Earth and Environmental Sciences

Embedding datasets encode complex relationships among multiple sources of Earth observation data into a compact format. Here, we evaluated the utility of a 10-m global Satellite Embedding product (SE) for classifying crop types in central California for the year 2020. We compared the classification accuracy of a random forest model based exclusively on the SE layer to an existing random forest model with multiple imagery inputs. Our results showed the SE-based classification had higher agreement with the reference dataset (California Department of Water Resources crop map) than the classification based on Landsat and National Agricultural Imagery Program inputs (94.7% versus. 91.9% overall accuracy, respectively). The performance of individual crop types was consistent across models, ranging from high agreement for rice (98.4% versus 98% accuracy) to lower agreement for pasture, grain, and fallow/young perennial classes (< 65% accuracy in both models). The SE-based workflow used three times less cloud-based computational resources and represented substantial savings of predictor development time. Geospatial embedding products can aid classification efforts by reducing predictor development time and processing demands while maintaining classification accuracy.

Forest canopy height (FCH) is a key parameter for estimating forest carbon stocks and sequestration. The Global Ecosystem Dynamics Investigation (GEDI) satellite emits laser pulses, which obtain vertical forest structure information. To solve the difficulties in estimating FCH in complex mountainous areas and the low accuracy of regional scale estimation, the GEDI LiDAR, ZY-3 stereo imagery and OHS hyperspectral data were integrated to estimate FCH of Jiyuan City in China. The spectral, textural and topographic variables derived from OHS data were selected using the method of variance inflation factor (VIF) and the Generalized Additive Model – Least Absolute Shrinkage and Selection Operator (GAM-LASSO), which aimed to improve the estimation accuracy and precision, and the variables were used to construct three machine learning models including random forest (RF), bidirectional recurrent neural network (BRNN) and convolutional neural network (CNN) to estimate FCH. Results indicate that multi-source data (GEDI and ZY-3) with RF method has the highest estimating accuracy (R2 = 0.529, RMSE = 3.701 m), followed by BRNN (R2 = 0.508, RMSE = 3.782 m) and CNN (R2 = 0.494, RMSE = 3.836 m). These results demonstrate that the integration of LiDAR, OHS and spectral remote sensing data is an effective approach to improving the accuracy of FCH estimation in mountainous areas.

Earthquake Early Warning Systems (EEWS) are now operational across multiple jurisdictions, including Canada, issuing messages designed to trigger protective behaviour such as Drop-Cover-Hold-On in the seconds before damaging waves arrive. Because most recipients will have no prior experience of an EEW alert, the experiential background they bring to a first alert is general earthquake comprehension, the phenomenon this study examines. Existing models frame recipients as rational decision-makers, yet the empirical record of lived earthquake experience suggests that the early seconds are dominated by confusion, and what recipients actually need from messaging in those moments has remained underexamined. Through a phenomenological-linguistic analysis of nearly 1,000 first-hand accounts from the United States Geological Survey’s Did You Feel It (DYFI?) survey across five North American earthquakes, this study traces how respondents describe and make sense of their experiences. Confusion was the dominant emergent feeling and was almost always paired with active resolution efforts; the patterns yielded a four-phase Earthquake Comprehension Cycle that converges with foundational phenomenological and perceptual cycle models. EEWS messages function primarily as early comprehension aids, interjecting into the cycle ahead of ground shaking and supplying a theory that the recipient would otherwise need to generate themselves. The reframing applies broadly to other early warning systems facing confusing emergent events and offers a phenomenological lens for risk communication design.

H) remained virtually invariant throughout the simulation. Crucially, these ratios precisely matched authentic field residues, including 10-year archived samples from the 2008 Morning Sun incident and dynamically weathered residues from the 2017 Green Island spill. These findings establish C30-hopane as a highly recalcitrant, conservative internal standard for quantifying NAPL mass transfer in complex coastal systems. Furthermore, the systematic depletion of alkylated polycyclic aromatic hydrocarbons (APAHs) marks a kinetic shift from initial physical sequestration to chemical and microbial oxidation. This oxidative transition governs the leaching of water-soluble, toxic oxygenated PAHs (OPAHs) into underlying coastal aquifers. Ultimately, our results provide a robust, tiered forensic framework to calibrate groundwater transport models and predict the environmental persistence of legacy petroleum sources at vulnerable maritime-terrestrial interfaces.

The utilisation of microwave radiances is crucial for enhancing the precision of weather forecasts. Despite existing uncertainties over land and ice-covered surfaces, recent advances have enhanced their use. This study examines the impact of assuming either Lambertian or specular surface reflection on the simulation of brightness temperatures for surface-sensitive, clear-sky AMSU-A microwave radiances across land and snow-covered areas. It represents the preliminary work before running a full assimilation and forecast impact study. Using the high-resolution HARMONIE-AROME regional modelling system, experiments were conducted to retrieve and analyse the retrieved emissivity in different conditions/seasons. The emissivity was also used as input to the radiative transfer model to simulate brightness temperatures of surface-sensitive sounding channels. The results show that the Lambertian assumption produces higher variability in dynamic surface emissivity, while the specular approach yields smaller and more consistent deviations. During winter, specular reflection shows higher first-guess departures (e.g., observations minus simulations) for surface-sensitive sounding observations, whereas in summer it performs better over land surfaces. Over snow-covered regions, the use of the Lambertian reflection to simulate the brightness temperature gives smaller mean errors for AMSU-A channels 4 (52.8 GHz) and 5 (53.59 GHz). These findings encourage further investigation into implementing a parameter that accounts for the Lambertian component of surface reflection when simulating brightness temperature in high-resolution limited-area models. Additionally, these findings provide practical guidance for configuring complex Nordic surface regional models and for future Arctic Weather Satellite microwave radiance assimilation.

Abstract The agriculture sector has consistently attracted substantial attention due to its high vulnerability to climate change in pre-industrial societies. This study examines the role of land reclamation in stabilising rice prices and improving the resilience of the rice market to climate change, focusing on Guangdong Province during the 18th century. Using a variety of statistical tools, the results indicate since land reclamation could limitedly alleviate climate-induced stresses, the agrarian economy of Guangdong Province remained highly dependent on natural conditions. For the first time, this study emphasises the use of a mesoscale approach, which combines the quantitative and narrative methods, as an effective scalar mode for simultaneously identifying the impacts of climate change and human responses. Furthermore, the study contributes to a deeper understanding of industrial development in Guangdong Province during the Qing Dynasty by examining agricultural resilience to climate change, thereby encouraging additional studies on past industrialization from an environmental perspective.

Study region: Kizito Huonjwa Street in Kimbiji Ward, Kigamboni Municipality, is a peri-urban coastal settlement in Dar es Salaam, Tanzania. Study focus: This study evaluated how seasonal rainfall structure, moderate extremes, projected climate forcing, and engineering design were associated with simulated household rooftop rainwater harvesting system (RWHS) reliability. Gridded rainfall and gauge observations were combined with six scenario-complete regional climate model chains under RCP4.5 and RCP8.5. Seasonal diagnostics, peak-over-threshold (POT) analysis, empirical quantile mapping (EQM), deterministic daily water balance simulation, and Type II analysis of variance were used to distinguish local design responses from formal variance attribution. Different roof catchment areas, RWHS storage capacities, and household demands were evaluated. New hydrological insights for the region: Across the 864 scenario simulations, engineering design terms accounted for 75.7% of simulated reliability variance, compared with 18.9% for the included climate forcing terms. Storage capacity had the largest individual contribution within the fitted design, followed by household demand, climate model identity, and roof area. For the climate model-based stress testing results, the rainfall scenario median differences were smaller than the corresponding inter-modal spread. These results are conditional on the evaluated factor ranges, six-model ensemble, daily rainfall forcing, deterministic RWHS formulation, and the absence of validation against observed household tank operation.

Estuarine ecosystems located in tropical urban regions are increasingly exposed to cumulative anthropogenic and hydroclimatological pressures. This study assessed the environmental health of the Pina Basin estuary (Pernambuco, northeastern Brazil) through an integrated approach combining water quality assessment, polycyclic aromatic hydrocarbon (PAH) monitoring, and biomarker responses in resident mullets ( Mugil sp.) during dry and rainy seasons. Physicochemical analyses revealed severe environmental degradation, particularly during the rainy season, when dissolved oxygen levels reached critical hypoxic conditions (0.20 mg L −1 ). Although dissolved Σ16 PAH concentrations remained relatively low (<100 ng L −1 ), biological responses indicated persistent physiological stress. Micronucleus frequencies showed no significant seasonal variation (0.38% dry; 0.43% rainy; p = 0.719), suggesting chronic and continuous genotoxic exposure throughout the year. In contrast, hematological and biochemical biomarkers exhibited pronounced seasonal responses. Platelet counts decreased significantly during the rainy season ( p < 0.001), a pattern associated with handling stress and hypoxia-induced cellular aggregation rather than thrombocytopenia. Rainy-season lymphopenia was accompanied by significant reductions in aspartate aminotransferase activity ( p = 0.022) and seasonal variations in protein and lipid metabolism. Multivariate correlation analyses linked these physiological alterations primarily to severe hypoxia, nutrient enrichment, and diffuse urban runoff, rather than to isolated chemical contamination. The findings demonstrate that biomarker networks in wild mugilids are effective early-warning indicators of ecosystem degradation, reflecting the cumulative impacts of urban stressors in tropical estuaries. Integrating physiological and genotoxic biomarkers into environmental monitoring programs is therefore essential for improving coastal management and conservation strategies.

Abstract. The increasing availability of kilometer-scale climate simulations presents major challenges for data access, processing, and analysis due to the unprecedented volume and heterogeneity of the outputs. Different data formats, structures, and metadata conventions, require dedicated solutions to ensure interoperability and usability. We introduce AQUA (Application for QUality Assessment), a Python-based framework developed within the Climate Change Adaptation Digital Twin of the Destination Earth (DestinE) initiative, designed to support the automated evaluation of high-resolution global climate simulations. Although several diagnostic suites for the analysis of global climate model data are already available, AQUA provides a flexible and modular infrastructure for accessing and processing climate model output across various formats. By building on widely adopted Python libraries, it enables scalable, out-of-core computations. Its design supports integration into automated workflows and user-defined pipelines, facilitating both operational and research-oriented applications. This paper focuses on the architecture and core functionalities of the AQUA core, which handles data ingestion, standardization, and pre-processing. AQUA is open source and actively maintained, and aims to serve as a community tool for robust, reproducible, and efficient climate data analysis across projects and institutions.

Canopy Leaf Area Index (CLAI) is a stand attribute containing information on the real-time health and growth potential of managed pine plantations. Current remote sensing techniques for quantifying CLAI rely on simple linear models applied to satellite multispectral imagery, or on techniques based on light detection and ranging (LiDAR) data that are costly and less frequently collected. This study demonstrates a convolutional neural network (CNN) approach to retrieving CLAI from 10 m Sentinel-2 multispectral imagery with a model trained on gridded LiDAR-based CLAI estimates. We demonstrate large gains in accuracy with the CNN compared to traditional linear models based on vegetation indices (e.g., Simple Ratio), but also clear shortfalls in model skill when predicting “blind” in some spatial domains that were completely excluded during model training. Pixel-scale root mean squared error ranged from 0.34 to 0.64 by domain when exposed to CLAI training data from all available spatial domains, but rose to 0.58–1.74 when predicting without prior domain-specific training. Prediction accuracy was consistently lower when applied to completely unobserved USGS LiDAR-based CLAI estimates. Traditional linear models, in contrast, had the advantage of usually lower prediction error across unobserved spatial domains (0.43–1.98), but with lower maximum accuracy. These results demonstrate a potential route for deploying more complex models for LiDAR “mimicry”, e.g., between data acquisitions widely separated in time, but advocate for the development and use of more stable generalized approaches for use in unobserved managed pine stands.

Abstract Porous volcanic rocks form volcanic edifices and host key geothermal reservoirs worldwide. When subjected to constant stress above dilatancy in the brittle regime, these materials undergo time‐dependent failure known as brittle creep. In the ductile regime, where deformation is compactive, time‐dependent compaction—compaction creep—has been documented in sandstones but never in lavas. Here, we present the first experimental investigation of compaction creep in porous lava from Volvic, France. We performed triaxial constant strain rate and compaction creep experiments at an effective pressure of 100 MPa. Our results show that lava accumulates inelastic strain and loses porosity continuously with decreasing rate during creep deformation. Creep strain rates weakly depend on the applied stress, and microstructural observations reveal that time‐dependent compaction localizes into compaction bands. These findings demonstrate that compaction bands can form slowly over extended periods in lava with implications for the long‐term evolution of volcanic systems and geothermal reservoirs.

Abstract Seismic data for Mars has provided constraints on the thickness of its crust, but its density and how much it varies spatially, especially across the hemispherical dichotomy, is still undetermined. We apply a novel constraint in the analysis of spacecraft tracking data to determine Mars' gravity field that allows for the determination of the density of the north and south directly from the data. We find a density of kg for the north and kg for the south. This result is independent from the seismic results, and consistent with seismic constraints that indicate the difference between the north and south should not be greater than 200 kg . Accounting for the higher densities of volcanoes changes the difference slightly, still within the seismic constraint. Our density results provide additional, independent constraints for crustal thickness modeling and elucidating the origin of the dichotomy.

Abstract The atmospheric chlorine radical (Cl•) is a highly reactive oxidant. However, its formation on particulate matter (PM) surfaces and its role in nitrogen chemistry remain insufficiently understood. This study investigated the Cl• chemistry on photoactive PM and evaluated its potential influence on atmospheric nitrogen cycling. Experimental evidence confirms that the chloride ion (Cl–) on the photoactive surface undergoes oxidation by photogenerated hole (h+) to form Cl•, which subsequently accelerates nitrite intermediate conversion to gaseous nitrogen oxides (NOx). Under the model experimental conditions, this process enhanced NOx formation by approximately 40-fold, while Cl– was regenerated during the reaction. Additional experiments with different chloride salts and acidity controls indicate that the enhancement is mainly associated with Cl–-involved photochemistry. This mechanism maintains efficacy across various humidity conditions and can be extended to diverse photoreactive PM (e.g., soot) and different nitrogen-containing components (e.g., ammonium, 4-nitrophenol). Experiments with reduced photoactive matter loadings and simulated atmospheric PM further reveal that the reaction rate is regulated by PM composition and the abundance of photoactive matter. Environmental simulations suggest that this surface Cl• cycling process may contribute 0.7%–6.2% to NO2 variability in the study area. Overall, this study identifies a previously overlooked Cl•-mediated surface process on photoactive PM and highlights its potential contribution to tropospheric nitrogen transformations.

Leaf area index (LAI) is widely used to characterize foliage amount and seasonal canopy development, but it captures only selected aspects of forest structure and can be difficult to retrieve reliably in heterogeneous, multilayered stands. This study evaluates Sentinel-2-based LAI information across eight sites in the Slovenian Classical Karst encompassing post-disturbance regeneration and established forest stands in dolines and relatively level inter-doline terrain. Field effective LAI measured during six periods in 2021 was compared with six Sentinel-2 spectral variables, LAI derived using the Sentinel Application Platform (SNAP), and the Copernicus Land Monitoring Service High-Resolution LAI product. The analysis explicitly distinguished two dimensions of retrieval performance that are often conflated: seasonal fidelity within sites and preservation of structural differences among sites. Most satellite-derived variables and LAI products captured the broad phenological progression from canopy development to senescence. However, strong temporal agreement within sites did not consistently translate into preservation of the ordering or magnitude of structural differences among sites. Several methods compressed the range of high effective LAI values at dense regeneration sites with substantial lower-layer vegetation. The study therefore provides a more informative framework for evaluating LAI products by identifying which component of variation drives apparent agreement. These findings indicate that Sentinel-2 can support phenological monitoring and broad screening of post-disturbance vegetation development. However, quantitative comparisons of canopy density or structural recovery across heterogeneous stands require consideration of canopy heterogeneity, potential spectral saturation, and plot-to-pixel support. The evaluation framework and the observed retrieval limitations are relevant beyond karst forests, particularly to post-disturbance stands, fragmented forests, open woodlands, and sites with dense understory or regeneration layers.

Low-light conditions degrade remote sensing imagery by reducing contrast, distorting color, and obscuring fine terrain structures and small objects critical for Earth observation. Accurate illumination adjustment under spatially varying scene content remains challenging for existing enhancement methods, and many prior-guided approaches operate exclusively in either the spatial domain or the frequency domain. In this work, we propose a Dual-Domain Illumination Prior (DDIP), a trainable dual-domain illumination-prior module that is jointly optimized with each host backbone and exploits frequency-domain and spatial-domain illumination statistics. DDIP comprises three components: a Frequency-Domain Illumination Distribution Prior (FIDP) that performs per-color-channel amplitude calibration in Fourier space to improve global brightness; a Spatial-Domain Illumination Distribution Prior (SIDP), adapted from IDP-Net, that performs multi-scale sub-region statistical correction for local illumination adjustment; and a Selective Core Feature Fusion (SCFF) module that adaptively combines the frequency-domain output, the spatial-domain output, and the original input through an attention-based gating mechanism with dual pooling. DDIP is integrated with each host backbone while leaving its main restoration blocks unchanged. In the controlled reconstruction comparisons on iSAID-dark and the evaluated general low-light benchmarks, equipping the tested backbone networks with DDIP improves PSNR and SSIM over their corresponding baselines. Complementary LPIPS and CIELAB lightness measurements characterize perceptual similarity and lightness behavior, while a fixed-detector object-detection evaluation on the tested high-resolution iSAID-dark scenes examines the effect of the enhancement pipelines under the reported synthetic low-light conditions. The ablation studies further examine the contribution of the module components within the reported experimental settings.

Two types of extremes are often used as indicators of future change: Extreme climate events and extreme habitats. These extremes can be useful gauges of environmental change, but only if biological, spatial, and temporal scopes are aligned between the extreme and the future change it is hypothesized to mirror. In general, short-lived extreme events provide insights towards acute responses but are not as suitable for assessing long-term ecological or evolutionary change. Extreme habitats can reveal potential long-term adaptations, but community structure and trophic interactions can be challenging to assess due to species often following habitat-dependent evolutionary paths, and these environments tend to undergo gradual rather than rapid ecological change. Community structure and species interactions may further differ for both types of extremes between biological response levels. In cases of a mismatch between the size of an extreme habitat or event and its comparative system, it will be more appropriate to study biological responses at smaller geographic scales than the habitat or event occurs. Here, we explore the opportunities and limitations of a variety of extremes in marine and freshwater systems.

Atmospheric particulate matter (PM) is strongly influenced by both combustion and non-exhaust traffic emissions, yet their simultaneous molecular characterization over seasonal cycles remains rare in the French monitoring literature. This study addresses this question through the first year-round application of a suite of organic tire-related markers (TRMs) applied alongside polycyclic aromatic hydrocarbons (PAHs), monosaccharide anhydrides (MAs), and methoxyphenols (MPhs) in 36 weekly PM 10 samples collected at a semi-urban, traffic-adjacent site in Cronenbourg (northeastern France) over four seasons (winter 2023-autumn 2024), using a validated multi-residue method targeting 48 compounds. Combustion-derived PAHs (median Σ 13 PAHs 16.6 ng m -3 ) and organic TRM (median Σ 12 TRMs 4.5 ng m -3 ) showed a weak, non-significant positive association (r = 0.26 , p = 0.132), indicating that a monotonic co-variation between the two tracer groups was not detected under the present sampling conditions. PAH concentrations were dominated by high-molecular-weight, traffic and heating-associated compounds in winter (BghiP 5.8 ng m -3 , BaP 5.6 ng m -3 ), consistent with pyrogenic combustion amplified by residential heating demand. Complementary MA and MPh diagnostic ratios (L/M = 0.57-6.35) identified a biomass-burning signature consistent with predominantly softwood-related combustion. Biomass-burning markers showed a directional 1.5-3.2-fold enrichment in autumn-winter relative to spring-summer, though this difference did not reach statistical significance for any individual marker (Kruskal-Wallis, p > 0.12), consistent with a moderate rather than dominant local biomass-burning source. Compared with previous studies from French and Mediterranean sites, the molecular marker patterns suggest relatively less evidence of hardwood or agricultural-residue burning. This study also represents the first simultaneous four-season MA-MPh profiling investigation conducted in northeastern France. Together, these results suggest that exhaust combustion, non-exhaust emission, and biomass burning represent chemically distinguishable contributors to urban PM 10 , supporting consideration of source-specific rather than aggregated approaches to traffic-related air quality management.

Remote sensing change detection technology is widely used in land-use monitoring, urban planning, and disaster assessment. However, during imaging and transmission, bi-temporal remote sensing images are vulnerable to Gaussian noise, which makes it difficult for change detection algorithms to distinguish truly changed areas from noise-affected regions. To address this issue, this study proposes an unsupervised Gaussian-noise-robust change detection algorithm, termed FRIH-SEEDSAM. The proposed method first applies the Fast and Robust Fuzzy C-Means (FRFCM) algorithm to perform noise-resistant fuzzy clustering on bi-temporal remote sensing images. To establish reliable correspondences between the clustering results, the Integrated Region Matching (IRM) algorithm is introduced to construct weighted matching relationships while reducing the influence of abnormal memberships. The change intensity of spatially corresponding pixels is then calculated to generate a more stable change intensity map. Subsequently, the change intensity map is input into the Hybrid Conditional Random Field (HCRF) to infer pixel-level change labels, where the object potential function is constructed from the segmentation results of the Energy-Driven Sampling (SEEDS)-guided Segment Anything Model (SEEDSAM), which uses the centroids of the SEEDS superpixel regions as point prompts for SAM, thereby enhancing change-label consistency within the same changed object region. The experimental results show that the FRIH-SEEDSAM algorithm maintains stable change detection performance across different datasets and under varying Gaussian noise levels. It outperforms the comparison algorithms in terms of several accuracy evaluation indicators, including Kappa and F1. Furthermore, even when the Gaussian noise variance increases to 0.05, Kappa remains at 0.8 or above on multiple dataset images.

The shallow shear-wave velocity structure obtained from surface-wave dispersion inversion is essential for seismic site characterization; however, deterministic inversion provides only point estimates without uncertainty, while existing probabilistic approaches often lack coverage calibration, systematic robustness evaluation, and physical-consistency checks. A mixture density network is trained on a large-scale public dispersion-inversion benchmark to output a depth-wise probability distribution of shear-wave velocity V s ; conformal calibration is introduced to provide coverage guarantees, an augmentation-calibrated robust variant handles distribution shift, and dispersion forward modeling examines the physical consistency of predictions. On the test set, the shear-wave velocity prediction attains a coefficient of determination R 2 of 0.904 and a root-mean-square error of 0.199 km/s; this calibration corrects the empirical coverage of the nominal 90% interval from an over-covered 0.950 to a precise 0.903 while narrowing the interval width by 14.3%. Under observational noise and out-of-distribution geology, the standard conformal coverage degrades, decreasing to 0.829 for faulted sites, and augmented calibration partially restores it to 0.868; a contrast between generic and frequency-dependent physics-guided perturbations yields nearly identical coverage, indicating that the recovery stems mainly from augmentation breadth rather than perturbation spectral shape, with residual under-coverage remaining under severe shift. A dispersion-consistency diagnostic reveals a positive correlation (0.558) between physical residual and predictive uncertainty, showing that the uncertainty captures physical inconsistency. The framework delivers calibrated, robust, and physically consistent uncertainty quantification for probabilistic surface-wave dispersion inversion.