field-scale soil mapping.

Abstract: Industrial parks and company growth are vital for urban planning and development. With temporal data collection, mathematical model based growth predictions of industrial parks becomes feasible. In this study, an industrial park Land development Simulation and Company Growth Projections (LSGP) framework is proposed by applying satellite data for land development simulation and social big data for company growth analysis. The cellular automata model is developed to simulate land use development, and the logistic function is applied to set up company growth projection model for selected industrial parks. Results show that the LSGP framework can effectively simulate industrial park land expansion and growth patterns with the following main findings: (1) Underlying land growth follows similar patterns, and key influencing factors are similar, such as traffic convenience, adjacent to urban centers; (2) Company growth pattern follows a logistic growth curve, with each subgroup of companies following the same pattern; (3) Models developed based on a selected industrial park can be transferred to another, i.e., case-based studies can be applied to newly planned industrial parks; (4) The LSGP framework may be further developed to simulate various land use projects and emulate growth patterns for additional areas of interest.

Abstract: Analysing the spatial and temporal variability of near-surface air temperature (T_a) is essential for assessing climate change, particularly in polar regions such as the maritime Antarctic, where observational studies remain comparatively scarce. In areas with sparse meteorological networks, satellite observations provide a valuable means of extending station-based analyses and generating spatially distributed Ta maps. This study analyses updated T_a records from six meteorological stations on the Hurd Peninsula, Livingston Island, for 2000–2024, with particular attention to temperature changes since 2016, and evaluates T_a estimation models based on MODIS Terra daytime LST Collection 6.1 data. The analysis of the T_a time series showed significant warming over the complete 2000–2024 period, with trends ranging from 0.53 to 1.27 °C decade⁻¹, and a clear shift towards warming after 2016. While the 2000–2015 period showed spatially contrasting trends, including cooling at JCI, positive trends were observed at all analysed stations during 2016–2024, ranging from 0.40 to 1.33 °C decade⁻¹ and highlighting substantial local variability across the Hurd Peninsula. The global Ta model, developed using data from all six stations and based on MODIS Terra daytime LST and spatiotemporal variables, achieved a performance comparable to that obtained with Collection 6, with R² = 0.77 and a cross-validated RMSE of 2.23 °C. The results also showed that rigorous filtering of the MODIS LST observations is essential: retaining only good-quality data and excluding observations with |view angle| > 45° substantially reduced dispersion in the Ta–LST relationship and removed unrealistically low LST values. The introduction of this view-angle threshold represents a new filtering step with respect to our previous studies.

Abstract: Hyperspectral Image Classification is an importanttopic in remote sensing with numerous applications in precision agriculture, environmental monitoring, defence, and more. It helps classify material composition by mapping spectral signatures from each pixel. Despite its potential, it faces persistent challenges due to the scarcity of labelled data and high dimensionality. This makes it difficult to achieve robust performance that can generalize well. To address this problem, we propose SRSSF (Stability-Reliability Spectral-Spatial Fusion), a hybrid mixture-of-experts framework for hyperspectral image (HSI) classification. It fuses a compact, stability-selected spectral representation with full-spectrum spatial and joint spectral-spatial descriptors (extended morphological, attribute-like, and Gabor texture profiles), combined through validation-derived reliability weights rather than fixed or heuristic fusion rules. SRSSF is evaluated under a strict, footprint-level spatially disjoint protocol in which entire patch receptive fields are guaranteed non-overlapping across training, validation, and test partitions, not mere center pixels. It has an automated, geometry-only (label- and accuracy-independent) seed-acceptance procedure and requires a minimum dataset-specific test-set size for training. The framework is benchmarked against fourteen classical and CNN spectral/spatial baselines and a dedicated ablation suite, across four benchmark datasets (Indian Pines, Pavia University, Salinas, and the Kennedy Space Center (KSC)). Our framework achieves the best OA, AA, and κ on every dataset, with OA scores of 56.43,91.94,90.50, and 75.96 at Indian-Pines, Pavia University, Salinas, and KSC, respectively.

Abstract: Total suspended matter (TSM) is a key coastal water-quality indicator, yet the operational GOCI-II retrieval algorithm is constrained by an empirical ceiling of approximately 28.13 g m⁻³ because of the inherent dynamic-range limitation of visible band-ratio formulations in the current operational product; this limitation precludes accurate monitoring of turbid nearshore environments, which is critical for ecosystems and coastal management. Here, we developed and evaluated a physically interpretable stacking ensemble (PISE) model that fuses three complementary base learners via a ridge meta-learner on five physically interpretable GOCI-II Rᵣₛ (555, 620, 709 nm) features trained on 212 matchup samples (2022–2025). On an independent held-out test set, PISE achieves R² = 0.793 and RMSE = 3.670 g m⁻³, improving the operational GOCI-II product by 8.9 percentage points in R² and 16.3% in RMSE, with 90% conformal prediction intervals achieving 90.6% empirical coverage. On 4 May 2025 GOCI-II imagery, PISE captures the Gyeonggi Bay flood-ebb TSM cycle, with a +0.006 g m⁻³ per cm regression against the KHOA tidal-stage gauge. PISE thereby provides a quantitative satellite-derived characterization of diurnal TSM dynamics across the flood-ebb tidal cycle over Korean coastal waters, extending operational TSM retrieval above the 28.13 g m⁻³ retrieval ceiling up to its trained limit of 50 g m⁻³.

Abstract: Periodic condition assessment of Natura 2000 (N2K)—the EU's network of protected areas under the Habitats Directive—is legally mandated but cannot be sustained at the required spatial scale through field survey alone. We introduce a multi-country Leave-One-Park-Out (LOPO) evaluation framework for binary per-habitat segmentation of N2K grasslands from Sentinel-2 multitemporal imagery, implemented using a parameter-efficient adaptation of the Prithvi-EO-2.0 geospatial foundation model. The framework spans four protected areas across Spain, Portugal, and France and three European biogeographic regions. Low-Rank Adaptation (LoRA) adapters (r=16, 6.3 M trainable parameters) fine-tune the frozen ViT-L backbone alongside a UPerNet decoder, ingesting three stochastically sampled monthly composites per training iteration. LOPO evaluation on three habitat configurations—semi-natural calcareous and Nardus grasslands (6210+6230), lowland and mountain hay meadows (6510+6520), and thermo-Mediterranean grasslands (6220)—assesses cross-park generalisation across ecologically distinct parks, with mean LOPO mIoU of 45.8% for 6210+6230 and 52.2% for 6510+6520 following a controlled data-expansion experiment that increases habitat-positive supervision by a factor of 25.3. A same-architecture within-park baseline inflates mean mIoU by 14.0 pp, directly quantifying the optimistic bias of conventional single-site evaluation.

Abstract: This study presents a multi-instrument remote sensing approach to characterize the optical properties and vertical distribution of aerosols within the planetary boundary layer (PBL) over New York City during the COVID-19 pandemic and post-pandemic periods. The analysis focuses on changes in aerosol loading associated with variations in anthropogenic activity and examines the relationship between column-integrated aerosol optical properties, vertically resolved aerosol distributions, and near-surface fine particulate matter (PM2.5). Continuous ceilometer observations are used to characterize profiles of attenuated backscatter and to investigate aerosol vertical structure and variability within the PBL. These measurements are complemented by sun-photometer observations of column-integrated aerosol optical properties, radiosonde profiles of atmospheric thermodynamic structure, and ground-based PM2.5 measurements. The synergistic use of these datasets provides a more comprehensive characterization of aerosol optical and physical properties under varying meteorological and boundary-layer conditions than can be obtained from individual instruments alone. In addition, air-mass back trajectories calculated using the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model are employed to investigate the potential influence of local, regional, and long-range aerosol transport on observed aerosol variability over New York City. Comparisons between the pandemic and post-pandemic periods reveal substantial differences in aerosol loading and pollution levels, highlighting the combined influence of changes in anthropogenic emissions, PBL dynamics, meteorological conditions, and transported aerosol on the urban aerosol environment. The results demonstrate the value of integrating vertically resolved active remote sensing with columnar and surface observations and atmospheric transport modeling for assessing temporal changes in urban aerosol distributions and their controlling processes.

Abstract: Reliable shoreline monitoring is essential for understanding how river deltas respond to changing sediment supply, marine processes, and human activities, and for supporting coastal management. However, accurately locating complex land–water boundaries and maintaining consistency across satellite sensors remain challenges that conventional segmentation metrics alone cannot adequately assess. This study developed MFGR-DeepLabV3+, a lightweight deep-learning model combining progressive feature fusion with boundary refinement to improve shoreline delineation in the Yellow River Delta. Landsat and Sentinel-2 imagery were used to reconstruct instantaneous waterlines for eight epochs between 1990 and 2025, followed by positional validation, cross-sensor comparison, and change analysis using the Digital Shoreline Analysis System (DSAS). Compared with the baseline, the model increased boundary intersection over union by 4.08–4.44 percentage points and reduced average symmetric surface distance by 18.9–43.2%, with benefits concentrated in complex shoreline settings. Validation against manually interpreted shorelines yielded root mean square errors of 18.98–38.13 m, while near-synchronous cross-sensor comparisons showed small mean biases but persistent local discrepancies. Across the study area, 60.09% of transects exhibited net retreat, with a median displacement of −503.51 m despite a positive regional mean. Strong local progradation at the active river mouth contrasted with persistent erosion along Laizhou Bay and the abandoned Diaokou River sector, while engineered shorelines showed mixed responses. By linking boundary-focused learning, positional reliability assessment, and long-term geospatial analysis, this framework strengthens AI-supported shoreline monitoring and provides evidence for identifying priority erosion-monitoring sectors.

Abstract: Assimilating satellite-retrieved profiles is an important approach to improving numerical model initial conditions and typhoon forecasting skill. This study uses temperature and humidity retrieved profile products from the Geostationary Interferometric Infrared Sounder (GIIRS) onboard Fengyun-4B for three typhoons that affected Zhejiang Province in July 2025 with different track types. Sensitivity experiments are conducted under clear-sky, cloudy-sky, and all-sky conditions. Vortex relocation experiments are designed to separate the contributions of typhoon vortex structure and environmental field to track forecasts. Results show that assimilation systematically adjusts temperature and humidity in the typhoon inner core, rainband, and outer environment. All schemes produce lower–middle tropospheric warming and upper tropospheric cooling. Humidity decreases in the low-level outer environment but increases around the typhoon vortex and along the Western Pacific Subtropical High edge. All-sky assimilation is optimal for track forecasts, whereas clear-sky or cloudy-sky assimilation alone can be adverse. For 72-h precipitation, all-sky assimilation increases the heavy rain threat score by 0.09 versus control and the torrential rain threat score by 0.09 versus cloudy-sky assimilation. This study reveals the comprehensive advantages of all-sky GIIRS profile assimilation for typhoon forecasting and supports the operational application of GIIRS-retrieved temperature and humidity profiles.

Abstract: Accurate image correspondences are essential for geometric inference in vision-based UAV sensing because camera pose, 3D reconstruction, and spatial measurement depend on the geometric validity of matched points rather than on match count alone. This study develops a hierarchical Fourier–Mellin/phase-correlation framework for sequential RGB imagery acquired by a UAV camera sensor. Global similarity registration is followed by recursive local refinement with spectral, forward–backward, overlap, and radiometric acceptance criteria. Performance was assessed on 338 held-out stereo pairs using independently bundle-adjusted camera geometry as an external reference and a common relative-orientation pipeline for all methods. The proposed method achieved 63.91% matching success and 60.06% reference-valid success, compared with 89.05% and 27.81%, respectively, for Area-LSM. For successfully oriented pairs, median relative-rotation and baseline-direction errors were 0.0322° and 3.7987°, respectively. On common successfully oriented pairs, the proposed approach yielded lower errors than Area-LSM and ORB, whereas SIFT, Harris, Förstner, Moravec, and FAST generally retained higher success rates and lower errors. Baseline-direction accuracy was strongly dependent on parallax and homography dominance. These results support independent sensor-geometry validation and geometry-aware hybrid matching for UAV vision-based sensing.

Abstract: Urban flooding recurs in the same low-lying districts of coastal megacities every rainy season, and Lagos, Nigeria, is no exception. Existing flood risk maps for the city are fragmented across single-hazard studies, rely on expert-assigned weights never checked against observed flooding, and offer no explanation for why one area is classified as riskier than another. This study developed and validated a three-layer geospatial disaster intelligence framework, GeoAID, for Amuwo Odofin Local Government Area. Fourteen satellite-derived conditioning factors and a Sentinel-1 SAR-derived flood inventory were assembled via Google Earth Engine. Logistic Regression, Random Forest, and XGBoost were trained and compared; Random Forest was selected, achieving 0.798 ROC-AUC under random train-test splitting, falling to 0.642 under spatial five-fold cross-validation, a gap reported rather than concealed. SHAP identified vegetation cover (NDVI) as the dominant predictor, converging with an independent city-scale land-cover study using unrelated data and methods. The resulting map showed 229,402 residents (61.1% of the LGA's population), 16 of 35 schools, and 55 of 102 health facilities within High or Very High risk zones. All outputs were deployed in a public dashboard linking structural risk to near-real-time rainfall. The framework demonstrates that explainable machine learning on sparse, SAR-derived training labels can produce validated, auditable flood risk information in a data-scarce urban setting, without requiring ground-truth flood records.

Abstract: Ground-point classification (GPC) is a fundamental step in LiDAR data processing because classification errors propagate into digital terrain models (DTMs), height normalization, vegetation structure metrics, and other derived products. Existing GPC algorithms depend on user-defined parameters whose effectiveness can vary with terrain complexity, vegetation structure, sampling density, and sensor-platform acquisition geometry. We present FAST-GC, a fully automated and self-tuning GPC algorithm that does not require scene-specific parameter calibration. FAST-GC was evaluated through complementary experiments encompassing a controlled six-site comparison with established ground-filtering algorithms, terrain-complexity and slope stress testing, systematic point-density reduction across 20 ALS sites, cross-platform ALS–ULS–TLS assessment, and large-area processing. Across the controlled comparative sites, FAST-GC maintained high ground completeness while limiting commission error, with its advantage over the established algorithms becoming increasingly pronounced as terrain complexity increased. At slopes of 70–80°, FAST-GC achieved a DTM mean absolute error (MAE) of 1.00 m compared with 3.12–12.75 m for the evaluated algorithms. Extended assessment across 20 ALS sites further demonstrated stable performance under systematic sampling-density reduction: reducing median density from 9.84 to 3.84 pts./m² maintained F1 ≈ 0.97, intersection over union (IoU) ≈ 0.95, and Matthews correlation coefficient (MCC) ≈ 0.85, while DTM MAE increased only from 0.07 to 0.11 m. Independently reconstructed 0.5 m terrain surfaces from ALS, ULS, and TLS observations showed strong cross-platform correspondence despite substantial differences in acquisition geometry and sampling conditions (R² and concordance correlation coefficient ≥ 0.99; RMSE = 0.10–0.16 m). Large-area processing of 382 ALS tiles (~23 km²) further demonstrated stable tile-wise execution and automated terrain-product generation. Collectively, the results show that GPC performance is governed not only by point-wise classification agreement but also by the ability to preserve terrain morphology under increasing topographic complexity, reduced sampling support, and contrasting acquisition geometries. By adapting terrain-support estimation internally rather than requiring scene-specific parameter calibration, FAST-GC provides a reproducible and scalable framework for heterogeneous and multi-platform LiDAR processing.

Abstract: Earth observation (EO) is entering a petabyte-scale era in which multi-sensor satellite archives capture environmental processes across broad spatial, spectral, and temporal dimensions. EO data-processing workflows, however, still rely heavily on task-specific supervised models that require large amounts of labeled data and may generalize poorly when applied to new regions, time periods, sensors, or environmental conditions. This review provides a structured synthesis of the rapidly evolving literature on EO foundation models (EO-FMs), which seek to address these limitations through large-scale pretraining and efficient downstream adaptation. We organize the field of EO-FMs through a data-objective-architecture-adaptation-evidence lens that links training data and harmonization, pretraining objectives, backbone design, adaptation strategies, and empirical evidence across downstream applications. We examine representative model families including Prithvi-EO-2.0, Copernicus-FM, TerraFM, DOFA, SatCLIP, EarthPT, and Aurora, and compare their capabilities across tasks, regions, time periods, and sensing modalities. Across the literature, EO-FMs show strong potential for improving label efficiency, supporting multimodal learning, and enabling the reuse of pretrained representations across applications such as land-cover mapping, hazard response, ecosystem monitoring, urban climate analysis, and Earth-system forecasting. At the same time, empirical evidence on downstream performance, label efficiency, and generalization remains heterogeneous across datasets, sensing modalities, metrics, and benchmark protocols, which limits direct model ranking and underscores the need for task-aware interpretation. We argue that the next phase of EO-FM research should move beyond proof-of-concept gains toward more rigorous and capability-aware benchmarking, explicit assessment of predictive uncertainty and calibration, transparent preprocessing and adaptation procedures, and geographically and temporally robust evaluation. Meeting these requirements would strengthen the reliability, comparability, and scientific utility of EO-FMs for scalable and decision-relevant Earth observation analysis.

Abstract: Accurate and transferable land-use and land-cover (LULC) mapping in the Brazilian Cerrado is challenged by spatial and temporal heterogeneity, domain shift, and the high cost of expert annotation. This systematic review with qualitative methodological synthesizes evidence on Active Learning (AL), domain adaptation, remote-sensing foundation models, and Remote-Sensing Vision–Language Models (RS-VLMs), focusing on their complementary roles in annotation-efficient and transferable LULC mapping. Evidence was synthesized qualitatively because the reviewed studies differed substantially in sensors, tasks, datasets, domain-shift conditions, and evaluation protocols. The literature indicates that uncertainty-, diversity-, and representativeness-based AL can improve annotation efficiency, while foundation models and RS-VLMs provide increasingly transferable representations and few-shot or open-vocabulary capabilities across remote-sensing applications. However, direct empirical evidence integrating transferable representations, AL, and domain adaptation for Cerrado LULC monitoring remains limited. Based on the convergence of these evidence streams, we propose a conceptual human-in-the-loop framework that combines transferable representations with selective target-domain annotation and subsequent adaptation. The framework is hypothesis-driven rather than experimentally validated. The review identifies priorities for empirical evaluation, including multimodal benchmarks, standardized AL protocols, spatially and temporally independent validation, explainability, and continuous human-in-the-loop adaptation. This synthesis provides a methodological basis for future domain-adaptive LULC monitoring in the Brazilian Cerrado.

Abstract: Mobile mapping systems (MMS) are moving from sensor-centric survey platforms toward intelligencedriven components of the remote sensing ecosystem. This review analyzes rather than catalogs that shift, organizing it around four structural trends: the replacement of per-frame geometric optimization in simultaneous localization and mapping bylearned, feed-forwardgeometry, arrangedalongalatencyfidelity axis; the emergence of tokenization economics—how points become, and are pruned from, attention tokens—as the governing principle of 3D perception efficiency; the migration of radiancefield reconstruction from visualization toward measurement, where explicit representations such as 3D Gaussian Splatting yield updatable digital twins; and the transition from perception to agency, coupling MMS evidence to Earth foundationmodels. Eachtrendistested against the two environments where mobile mapping is most deployed, city and forest, whose contrasting geometry and satellite visibility expose how far each advance generalizes. Seven recurring pitfalls are then consolidated, among them efficiency claims measured on heterogeneous hardware, parameter count used as an efficiency proxy, and architecture rankings assumed to transfer between domains as dissimilar as urban corridors and forest plots. Controlled same-hardware benchmarks supply measured evidence for the efficiency pitfalls, and a research agenda for the photogrammetry and remote sensing communities follows.

Abstract: The 2019–2020 Black Summer fires affected approximately 33.5% of Kosciuszko National Park, Australia's largest alpine reserve; post-fire recovery models in Australia have rarely used spatial validation. We quantified recovery across severity classes and elevation zones (2021–2025), tested whether driver importance shifts over time, and whether predictors generalise to unseen locations. A Sentinel-2 Normalized Difference Vegetation Index (NDVI)-derived recovery index was modelled with Random Forest at one, three and five years post-fire using twelve predictors and random/spatial cross-validation; SHapley Additive exPlanations (SHAP) importance used bootstrap confidence intervals, and recovery was cross-validated against the Normalized Burn Ratio (NBR). Severity's importance declined sharply (mean |SHAP| 0.116 to 0.019); elevation remained important throughout. Fast-recovery clusters occurred at lower elevation (877 versus 1284 m; P < 0.001), and montane and subalpine trajectories were statistically distinguishable. NDVI overstated recovery relative to NBR every year (P < 0.001). Spatial cross-validation R² stayed near zero regardless of block resolution, versus 0.10–0.16 under random splitting. Fire severity's influence is transient; elevation-mediated differences persist over at least five years, and static covariates cannot predict recovery at unseen sites. Post-fire monitoring should be reassessed periodically rather than prioritised on severity alone; spatial cross-validation should become standard for spatially autocorrelated recovery data.

Abstract: Forest and land fires in Southern Riau, Indonesia, vary seasonally with environmental and landscape conditions. This study evaluated whether separate seasonal models reveal susceptibility patterns not captured by an all-season model and whether preceding environmental change adds predictive information. All-season, dry-season, and rainy-season models were developed from multi-source satellite and geospatial data for 2010–2024 using Logistic Regression, Random Forest, XGBoost, and LightGBM under condition-only and condition-change configurations. Models used a 1:5 fire-to-non-fire ratio, a stratified 70:30 train–test split, and PR-AUC as the primary metric. LightGBM was selected for all periods, with held-out PR-AUC values of 0.921, 0.905, and 0.966, respectively. Historical change improved PR-AUC by 0.008 for all-season and 0.009 for dry-season but decreased it by 0.001 for rainy-season, where condition-only was retained. All-season and dry-season maps were more spatially similar to each other than to the rainy-season map, whose model-derived contributions were led by wind speed, vegetation condition, and peat-related context. Independent 2024 POLRI records showed that 77.60%, 76.79%, and 78.57% of matched incidents fell within High or Very High susceptibility classes. Overall, separate seasonal modelling revealed patterns not fully represented by the all-season model, while the value of preceding environmental change differed among periods.

Abstract: Satellite remote sensing, machine learning, ecological monitoring, and household-level economic analysis increasingly provide complementary information on pastoral systems. Yet indicators in these domains are often interpreted as measuring the same underlying construct. Grassland gross primary productivity (GPP), vegetation condition, soil organic carbon (SOC), livestock production, value capture, and household welfare are related but distinct outcomes. This perspective develops the Productivity, Condition, Carbon, and Livelihood (PCCL) framework for Mongolian rangelands with explicit inference boundaries. The framework distinguishes six analytical components: ecological productivity, rangeland condition, soil carbon and related belowground processes, livestock production, value capture efficiency, and household welfare and resilience. It also separates measurement, prediction, statistical association, and causal inference. Recent studies in Mongolia that use remote sensing, machine learning, panel data, soil carbon information, and household outcomes illustrate different parts of this system rather than independently replicating a single relationship. Broader evidence shows that climatic variability can dominate livestock effects on primary productivity at some temporal scales, while grazing effects on SOC vary with intensity and environmental context. Accordingly, high remotely sensed productivity should not be interpreted automatically as healthy rangeland, increased SOC, or higher pastoral income. The paper introduces no new data or empirical estimates. Instead, the paper provides an organizing architecture for integrating ecological and economic evidence, preserving variable provenance, limiting interpretation to what research designs support, and generating testable propositions for future research across multiple scales.

Abstract: Phragmites australis (common reed) is a resilient wetland species providing key ecosystem services, including carbon sequestration. However, its carbon storage potential in hyper-arid environments remains insufficiently quantified. This study develops an integrated framework combining destructive sampling and remote sensing to assess biomass distribution and carbon stocks in Oman. Results show that destructively measured AGB correlates only weakly with reed height (R² = 0.15), indicating that height alone is not a reliable standalone predictor of biomass in this system; an initially high fit (R² ≈ 0.99) reflected the mathematical dependence of the height-based allometric equations on their own input variable rather than genuine predictive validity. While predictive models based on dry biomass and TOC showed moderate to low performance (R² < 0.16), vegetation indices derived from remote sensing demonstrated high accuracy in estimating spatial biomass variability. Carbon stock estimates ranged between ~0.36 and 1.95 kg/m² depending on the method used. While localized allometric models showed limited predictive power due to high variability in tiller density, vegetation spectral indices demonstrated high accuracy (R² = 0.78), highlighting remote sensing as a viable tool for spatial carbon assessment in arid wetland ecosystems. These findings highlight the potential of integrating field-based and remote sensing approaches to improve carbon accounting in arid wetlands. The study provides a scalable framework for ecosystem-based carbon assessment and supports the role of P. australis as a nature-based solution for climate mitigation in water-stressed regions.

Abstract: India has not held a decennial census since 2011, leaving the 2015–2025 decade without a population count against which claims of small-town urbanisation can be verified. This study tests whether such growth is empirically demonstrable along a sixty-kilometre stretch of National Highway 63 in Jagtial district, Telangana, covering three corridor towns: Jagtial, Korutla and Metpally. Sentinel-2A and LISS-4 imagery for December 2015, 2020 and 2025 were segmented using Large Scale Mean Shift and classified with nine independently trained Random Forest models, achieving overall accuracies of 72.73 to 90.59 per cent. Results were cross-checked against electricity, motor-vehicle and beedi-establishment records sharing no error structure with the imagery. Built-up area grew 80.2 per cent at Jagtial, 62.6 per cent at Korutla and 31.5 per cent at Metpally between 2015 and 2025, a gradient tracking administrative rank rather than a corridor-wide wave; Korutla combines the fastest growth in service connections and load intensity with the weakest formal streetlight-network expansion, consistent with its home-based beedi cottage-industry base. An XGBoost-anchored trend model extends the built-up, electricity and vehicle series to 2040, corroborated by the Union Cabinet's June 2026 approval of highway widening for the corridor, expressly justified by congestion from this built-up growth.