Atmospheric and Oceanic Sciences

Abstract Atmospheric rivers (ARs) over western North America drive extreme precipitation and flood hazard, yet their intensity upper bounds remain poorly constrained by the short observational record. We combine a differentiable global climate model (~1.4°) with high-resolution dynamical downscaling (~0.11°) to construct physically plausible storylines of five unprecedented AR events in British Columbia. By optimizing minimal perturbations to historical initial conditions, we generate events that maximize integrated vapor transport (IVT) and exceed the observational record under present-day conditions. Pseudo-global warming perturbations under SSP5–8.5 provide a second pathway through end-of-century warming. Both approaches amplify AR intensity through distinct mechanisms: the optimization primarily modifies the wind field, while the pseudo-global warming signal primarily increases atmospheric moisture. These contributions act on largely independent components, and when applied simultaneously, the combined effect is nearly additive, with differences generally below 15%. Non-linear effects, though small, are always positive, and within these storylines the most extreme physically plausible ARs arise from compounding both drivers. The pathways differ in precipitation efficiency: the dynamical amplification largely preserves the conversion of moisture transport to precipitation, whereas the thermodynamic amplification reduces it by up to 29%. Bounding the upper tail of AR hazard may therefore require accounting for both dynamical variability and thermodynamic change.

Tropical cyclones (TCs) are one of the most destructive natural disasters. It is important to document long-term changes in TC properties to better prepare for the damages associated with TCs. TC lifespan is one of the little-studied TC characteristics. Here we show that annual mean TC duration has been decreasing at rates of -13 ± 6 and -7 ± 6 hours per decade, corresponding to approximately 56 and 30 h shorter average lifespans, in the Northeast and Northwest Pacific, respectively, from 1982 to 2024. The decreasing trend is primarily due to shorter TC tracks associated with poleward migration of TC genesis latitude. The TC intensification (weakening) rate before (after) the first (last) lifetime maximum intensity has increased, a phenomenon in which environmental conditions and internal convective structure may both play a role. The shorter TC lifespan over the open ocean before entering coastal zones compresses the time available for weather forecasting and disaster preparedness.

Abstract We investigate the North Atlantic (NA) tropical cyclone (TC) response to a substantial weakened Atlantic Meridional Overturning Circulation (AMOC) under external freshwater forcing using a high‐resolution coupled model. With explicit TC trackings instead of indirect inferences from large‐scale environmental changes, we detect a non‐uniform AMOC‐linked NA TC reduction. The transient TC reduction is more rapid and pronounced over the eastern NA. After several decades, the TCs are almost absent over the eastern NA and the remaining substantially reduced TCs are mostly confined over the western subtropical NA. This non‐uniform TC reduction is related to the equatorward propagation of negative extratropical surface temperature and near‐surface humidity anomalies along a horseshoe pathway as the AMOC weakens. Concurrently, vertical wind shear strengthens over the tropical NA, reinforcing the TC suppression there. In contrast, the AMOC weakening induces TC‐favorable environmental changes (e.g., surface warming) and thus the emergence of TCs over the South Atlantic.

Abstract The deployment of a multistatic radar network in the Oklahoma City metropolitan area in Spring 2024 has allowed for unique observations of several severe weather events. This passive multistatic network operates in conjunction with the operational KTLX WSR‐88D, allowing for synchronous multi‐Doppler analyses. This work presents a case study of a tornadic mesovortex in central Oklahoma. Prior to tornadogenesis, a horizontal rotor circulation is tilted into the mesovortex, providing a pathway for an intense low‐level updraft to develop. The release of horizontal shearing instability beneath this updraft is shown to supply near‐surface vertical vorticity, which is then amplified by vertical stretching, which is followed by tornadogenesis. The ability to derive kinematic fields from a single operational radar represents a major expansion of the current observational capabilities of the WSR‐88D network.

Abstract Tropical convection over rainforests modulates atmospheric circulation and the energy and hydrological cycles across multiple scales. However, the scarcity of observations still limits our understanding of these processes. Although numerical models are utilized to investigate atmospheric physical processes, their performance depends on the choice of parameterizations and grid resolution. Here, we introduce and apply a multi‐variable, multi‐physics framework to evaluate and rank the Advanced Research Weather Research and Forecasting (WRF–ARW) configurations at 1‐km resolution over the central Amazon during the wet season. A 48‐member ensemble combines three land‐surface models (LSMs), four planetary boundary layer (PBL), and four microphysics (MP) schemes. Model performance for seven convection‐related near‐surface and boundary‐layer variables is assessed using Taylor diagrams and the Taylor skill score (TSS), analysis of variance (ANOVA)‐based sensitivity metrics, and non‐parametric rank tests. We then construct a combined, weighted TSS to identify configurations that are comparatively robust across variables. Results showed that most configurations reproduce near‐surface temperature, sensible and latent heat fluxes, and boundary‐layer height reasonably well, whereas humidity and rainfall remain challenging. LSM choice has the strongest impact on the surface fluxes and a secondary influence on near‐surface temperature and humidity, PBL schemes dominate boundary‐layer height, and MP schemes exert the largest control on rainfall. No single configuration is optimal for all variables, but the combination of Noah (LSM), Yonsei University (PBL), and Morrison (MP) emerges as the most robust configuration for this case, with the WRF single‐moment six‐class scheme (WSM6) providing a competitive, computationally cheaper MP alternative. The framework is general and can be applied to other regions, seasons, and convective regimes.

Quantitatively predicting new particle formation (NPF) remains challenging, largely because the true origin of events observed on the surface is often ambiguous. This potential for misattributing particles formed aloft as ground-level events fundamentally compromises predictive models built on local measurements. Here, we propose and validate an analytical framework using airborne eddy covariance and continuous wavelet transform to directly quantify the vertical turbulent flux of newly formed particles and determine their dominant vertical transport direction. Analyzing data from a dedicated airborne campaign over the Southern Great Plains, we observed a consistent and strong downward turbulent flux of ultrafine aerosols during NPF events, with a mean value of −133.8 cm –3 m s –1 in the entrainment zone, whereas fluxes on non-NPF days were negligible. Our spectral analysis further confirms that these directional fluxes can be reliably captured by using standard 1 Hz aerosol instrumentation. These findings suggest that NPF events associated with entrainment from the overlying residual/stable layer during planetary boundary layer growth represent a significant and potentially underestimated source of boundary layer aerosols. The framework presented here provides a new methodology to correctly attribute NPF events to specific altitudes, thereby improving our process-level understanding of particle formation in the atmosphere.

Abstract Accurate retrieval of dry air mole fraction of CO 2 (XCO 2 ) is essential for tracking emissions and supporting mitigation. However, aerosols, especially their particle size distribution (PSD), introduce significant uncertainties via scattering and absorption, yet are often overlooked in current methods. Here we present a boosted aerosol‐size‐Integrated XCO 2 (BASIC) retrieval algorithm that flexibly accounts for aerosol‐induced lightpath modifications. Validation at five TCCON sites in East Asia shows that BASIC reduces RMSE by 30% and 13% compared to standard and bias‐corrected OCO‐2 products, respectively. Moreover, BASIC more accurately reproduces the observed spectra, particularly in the aerosol‐sensitive O 2 A band, outperforming the ACOS algorithm. These improvements highlight the importance of incorporating variable aerosol PSD in retrievals and demonstrate that BASIC enables a more accurate representation of aerosol effects on radiative transfer. Our results suggest that PSD‐aware retrievals can significantly improve the accuracy of satellite‐derived XCO 2 estimates under heavy aerosol loading conditions.

Abstract The tropical cyclone (TC) season, defined by the timing of storm genesis within the year, is an important but still uncertain aspect of the TC response to climate forcing, especially across different ocean basins. Here we examine changes in Northern Hemisphere TC genesis timing using statistically downscaled tropical cyclone data sets derived from large‐scale environmental fields in three idealized carbon dioxide climate states. We focus on basin‐dependent responses in the western North Pacific (WNP) and North Atlantic (NA) and evaluate the contributions of vertical wind shear (VWS), potential intensity, and the thermodynamic parameter to seasonal asymmetry. Under moderate warming, the NA season lengthens, whereas the WNP shows a modest overall contraction that reflects opposing seasonal changes in its northern and southern subregions. Seasonal‐asymmetry diagnostics show that the controls on TC seasonality are basin‐dependent under moderate warming, with VWS providing the leading contribution in most basins. In the warmest climate state, treated here as an idealized end‐member sensitivity experiment, these regional contrasts become more pronounced, with the NA becoming later and shorter and the WNP exhibiting a longer season with subregional differences. Overall, TC seasonal responses are basin‐dependent and nonmonotonic, and their dominant environmental controls differ across climate states.

Abstract The Ross Ice Shelf Polynya (RISP) is one of the primary production sites of the High Salinity Shelf Water (HSSW), the precursor of Antarctic Bottom Water (AABW) that drives the lower limb of global meridional overturning circulation. Utilizing a multi‐platform historical data set gathered over the western Ross Sea continental shelf, this study employs a box model to estimate heat and salt budgets in the western RISP and Ross Ice Shelf Cavity (RISC) system. Our analysis identifies three dominant controls on heat and salt budgets in the western RISC: cavity‐polynya exchange (EX), glacial basal melting (GM), and subglacial discharge (SD). The rate of SD is estimated at 87.0 Gt·yr −1 , higher compared with previous studies. This study highlights the role of subglacial discharge in modulating water mass evolution in Antarctic marginal seas, a component that requires incorporation in future climate models.

Abstract Mesoscale convective systems (MCSs) are key contributors to heavy rainfall in the East Asian summer monsoon, yet their statistics and internal structure remain difficult to simulate, even in convection‐permitting models (CPMs). This study evaluates the performance of the newly developed Unified Forecast System Double‐Moment microphysics scheme (UDM) in a convection‐permitting configuration of the Weather Research and Forecasting (WRF) model for June‐September 2020 over East Asia. We compare UDM and the WRF Double‐Moment 7‐class microphysics scheme (WDM7) against satellite‐based identification of meso‐ α ‐ and meso‐ β ‐ scale convective systems ( α MCSs and β MCSs), and perform component‐wise sensitivity experiments in which individual UDM components (reduced ice‐phase accretion rates, warm‐rain autoconversion, in‐cloud microphysics, and semi‐Lagrangian rain sedimentation) are replaced by their WDM7 components. UDM substantially mitigates the systematic overestimation of βMCSs, reducing it from approximately 66% in WDM7 to 6%, while slightly alleviating the underestimation of αMCSs (by ∼11%). Composite analyses show that UDM produces broader, colder cloud shields and richer upper‐level ice and snow, along with stronger mid‐tropospheric updrafts and higher equivalent potential temperature in convective cores. In‐cloud microphysics and ice‐phase accretion rates are the dominant contributors to these improvements, whereas warm‐rain autoconversion plays a secondary role. They shift hydrometeor partitioning from fast‐falling hail toward slower ice and snow, resulting in a positive microphysics‐dynamic feedback that sustains extended stratiform anvils. These results highlight the importance of physically consistent microphysics parameterizations for realistic simulation of MCS statistics in CPMs, and UDM provides a physically tested microphysics configuration that can be applied in future climate change experiments.

Abstract Employing orographic perturbation experiments from a state‐of‐the‐art climate model, this study investigates the topographic forcing associated with the Tibetan–Iranian Plateau (TIP) in modulating Northern Hemisphere boreal winter blocking frequency. The results show that TIP‐related topographic forcing profoundly impacts blocking frequency, with distinct roles of mechanical and thermal components. The increased hemispheric‐mean blocking under full TIP‐related topographic forcing is dominantly attributed to its mechanical forcing, while localized slight reductions stem from its thermal forcing. In general, the response of Northern Hemisphere winter blocking to TIP‐related topographic forcing is dominated by mechanical forcing. TIP‐related mechanical forcing favors blocking occurrence, particularly over Europe and the Ural region, via strengthened stationary waves and positive geopotential height tendencies induced by weakening synoptic‐scale transient eddies. In addition, decelerating background westerly wind and reducing the meridional gradient of potential vorticity also provide favorable conditions for the formation of blocking. In contrast, TIP‐related thermal forcing mildly suppresses North Pacific blocking, primarily through weakened stationary waves and altered background lower‐level temperatures.

Abstract Understanding spatiotemporal patterns in trace gases aids in separating sources and sinks and reducing model biases. As a proof‐of‐concept, we apply power spectral analysis to monthly total‐column CO from satellites (MOPITT and IASI) and models (CAMS reanalysis and CAM‐chem), along with tagged‐tracer simulations, to disentangle dominant patterns of CO. The dominant time scales (seasonal, interannual, and transport modes) derived from power distributions are used to attribute model and satellite biases to physical processes driving CO variability. We show that the interannual modes reflect interhemispheric exchange while seasonal modes reflect regional transport. A combined analysis of the dominant scales of CAM‐Chem tags showed that IASI and CAMS are able to detect more interannual variability due to biomass burning (BB) plumes in the extratropics compared to MOPITT and CAM‐Chem. IASI also shows stronger interannual variability that is associated with biogenic and non‐methane volatile organic compounds (NMVOCs) in the tropics. CO‐tagged sources indicate that anthropogenic CO is dominated by an interannual mode tied to seasonal shifts in the Intertropical Convergence Zone (ITCZ), whereas CO from fires shows pronounced zonal gradients in both hemispheres during boreal fall and winter, consistent with tropical BB plume transport to extratropics. Dominant time‐scale analysis of total‐column CO, together with tagged model tracers, allows an alternative diagnostic of models and satellite retrievals by decomposing source variability into periodic modes. By considering bias terms in frequency space, we visualize model and instrument bias as a convolution of the biases of each source term acting at their respective time scale.

Abstract Hydrologic modeling supports flood forecasting and water resources management, but complex preprocessing, parameterization, and configuration limit broader use. This study defines a six‐level framework for artificial intelligence (AI)‐agent autonomy in hydrologic modeling and develops a Level‐4 agent, powered by large language models, that translates natural‐language requests into data retrieval, model execution, diagnostics, and reports with human oversight. In a proof‐of‐concept application to the July 2025 Texas flash flood, the agent reproduced key flood dynamics in the tested basin and reduced manual workflow effort. Observation‐driven simulations aligned with streamflow records, whereas a forecast‐driven run missed the flood response because the deterministic rainfall forecast displaced the storm core. These results suggest the feasibility of AI‐agent‐assisted hydrologic modeling in the tested case, while robustness and generalization across broader basins and events remain to be established through systematic validation, probabilistic meteorological forcing, and expert review of automated outputs.

Abstract The unpredictability of precipitation over land is a major forecasting challenge, particularly when local processes dominate. This study introduces a classification scheme for precipitation events based on synoptic‐scale forcing for ascent, quantified via the quasigeostrophic (QG) omega equation, motivated by a broader goal of studying the role of land‐atmosphere coupling in regimes of weak synoptic forcing. The scheme classifies 3‐hourly precipitation as synoptically or non‐synoptically forced based on threshold values of QG omega. It is first developed using NASA's MERRA‐2 reanalysis, and QG omega fields computed from MERRA‐2 are subsequently used for classification of the IMERG precipitation data set over the continental United States (CONUS). The mean ratio of non‐synoptic to total precipitation events exhibits a strong south‐north gradient from 0.7 at 25°N to 0.35 at 50°N in the zonal average. There is strong seasonality, with synoptically driven precipitation dominant in winter and a non‐synoptic precipitation ratio exceeding 0.8 in the southern half of CONUS in summer. Interannual variability in the non‐synoptic precipitation frequency is largest in the southern CONUS, but secular trends are weak and largely insignificant except for a decreasing trend over the northern Plains and Midwest in both DJF and MAM. Classification of precipitation using QG omega computed directly from satellite sounder data is explored. The sounder‐based non‐synoptic precipitation ratio is significantly overestimated due to a clear‐sky coverage bias in the sounder retrievals.

Abstract Katabatic flows over sloping terrain exhibit low-level jets, with turbulent transport becoming dominant over local shear production near the jet maximum, challenging the applicability of classical similarity theory developed for horizontal terrain. This study investigates the influence of slope angle on local similarity scaling in katabatic flows using one-dimensional Reynolds-averaged Navier–Stokes (RANS) simulations with first- and second-order turbulence closures. The strengths and limitations of these models are assessed against observations from the Pasterze Glacier, the Vatnajökull ice cap, the MATERHORN experiment, and the Val Ferret field campaigns. The second-order closure reproduces the observed mean jet structure and momentum fluxes more accurately and is therefore used to analyse turbulence dynamics and similarity relations. A characteristic height scale, $$z_{TM}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>z</mml:mi> <mml:mrow> <mml:mi>TM</mml:mi> </mml:mrow> </mml:msub> </mml:math> , is proposed to identify the region where the recently-proposed slope-adjusted stability parameter of Hang et al. (2021) collapses the local dimensionless momentum gradients across observations and simulations. The observed dimensionless temperature gradient exhibits larger scatter, whereas the simulations show a more consistent trend. Overall results indicate that RANS provides an efficient and viable framework for studying katabatic flows, with the second-order closure resolving processes that the first-order closure does not represent but are essential to katabatic dynamics, and that flux–gradient relations based on the slope-adjusted stability parameter collapse the dimensionless momentum gradient well below $$z_{TM}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>z</mml:mi> <mml:mrow> <mml:mi>TM</mml:mi> </mml:mrow> </mml:msub> </mml:math> , while non-local transport processes limit flux–gradient relations near and above the jet peak.

Abstract Monitoring water vapor during extreme events is crucial for understanding atmospheric physics, as well as for improving the prediction of severe weather and enhancing early warning systems. The major gap in monitoring water vapor during extreme events lies in the limited spatial and temporal resolutions of existing techniques. This restricts our ability to capture rapid and localized atmospheric changes driving severe weather. The primary goal of this study is to evaluate the performance of unconstrained Global Navigation Satellite Systems (GNSS) tomography in resolving the three‐dimensional structure and evolution of atmospheric water vapor during severe weather conditions. In this study, we evaluated 10 heavy rainstorm events (each exceeding 200 mm), whose observations were processed tomographically over the Hong Kong region. Emphasis is given to the case of Typhoon Sarika and Super Typhoon Haima, 2 days apart, which is analyzed in greater detail to illustrate the technique's capability under intense weather conditions. All events were validated using radiosonde (RS) observations and ERA5 reanalysis data, allowing a comprehensive assessment of the accuracy and consistency of the tomographic retrievals. The results indicate that unconstrained GNSS tomography struggles to resolve near‐surface moisture but achieves performance comparable to ERA5.

Study region This study focuses on the Yunnan Plateau in southwestern China, examining nine major plateau lakes: Dianchi, Erhai, Fuxian Hu, Chenghai, Lugu Hu, Qilu Hu, Xingyun Hu, Yangzonghai, and Yilong Hu, all located in high-altitude settings with complex surrounding terrain. Study focus Using newly released Surface Water and Ocean Topography (SWOT) satellite data to monitor lake water-level dynamics, we propose an automated method to retrieve high-precision time-series water levels from SWOT observations. The method aims to improve spatiotemporal monitoring capability and provide a complementary alternative to the official SWOT LakeSP product. New hydrological insights for the region The results show that SWOT, combined with the proposed method, can monitor the nine lakes at a monthly scale with high accuracy (achieving an RMSE of less than 0.09 m). The retrieved water levels reveal clear seasonal variations, with minimum levels generally occurring in April–May and maxima from August to November, consistent with regional precipitation patterns. Intra-annual water-level fluctuations range from 0.23 to 1.20 m. Most lakes exhibit periodic annual variations, whereas Fuxian Hu and Yangzonghai show sustained water-level declines over the available observation period.

Our study focuses on an experimental demonstration of a novel method for CO 2 removal from ocean water, which has captured up to 40% of all anthropogenically released CO 2 emissions, that combines H 2 and redox salt looping to induce pH swings in electrochemical flow cells. Model-driven design optimization guides 3-D printed electrolyte flow channel designs to alleviate mass-transfer limitations. A ridged flow channel with 1 mm thick fins protruding at an angle of 30° from the base reduces ferri-/ferro-cyanide redox salt concentration boundary layer thickness by up to 35% while restricting the pressure drop to less than 415 Pa. A notable feature of this work is the experimental validation of the intrinsic operando cleaning capabilities enabled by the reversible looping process, eliminating what would otherwise be process down time. Over 4 acidification/basification cycles, 86% removal of fouling from electrode surfaces is demonstrated, while distinctly maintaining a constant electrochemical energy intensity. The lab-scale, proof-of-concept experimental demonstration shows promising results, which are interpreted to provide insights and guidelines toward further enhancing scalability and efficiency of oceanic carbon removal and utilization processes.

Abstract The deep Earth water cycle is a key process in Earth's evolution. Stishovite has been proposed as a major carrier of water from the surface into the lower mantle. However, its role remains unclear because its water solubility is controversial. We investigated the water solubility of Al‐free stishovite at pressures of 33–38 GPa and temperatures of 700–1200 K using in situ X‐ray diffraction in an advanced multianvil apparatus. Water solubility is high, around 3 wt%, at 700 K but decreases rapidly with increasing temperature and becomes negligible (&lt;1,000 ppm), at temperatures above 1000 K. These results suggest that pure stishovite is unlikely to transport significant amounts of water into Earth's lower mantle under comparable pressure‐temperature conditions. Previously reported high water contents in pure stishovite under deep Earth conditions likely reflect the solubility of post‐stishovite or experimental artifacts.

High Resolution Image Download MS PowerPoint Slide A potential approach for mitigating the climate impact of hard-to-abate, dilute (less than 1000 ppm; ppm) methane (CH 4 ) is gas-phase advanced oxidation (GPAO). Here, we present experimental results from a benchtop GPAO reactor oxidizing CH 4 with chlorine radicals (Cl · ) produced from the photolysis of chlorine gas (Cl 2 ), a process we term “Cl 2 -GPAO.” We find that at CH 4 concentrations from 2 to 90 ppm, supplying Cl 2 in a ratio of 1:1 with CH 4 uses photogenerated Cl · efficiently, with 3 Cl · photogenerated for each CH 4 oxidized with only 20% of Cl · recombining to Cl 2 . When more Cl 2 is added, the product balance shifts from mainly carbon monoxide to mainly carbon dioxide. Cl 2 -GPAO performs similarly under relative humidities between 5 and 60%. In the presence of ppm-level ammonia, toluene, nitric oxide, and hydrogen sulfide, CH 4 conversion decreases, suggesting that gaseous contaminants compete with CH 4 for Cl · . Gas chromatography–mass spectrometry revealed that Cl 2 -GPAO with 30 ppm CH 4 and Cl 2 produces 1.05 ± 0.32 part per billion (ppb) methyl chloride and 1.25 ± 0.38 ppb chloroform. Methylene chloride and carbon tetrachloride were not detected. We discuss the implications for cost-effective and climate-beneficial Cl 2 -GPAO at scale.

Abstract Using simultaneous magnetic field observations from 10 satellites and an automated detection algorithm, we identify broad regions of electromagnetic ion cyclotron (EMIC) wave activity during the initial phases of geomagnetic storms between September 2015 and October 2019. Since an initial phase typically drives compression of the dayside magnetosphere, we expect the majority of activity to be found here. However, in over 50% of initial phases examined in this study, there is EMIC activity in the nightside magnetosphere. Occurrence of this nightside activity increases as an initial phase progresses, with a lag of at least 35 min before it begins. Additionally, the shock impact angle, solar wind dynamic pressure, and substorm activity level have strong positive correlations to dayside EMIC activity rates compared to nightside. With these observations, we can characterize the extent of magnetospheric response in the form of EMIC wave activity throughout the initial phase of a geomagnetic storm.

Abstract We present observations from two consecutive TRACERS‐2 orbits through the northern low‐altitude cusp. During the first crossing, TRACERS‐2 observed reversed cusp ion dispersion and sunward convection, consistent with magnetopause reconnection tailward of the cusp during this northward IMF interval. Simultaneous THEMIS‐D observations at the equatorial magnetopause show heated magnetosheath plasma captured on closed field lines, with similar particle spectra as in the low‐altitude cusp, indicating that reconnection indeed occurred tailward of the cusp and in both hemispheres. When TRACERS‐2 traversed the northern cusp again, 95 min later, the IMF was dominated by a negative B X component. Despite the different IMF conditions, TRACERS‐2 recorded nearly the same cusp signatures as before, that is, reversed ion dispersion and sunward convection. The observations indicate that tailward‐of‐cusp reconnection can occur for both northward and B X ‐dominated IMF and that these distinct IMF geometries can produce remarkably similar plasma and field signatures in the low‐altitude cusp.