Atmospheric and Oceanic Sciences

Hyperspectral infrared observations are an important data source for numerical weather prediction (NWP) because they provide rich information on the vertical structure of atmospheric temperature and humidity. However, most existing deep learning methods mainly focus on one-way retrieval from radiances to atmospheric profiles, while the reverse radiance simulation process and the consistency between atmospheric state space and radiance observation space are insufficiently considered. In this study, we propose SIMBA, a unified bidirectional retrieval–forward simulation framework for FY-4A/GIIRS hyperspectral infrared radiance modeling toward NWP applications. The framework jointly performs atmospheric profile retrieval and radiance reconstruction, introduces a cycle-consistency constraint to strengthen the coupling between the two processes, and employs a bidirectional Mamba state-space module to capture long-range dependencies along pressure levels. Using collocated FY-4A/GIIRS observations and ERA5 reanalysis data, the proposed method is evaluated for temperature retrieval, specific humidity retrieval, long-wave radiance reconstruction, and medium wave radiance reconstruction. Experimental results show that SIMBA outperforms several representative deep learning baselines across both retrieval and reconstruction tasks, while ablation experiments confirm the contribution of the bidirectional design and cycle-consistency mechanism. These results demonstrate that the proposed framework is effective for joint atmospheric profile retrieval and hyperspectral infrared radiance modeling, and suggests potential for future Jacobian-related analysis and NWP-oriented extensions.

Abstract We reconstruct the stable carbon isotopic composition of Neogene surface ocean dissolved inorganic carbon (δ 13 C DIC ) and atmospheric CO 2 (δ 13 C ATM ) to serve as input for carbon cycle modeling and proxy‐related studies. The largely observation‐based δ 13 C DIC reconstructions are inferred from Neogene planktic foraminifer carbonate δ 13 C records with offsets based on Holocene records and modern surface water δ 13 C DIC data. Estimates are corrected for foraminifer ontogenetic stage (size), modern anthropogenic surface water 13 C‐depletion since preindustrial times (Suess effect) and carbonate ion concentration changes. Sea surface temperature‐corrected air‐sea fractionations are applied to Neogene δ 13 C DIC records to reconstruct a δ 13 C ATM curve for the last 20 Myr. Neogene δ 13 C DIC and δ 13 C ATM values are somewhat higher than previously estimated. Parallel trends in the surface and deep ocean point to global exogenic δ 13 C trends and variations in the organic carbon cycle. Three features stand out: (a) a distinct δ 13 C peak at 16–15 Ma representing the climax of the Monterey Carbon Isotope Excursion (MCIE), (b) a Middle Miocene to Pleistocene decreasing δ 13 C trend, and (c) a well‐defined second Late Miocene δ 13 C maximum at 9–7.5 Ma superposed on the main trend (“Bengal Carbon Isotopic excursion,” BCIE) that precedes the well‐known Late Miocene Carbon Isotopic Shift (LMCIS). Explanations for the MCIE and subsequent δ 13 C decrease should consider variations in carbon storage on both continental shelves and land, p CO 2 ‐dependent carbon fractionation during photosynthesis, and aridity/vegetation affecting both the C and Si cycle. The BCIE reversal is most likely caused by intense weathering in the Himalayan‐Tibetan region affecting marine productivity and organic carbon burial.

Forest-steppe ecotones exhibit pronounced spatiotemporal heterogeneity and complex climate–vegetation interactions, posing significant challenges for vegetation dynamics prediction. Existing models often struggle to capture long-range temporal dependencies, preserve spatial continuity across heterogeneous transition zones, and provide ecologically interpretable insights. To address these limitations, we developed a bidirectional Geo-Spatial Mamba (Geo-S-Mamba) architecture with a multi-objective loss function incorporating spatial continuity constraints based on the first law of geography. The model was trained using multi-source geospatial datasets and independently validated during 2019–2023. The results show that Geo-S-Mamba achieved an R2 of 0.93. Moreover, both the bidirectional mechanism and the spatial-continuity loss improved the PSDI by approximately 0.08. The model effectively captured annual variations in NDVI and covariation among vegetation groups. Post hoc symmetric causal learning based on Pearl’s structural causal theory indicated that precipitation was the primary driver of grassland vegetation dynamics. Temperature and radiation influenced NDVI mainly through boundary-dependent effects. Overall, this framework can estimate changes in the spatial distribution of plant communities across heterogeneous environments and provides a scientific basis for further research on forest–steppe ecotones.

Abstract This study presents the first systematic catalog and long‐term statistical analysis of post‐sunset Equatorial Plasma Bubble (EPB) suppression events over northeastern Brazil, based on a 12‐year observational database spanning 2013–2024. The EPB activity was identified using GNSS‐derived Rate of Total Electron Content Index measurements over the São Luís longitudinal sector (40°–50°W). The results demonstrate that EPB suppression occurs under a broad spectrum of geomagnetic conditions, including intense, moderate, and weak geomagnetic storms, as well as during geomagnetically quiet periods (SYM‐H > −30 nT and Kp 3). A clear semiannual variation is observed, with maxima around the equinoxes, coinciding with periods of enhanced geomagnetic storm occurrence. Most suppression events occurred during the descending and ascending phases of the solar cycle. In some cases, suppression effects were observed more than 300 hr after storm onset, suggesting the influence of disturbance dynamo electric fields, modifications in the integrated E‐region Pedersen conductivity, and/or overshielding Prompt Penetration Electric Fields during the evening sector.

The recently developed Complex Amplitude Sensing (CAS) technique—in its two versions, CAS-v1 for liquid-borne particles and CAS-v2 for gas-borne particles—enables robust measurements of the complex forward-scattering amplitude S ( 0 ∘ ) of single particles, from which detailed analytical information about the particles—refractive index, size, and (via the polarization resolution of CAS-v2) shape—can be obtained. Here we develop a framework for quantifying the instrument-dependent errors in the S ( 0 ∘ ) measurement and demonstrate it on gas-borne particle measurement by CAS-v2. The framework infers, for each instrument, a four-dimensional error parameter vector comprising a relative systematic error and a random-noise standard deviation in each of Re S ( 0 ∘ ) and Im S ( 0 ∘ ) , by applying a Bayesian analysis to the S ( 0 ∘ ) data points of standard spherical particles with known refractive index and size. Because the three instruments probe the same standard-particle population, the data from λ = 453 n m , 638 nm, and 834 nm are combined in a joint inference that shares the wavelength-independent mean diameter across instruments; this coupling restores uniqueness in the large-diameter, short-wavelength regime where single-wavelength inference is defeated by Mie-amplitude oscillation. We apply the framework to three CAS-v2 instruments using aerosolized polystyrene latex and silica standard particles, obtaining posterior estimates of the relative systematic error and noise parameters separately for each instrument. The inference framework is general and applies to any S ( 0 ∘ ) -measuring instrument for gas-borne and liquid-borne particles. The resulting per-instrument error anchors supply realistic, data-driven priors on the S ( 0 ∘ ) measurement errors and thereby provide the foundation for uncertainty-aware retrievals of the refractive index, size, and shape of unknown particles.

Abstract. Satellite and reanalysis data sets are analyzed to explain sea level changes in the tropical North Atlantic margin off northwest Africa. The study domain sea level was rising as far back as 1986 and a pause in sea level rise (hiatus) began around 2010 and stopped in 2019. Characteristics of sea level anomaly and its drivers during a period of rise (1996–2004) and the hiatus period (2010–2018) are analyzed and compared. Results show that the most effective cause of domain-wide sea level rise during the period of rise is seawater expansion owing to changes in density structure (steric expansion), with almost equal contribution from temperature-driven (thermosteric) expansion and salinity-driven (halosteric) expansion. The cause of the domain-wide pause in sea level rise is a large thermosteric contraction that counteracted halosteric expansion and mass accumulation. Multidecadal sea level increase, defined here as the difference between the mean sea level during the period of rise and the hiatus period, is owing to steric expansion, vertical land motion, and mass accumulation, which contributed 56 %, 24 %, and 16 %, respectively. There are, however, regional differences in the patterns of multidecadal steric and mass adjustment. In the northern subdomain where ocean processes predominate mass-driven sea level variability, the steric adjustment is dominated by halosteric expansion, whereas in the southern subdomain where atmosphere-ocean processes predominate mass-driven sea level variability, the steric adjustment is dominated by thermosteric expansion. The accumulation of low-salinity water in the northern subdomain and precipitation in the southern subdomain appears to be associated with a mutual adjustment of vertical and horizontal velocity distribution inside the domain and west of it in the area of the Guinea Dome, a permanent upwelling region where isotherms are displaced upwards. The low-salinity water influx to the northern subdomain is linked to changes in the southward-flowing Canary Current. A probable hypothesis inferred from correlation and potential vorticity analysis is that the Canary Current source region was freshened by currents that supply water to the region via two pathways: an open ocean path that is consistent with the Azores Current, and a Western Europe coastal ocean path that is consistent with the Portugal Current and Portugal Coastal Current system. The results obtained highlight a multidecadal linkage between sea level anomalies in the eastern tropical North Atlantic margin and salinity anomalies elsewhere in the North Atlantic.

Abstract Pulsating aurora (PsA) is one of the common diffuse auroras typically generated by ∼10 keV electron precipitation. Precipitating electrons are scattered via wave‐particle interactions with whistler‐mode chorus waves in the magnetosphere. Previous studies have observationally demonstrated that, during PsAs, highly energetic electrons with energies ranging from a few hundred keV to a few MeV simultaneously penetrate into lower altitudes and can cause ozone depletion in the mesosphere and upper stratosphere. Recent attention has focused on the importance of high‐latitude propagation of chorus waves through magnetospheric density ducts for the precipitation of such highly energetic electrons. Our previous study showed that patchy structures of PsAs reflect the cross section of ducts. However, no statistical study has examined the relationship between ducted propagation of chorus waves and PsAs using ground‐satellite conjugate observations. Thus, we perform a statistical analysis to evaluate this relationship using data from the ground‐based optical network in Scandinavia and the Arase satellite. The results show that when chorus waves propagate to higher latitudes while maintaining their electromagnetic properties, PsAs exhibit patchy structures with more than 90% probability, and occurrence increases from midnight to the morning sector. In addition, the background electron density irregularities indicative of ducts were also observed in four events during high‐latitude propagation of electromagnetic chorus waves. For these events, the equatorial projection sizes of PsA patches ranged from several hundred to several thousand kilometers, comparable to the scale required to confine chorus waves. These observations suggest the horizontal scale of ducts at the magnetic equator.

Land surface temperature (LST) at high spatial resolution is needed for urban heat island analysis, but Landsat thermal sensors measure at approximately 100 m before resampling to 30 m, limiting effective thermal detail. This study presents a framework for downscaling Landsat LST from 30 m to 10 m using machine learning in Google Earth Engine (GEE), applied to Warsaw, Poland, across five summers (2021-2025). Eight predictors derived from Sentinel-2 (Normalized Difference Vegetation Index [NDVI], Normalized Difference Built-up Index [NDBI], Bare Soil Index [BSI], Modified Normalized Difference Water Index [MNDWI], and albedo) and Sentinel-1 Synthetic Aperture Radar (SAR) data (co-polarized VV backscatter, cross-polarized VH backscatter, and the VV/VH ratio) were used. Four GEE-native algorithms were compared: Random Forest (RF), Gradient Boosting Trees (GBT), Support Vector Machine (SVM), and Classification and Regression Trees (CART). RF achieved the best performance (mean root mean square error [RMSE] = 1.135°C, mean absolute error [MAE] = 0.827°C, coefficient of determination [R 2 ] = 0.896). NDBI was the most important predictor (23.1%), while SAR features contributed 27.6% of total importance. However, an ablation experiment showed that the full model (with SAR) outperformed the optical-only model in three out of five years (2021-2023), while the optical-only model performed better in 2024-2025. The mean RMSE difference was 0.110°C in favor of the full model, indicating that the SAR benefit is condition-dependent. Three strategies for computing shortwave infrared (SWIR)-based indices were compared: computing at native 20 m then resampling to 10 m reduced spatial artifacts at land cover boundaries while achieving the best average accuracy. Indirect validation using temperature transects confirmed that the downscaled LST captured sharper thermal gradients at urban-park and land-water boundaries than the 30 m input. The entire workflow is provided as open-source GEE code, requiring no local computing resources.

Snow depth, as a key parameter characterizing snow cover properties, plays a crucial role in climate change research, water resource regulation, and ecosystem stability. With the rapid development of Global Navigation Satellite System Reflectometry (GNSS-R) technology, its potential in snow depth retrieval has become increasingly evident. However, existing snow depth retrieval from GNSS-R has mainly focused on ground-based GNSS-R measurements, which is inherently limited to point measurements and cannot support large–scale spatial monitoring such as the Tibetan Plateau. In this study, SNR is used for snow depth retrieval from Cyclone-GNSS (CYGNSS) GNSS-R signal data over the Tibetan Plateau region. The CatBoost machine learning model was developed to estimate snow depth at a spatial resolution of 0.25° × 0.25° and a temporal resolution of 1 day, and its performance was subsequently evaluated. The results demonstrate that the CatBoost model performs the best performance than Random Forest (RF) and Artificial Neural Network (ANN) with a mean absolute error (MAE) of 1.065 cm and a correlation coefficient (R) of 0.886 on the test dataset. This model can accurately capture the spatiotemporal distribution characteristics of snow depth across the Tibetan Plateau. Further analysis , including ablation experiments and validations across different snow depths, seasons, DEM conditions, and independent stations, confirms the model’s overall robustness. The model captures seasonal snow depth variations and reproduces snow accumulation and melt processes well. Overall, the model retrieves snow depth well and reproduces seasonal snow depth variations effectively with the best performance during periods of abundant snow accumulation and stability.

ABSTRACT This study investigates the meteorological characteristics and associated human thermal discomfort of downslope windstorms known as ‘Vento Norte’ (VNOR; ‘North Wind’ in Portuguese) occurring in the city of Santa Maria, southern Brazil, over 20 years (2004–2023). VNOR events are characterized by unseasonable strong northerly winds, abrupt temperature rises and significant drops in relative humidity. Utilising hourly meteorological data, we identified 189 VNOR episodes, predominantly occurring during austral winter months (June–August), reflecting a seasonal pattern. VNOR events significantly altered local thermal comfort conditions, as demonstrated through a comparative analysis of five biometeorological indices: Effective Temperature with Wind (TEFW), Thermal Discomfort Index (TDI), Temperature–Humidity Index (THI), Human Discomfort Index (HDI) and Humidex (HU). Statistical analyses revealed significant differences between VNOR and non‐VNOR periods, with VNOR episodes consistently associated with elevated human thermal discomfort. A detailed examination of representative events further illustrated abrupt changes in meteorological parameters, highlighting the potential for rapid thermal stress in the local population. The present analysis demonstrates the importance of monitoring VNOR phenomena and their thermal impacts and suggests how these downslope windstorms alter local atmospheric conditions, providing an environmental baseline that could support future risk‐awareness frameworks and biometeorological monitoring protocols in southern Brazil.

Abstract Simulated supercell thunderstorms were initiated using four base-state thermodynamic profiles to test the influence of variation in both planetary boundary layer and free tropospheric relative humidity on downdrafts and cold pool characteristics. The selected base state environments include different combinations of boundary layer and free tropospheric relative humidity and closely resemble those of four observed supercells. Bulk cold pool properties were more dependent on boundary layer relative humidity, with more negatively buoyant outflow when the base-state boundary layer was dry. Free tropospheric relative humidity had a secondary effect on cold pool intensity by modifying precipitation production. The most intense cold pool was the result of the combination of a dry boundary layer that promoted greater rates of evaporation and a moist free troposphere that produced large amounts of precipitation. The relationship between different relative humidity regimes and cold pool characteristics was mostly insensitive to changes in CAPE, shear profile, microphysics, surface drag, random temperature perturbations, and temperature profiles.

Abstract A wide range of functions are currently available for simulating the calving of marine‐terminating glaciers, but there is no consensus on the best approach to represent the calving process in glacier and ice‐sheet models. Current assessments of calving functions are often crudely done by fitting functions to observed changes in terminus positions, neglecting the physical processes that drive changes in calving dynamics. Here, we use 3D simulations of synthetic tidewater glacier domains in Elmer/Ice, to determine whether natural behaviors emerge from the crevasse‐depth and von Mises calving functions, and to provide a basis for more robust assessments of the potential capabilities of calving functions. The crevasse‐depth calving function is shown to be able to simulate both serac and full‐thickness calving events and simulates how their relative proportion is altered by changing the ice freeboard or submarine melting. A clear distinction between rate‐ and position‐based calving is shown, with the von Mises calving function unable to respond to imposed changes in topography or freeboard ice. By comparing the two calving functions, it is apparent that the position‐based crevasse‐depth function more faithfully represents the calving behaviors observed in the natural world. Consequently, future projections should be made using position‐based calving functions. Using a position function, calving rates vary with time and glacier state, so cannot be assumed to be a constant function of stress. In essence, a calving function must be able to capture the key physical processes that drive calving. If so, the transitions in calving dynamics will inherently emerge.

Abstract A novel classification of three‐dimensional heat wave types is applied to Middle Europe over 1979–2022. Heat waves are classified according to their vertical structure of temperature anomalies in the ERA5 reanalysis into near‐surface (HWG), lower‐tropospheric (HWL), higher‐tropospheric (HWH), and omnipresent (HWO). Jenkinson–Collison classification is used to link circulation types (CTs) to individual heat wave types, and their climatological characteristics are studied in station‐based E‐OBS data. We show that the surface temperature anomalies are largest during HWG, which also are characterized by the driest conditions before onset. In all heat wave types, CTs with southerly flow are more common compared to the June–September climatology, but differences among the heat wave types are found for other CTs. In HWG, the CT occurring most frequently is indeterminate flow, corresponding to a little pronounced pressure field with no clear role of anticyclonic vorticity or flow direction. The expected pattern of increased anticyclonic and decreased cyclonic flow is clearly manifest only for HWH, while it is reversed during HWG. An important role of anticyclonic circulation supporting gradual warming is found before the onset of most heat wave types, except for HWH. The differences in circulation characteristics among the heat wave types are smaller after their termination, which is often associated with relatively cold westerly flow. The analysis contributes to better understanding the interrelationships among different heat wave types, atmospheric circulation, and other driving mechanisms.

Every year, an additional 11 million tons of microplastics permeate the water column and marine ecosystems within it. Geoscience is contributing to strategies that aim to identify microplastic movements—and stop them.

Excessive infiltration and inflow (I&I) of environmental water into sewer networks renders conventional wastewater treatment ineffective and costly. Physical inspection techniques (e.g., Closed Circuit Television and Sonar) are the go-to methods for diagnosis but are expensive ($5–15 USD/m). Chemical fingerprinting is over an order of magnitude cheaper ($0.5–1.5 USD/m) but the diagnosis is delayed due to reliance on laboratory analysis and is error-prone using deterministic algorithms. To overcome such limitations, sewer samples ( n = 21) were collected from a small (0.5 km 2 ) coastal district of Shenzhen and analyzed initially for 9 tracers including pharmaceutical compounds unique to sewage, but later focusing on 3 tracers (electrical conductivity (EC), radon-222 ( 222 Rn), and ammonia nitrogen (NH 3 −N)). New data from 31 environmental water samples were critical for constraining endmember compositions. A Bayesian Chemical Mass Balance (CMB) model distinguished and quantified seawater, groundwater, and rainwater intrusion probabilistically. The model using 3 tracers under dry-weather conditions (or 4 when δ 18 O was included for rainwater inflow during wet-weather events) yielded sewage proportion estimates comparable to that of 8 or 9 tracers, differing only by 3.0% ± 2.5%, although the mean relative error was twice as high. At four sewage wells, raw sewage accounted for 32.3 ± 13.3%, 12.8 ± 5.3%, 41.9 ± 15.1% and 33.6 ± 13.4% along the flow paths in August 2021. Significant temporal variability was evident at one well, with raw sewage contributions ranging from 21.2 ± 6.7% in August 2021 to 58.8 ± 8.0% in November 2022 and 50.8 ± 37.5% in January 2025.

Study region The Zambezi River Basin (ZRB) is a large transboundary basin in south-central Africa shared by eight countries. It provides vital water resources for hydropower, agriculture, ecosystems, and human use across the region. Study focus The remote sensing-based Water Accounting Plus (WA+) framework was applied to the ZRB from 2003 to 2023 to quantify its water balance. The WA + methodology was used to overcome sparse ground data by leveraging satellite-derived datasets such as precipitation ( P ), actual evapotranspiration ( ET a ) and land cover and use (LULC). New hydrological insights for the region The WA + water balance for the ZRB reveals that P (1280.5 km³/year) is the dominant inflow, while ET a (1095.6 km³/year) constitutes 85.5% of the gross inflow. This leaves a long-term mean surface water yield ( P–ET a ) of 184.9 km³ /year. A substantial portion of the inflow (115.9 km³/year) is lost to storage, suggesting significant basin-scale water-storage potential. The exploitable water averages 106.6 km³ /year, of which 19.9 km³ /year is reserved outflow and 0.2 km³ /year is non-utilisable outflow; thus, about 86.4 km³ /year remains available for use. Currently, 44.5 km³ /year is utilised flow, leaving 42.0 km³ /year that is still undeveloped. Overall, the basin sustains a net-positive water balance, though water availability declined over the study period, coinciding with a general downward tendency in P and elevated ET a in recent years; however, the basin-scale trends in P and ET a were not statistically significant.

Study area North-eastern Ethiopia, Lower Awash Basin Study focus The study aims to investigate the processes influencing groundwater composition and evolution using ionic ratios, chemometrics, and inverse geochemical modeling based on major ion chemistry of 164 groundwater samples that were collected from hand-dug wells, springs, and boreholes New hydrogeological insights The Results from the Gibbs plot, Ca + Mg vs SO 4 +HCO 3, Ca/Na vs HCO 3 /Na vs Ca/Na vs Mg/Na, and Chloralkaline indices showed that rock water interaction, silicate weathering, cation exchange, and evaporation are the major geochemical processes that control the groundwater chemistry. Principal component analysis identified three components: PC-I and PC-II reflect geogenic processes, whereas PC-III indicates anthropogenic influences. Hierarchical cluster analysis grouped the groundwater samples in to two main clusters: cluster I represents fresh Ca-Mg-HCO₃ type from the recharge zone alongside a mixed water type from transition zone, while cluster II includes mineral-rich samples from the discharge zone (Na-HCO 3 and Na-Cl). Inverse geochemical modeling indicated that major geochemical process that affect groundwater chemistry and evolution are dissolution of primarily silicate minerals such as plagioclase, olivine, and pyroxene, along with the precipitation of calcite and clay minerals. This study addresses the gap in conventional hydrochemical approaches in the understanding of hydrogeochemical processes by employing a quantitative approach and contributes in the formulation of policies for sustainable groundwater resource management.

Total organic carbon (TOC) is a critical parameter for evaluating hydrocarbon source rocks, but its strong spatial heterogeneity challenges accurate regional prediction using limited well sample data. Therefore, this study presents an integrated geophysical workflow for the three-dimensional (3D) quantitative prediction of TOC in the post-salt marine source rocks of the Madingo Formation, Lower Congo Basin. Calibrated with 60 measured samples, three log-based TOC models (multiple regression, improved ΔLogR, and a Back-Propagation neural network [BPNN]) were evaluated, and the optimal log-derived TOC was integrated with seismic attributes to construct a 3D TOC volume via seismic impedance inversion. Comparative results demonstrate that the BPNN outperforms traditional empirical models, yielding a correlation coefficient of R = 0.9342. The 3D prediction reveals pronounced spatiotemporal heterogeneity in TOC distribution, which is strongly controlled by sedimentary facies, reaching maximum concentrations in the deep-water slope zone and decreasing towards shallow-water zones. This integrated data-driven approach effectively addresses the limitations of discrete sample analysis, providing a quantified predictive framework that helps constrain prediction uncertainty for deep-sea hydrocarbon exploration in heterogeneous marine source rocks.

Abstract In the induced magnetotail of Mars, the neutral sheet is an important channel for atmospheric ion escape, yet the role of heavy planetary pickup ions in shaping its geometry has been unclear. Using three‐dimensional hybrid simulations, we demonstrate that these ions induce a global neutral sheet asymmetry via mass loading. Ambient plasma flow deceleration bends draped interplanetary magnetic field (IMF) lines, extending the sheet along the direction at large IMF Parker spiral angles and tilting its ‐hemisphere segment toward the plane in the Mars–Solar–Electric coordinate system at small angles. The asymmetry's magnitude is modulated by both the planetary ion production rate and solar wind dynamic pressure. These results show that Mars' magnetotail is not solely governed by IMF draping and crustal fields, but also dynamically reshaped by pickup ions.

For effective operation of Floating Offshore Wind Turbine (FOWT), comprehensive investigations are required, not only focusing on the turbine aerodynamics, but also on the global motion behaviours including the motion stability and risk assessment under failure scenarios. This paper presents numerical simulations of the IEA 15 MW offshore wind turbine under mooring failure scenarios in extreme environments. For the numerical analysis, a coupled framework was employed, combining the Computational Fluid Dynamics (CFD) software, STAR-CCM+, and the lumped-mass mooring dynamics code, MoorDyn. The turbulent flow fields around the turbine and floating platform were resolved using the CFD solver, while the dynamic response of the mooring system was analysed using MoorDyn. In the scenario where one of the 3-catenary mooring lines is broken, the variations in the 6-DOF motion response of FOWT were investigated with respect to the turbine blade pitch angles (i.e., normal and feathered blades). Additionally, the dynamic tensions in the remaining two lines were observed. The results indicate that feathered blade pitch angle can effectively reduce the motion response of FOWT. For instance, the platform pitch motion was reduced by more than half compared to the normal blade configuration, highlighting the importance of blade feathering for maintaining stability. In contrast, the normal blade configuration may lead to significantly larger platform responses under failure scenarios.