Using minute-scale observational data obtained from OTT Parsivel2 disdrometers deployed at four stations located at different altitudes, namely, plain, mountain foot, mountainside, and mountaintop, during a stratiform precipitation event over the Qilian Mountains from August 22 to 23, 2020, this case study systematically investigates the influence of topographic height on the microphysical characteristics of raindrop size distribution during this event. The results indicate that with increasing altitude, the raindrop size spectrum exhibits a systematic narrowing tendency. The mass-weighted mean diameter Dm decreases monotonically from 0.98 mm (plain) to 0.86 mm (mountaintop), whereas the normalized intercept parameter log10Nw increases from 3.60 to 3.97, and the total number concentration log10NT increases from 2.16 to 2.49. Conversely, the liquid water content W and precipitation rate R exhibit no statistically significant altitudinal variation. Precipitation at low altitudes is characterized by a broad spectrum, dominance of large drops, and high variability of parameters, which indicate convective characteristics. In contrast, precipitation at high altitudes exhibits a narrow spectrum, a high concentration of small drops, and exceptionally low parameter variability, characteristic of stable stratiform clouds. In the Dm–log10Nw phase space, the high-altitude stations are located within the “maritime-like” cluster, whereas the low-altitude plain station belongs to the “continental-like” convective cluster, suggesting a strong modulation of precipitation microphysical characteristics by topographic height. A significant altitudinal difference exists in the quadratic polynomial relationship between the shape parameter μ and the slope parameter Λ of the Gamma distribution, with high altitudes corresponding to high Λ and low μ, reflecting an enhanced contribution from ice-phase processes. The power–law relationship between the radar reflectivity factor Z and the precipitation rate R systematically shifts with increasing altitude; for the same Z, a lower R is observed at higher elevations. The application of a fixed Z–R relationship would therefore lead to an overestimation of precipitation in these elevated regions. This case study elucidates the gradient evolution mechanism of precipitation microphysical processes driven by topographic height and provides crucial localized constraints to improve quantitative precipitation estimation by radar and refining microphysical parameterization schemes in numerical models over complex mountainous terrain.
The microphysical properties of stratus cloud systems vary significantly across different regions and stages of development. This study investigates the microphysical characteristics of a postfrontal stratus cloud system with snowfall based on aircraft observations conducted on 1–2 March 2025 over Xingtai, Hebei Province, China. Distinct variability is observed, with thinner supercooled liquid clouds at the periphery and thicker clouds near the surface front, characterized by mixed-phase cloud below and ice cloud above. Clear differences are observed between the mixed-phase and ice-phase layers, as the mixed-phase layer shows higher RH and lower ice crystal number concentrations (Ni) and ice water content (IWC). Distinct vertical variations in supercooled cloud droplet properties, such as cloud droplet number concentrations (Nc), effective radius (re), and liquid water content (LWC), are also observed across different regions. Additionally, the ice crystal habit varies with altitude, with small dendritic and plate-like crystals aloft and large irregular aggregates dominating the mid- and lower layers, where small crystals possibly produced by second ice production are also observed. During the dissipating stage, decreasing RH and rising temperature within cloud correspond to reductions in Ni and IWC, along with broader particle spectra in both ice crystals and supercooled cloud droplets. These findings highlight the pronounced spatiotemporal variability of stratus cloud microphysics within the observed postfrontal system and suggest potential mechanisms in cloud evolution.
Precipitation across China is strongly modulated by complex topography and the East Asian monsoon, resulting in pronounced regional contrasts and spatiotemporal variability. Combined with the uneven distribution of rain gauges, these controls limit event-scale monitoring of precipitation and extremes and motivate rigorous evaluation of satellite precipitation products for hydrometeorological applications. Using daily observations from 835 gauges during 2001–2022, we evaluated GSMaP-Gauge-NRT and IMERG Early over mainland China using conventional continuous metrics and an event-based matching framework. Event performance was summarized using the event probability of detection (EPOD) and event false alarm ratio (EFAR), while misses and false alarms were further separated into timing-mismatched and isolated categories. Both products performed best in the humid eastern monsoon region and deteriorated markedly over the Tibetan Plateau and arid northwestern China. IMERG showed slightly greater event-detection sensitivity than GSMaP but also produced more false alarms, whereas GSMaP maintained lower false-alarm rates at the expense of more missed events in some regions. For GSMaP, residual errors were dominated by timing mismatches in humid eastern China and by isolated misses across northern China and the Plateau margins. IMERG showed a larger contribution from isolated false events and isolated misses in northern and high-elevation regions. Even among overlapping events, substantial inconsistencies remained, with GSMaP more often showing same-day boundary alignment and IMERG exhibiting earlier onset and stronger underestimation of peak and event-total precipitation, particularly over the first geomorphological step. These results identify region- and product-specific error modes and support targeted product selection, event-aware post-processing, and multi-source precipitation fusion for hydrological applications.
Understanding the interplay between emissions and meteorology is critical for air quality management, yet quantifying emission-related changes under unfavorable atmospheric conditions remains challenging. This research employs a natural experiment during the meteorologically demanding 2025 Victory Day military parade to evaluate the effectiveness and meteorological vulnerability of air quality control measures. Analysis of a decade (2016–2025) of high-resolution observed PM2.5 chemical speciation data from Beijing, alongside data from the Beijing–Tianjin–Hebei and its surrounding regions (“2 + 36 cities”), demonstrates substantial and sustained air quality enhancements. The “2 + 36 cities” experienced a 55.9% reduction in mean PM2.5 concentrations and a notable convergence of inter-city variability. Beijing's PM2.5 levels decreased by 50%, with secondary inorganic components, particularly nitrate (NO3−), becoming more prominent. Five major sources were consistently identified by the Positive Matrix Factorization model, with secondary formation and vehicular emissions remaining the most significant contributors. Meteorological normalization reveals that the 33.1% PM2.5 decrease during the 2025 study period, compared to the 2016–2024 baseline, was primarily associated with the anthropogenic-related component (−34.4%), with a minor meteorological contribution (+1.3%). However, on the parade day, stagnant and humid conditions, together with regional transport, resulted in a significant +43.2% meteorological penalty, largely negating the concurrent anthropogenic-related decrease (−35.7%). Furthermore, the meteorology-related contribution increased secondary inorganic components by 41.1%, counteracting their anthropogenic decline (−40.3%). This study provides evidence for an adaptive, sustainable air quality management framework in regions where secondary particulate pollution can be strongly amplified by unfavorable meteorological conditions.
In November 2025, an unprecedentedly intense vortex named Senyar developed over the Malacca Strait—a location typically unfavorable for severe storm intensification. Shallow straits promote upwelling and sea surface cooling, narrow fetch limits energy supply, and weak low-latitude Coriolis force constrains vorticity. Nevertheless, Senyar formed under these adverse conditions, bringing extreme rainfall, damaging winds, and landslides that caused over 1000 fatalities across southern Thailand, the Malay Peninsula, and northern Sumatra. This event ranks among the most extreme precipitation events in recorded history. Planetary-scale diagnosis reveals that constructive interactions between an equatorial Rossby wave and an active Madden-Julian Oscillation (MJO) phase provided a favorable large-scale envelope, while a concurrent negative Indian Ocean Dipole (IOD) supplied sustained moist energy. At the synoptic scale, the anomalously lower mid-tropospheric geopotential heights and persistent low-level moisture convergence triggered intense deep convection with high cloud tops, leading to a sustained low-pressure system. A clearly defined mesoscale vortex emerged within the broader convective envelope, exhibiting a compact circulation center and organized spiral rainbands. The system crossed the Malacca Strait twice: first after crossing the Malay Peninsula and making landfall in northeastern Sumatra, and again after re-emerging over the strait prior to its second landfall in Selangor. In terms of predictability, TIGGE ensemble members reveal systematic forecast biases: extreme precipitation was widely underestimated, and the simulated low-pressure center exhibited substantial track errors. These findings underscore the challenge of predicting topographically constrained tropical cyclogenesis in historically immune regions like the Malacca Strait—a threat that may grow under climate warming, demanding heightened vigilance and improved forecasting.
Abstract Extreme precipitation over the Tibetan Plateau (TP) has shown substantial variability in recent decades, yet previous studies have reported inconsistent trends in summer extreme precipitation, largely due to differences in the selected analysis periods and the difficulty in separating internal climate variability from externally forced signals. This study employs the large-ensemble versions of the two CMIP6 models, ACCESS-ESM1–5 and CESM2, with the externally forced and internally generated components rescaled using scaling factors derived from the optimal fingerprinting, to quantify the relative contributions of external forcing and internal variability to summer extreme precipitation over the TP. The rescaled single-model large ensemble method effectively separates external and internal components, improving correlation with true internal variability by ∼14.9% and reducing root-mean-square error by ∼61.8%. On the temporal scale, internal variability dominates high-frequency (5–20 years) fluctuations, while external forcing increasingly drives low-frequency trends when the time scale exceeds ∼37 years, with a transition around the early 21st century associated with anthropogenic forcing accumulation. On the spatial scale, internal variability plays a dominant role in shaping the detailed patterns of TP summer extreme precipitation, whereas external forcing primarily drives the overall intensification, particularly in the southwestern TP. These results provide new insights into TP summer extreme precipitation changes and support improved projections and debris-flow hazard prevention under ongoing climate change.
Understanding the impact of urbanization on precipitating systems’ characteristics, the present study is first of its kind, focused on rain microphysics during the Indian summer monsoon (ISM) season in the upwind, downtown, and downwind sites of Hyderabad urban sprawl and a surrounding reference (rural) station. During the ISM, precipitating systems are deeper over the upwind and downtown locations of Hyderabad than over the downwind and rural sites. Convergence, high water vapor, and moderate low-level vertical wind shear favor convective initiation across the city. Enhanced mid-tropospheric moisture, weak upper-level divergence, and relatively weak shear further support the growth of deeper convective systems over the upwind and downtown areas. In contrast, stronger mid-tropospheric wind shear over the rural and downwind regions suppresses vertical cloud development and lowers storm-top heights. The upwind and rural regions contain abundant small raindrops; however, the upwind region also shows a higher concentration of large raindrops, whereas the rural region has fewer large drops than the downtown and downwind areas. Despite this, convective rain mass-weighted-mean-diameter (Dm) differs little between the upwind and rural sites because the reduced number of medium-sized drops in the upwind region shifts Dm toward smaller values. In stratiform rain, the downwind region exhibits larger Dm values than other locations, while downtown and rural regions show similar Dm. The ISM low-level jet transports continental air into Hyderabad, giving the urban raindrop size distribution a continental character. Urban-induced collisions among small droplets enhance the formation of medium and large raindrops over downtown areas.
The recent intensification of extreme precipitation events in the western Mediterranean basin raises questions about the physical mechanisms linking global warming to local hydrological response. This study analyzes the climate features and evolution of the western Mediterranean water-vapor recirculation cell and its role as a modulator of torrential rainfall on the Levante coast. Using ERA5 reanalysis data and daily station precipitation (1950–2024), we characterize a hydrological memory mechanism that links summer dynamics to autumn extremes. The results show that the closed circulation in summer acts as a moisture trap (r = 0.51 between Zonal Circulation and Recharge), creating a reservoir of potential energy that conditions the severity of rainfall months later. A robust predictive correlation (r = 0.52) was verified between cumulative summer recharge (Rec) and extreme autumn precipitation (Rp99), ruling out simultaneous local evaporation (r ≈ −0.05) as the main driver.Secular trend analysis reveals a highly nonlinear climate response. While summer water vapor content (TCWV) has increased by 6.4%, remaining in a Sub-Clausius-Clapeyron regime due to regional subsidence, extreme precipitation (Rp99) has experienced a disproportionate amplification of +68.0%. This discrepancy in magnitude indicates that the system has transitioned to a more efficient regime, in which the intensification of the blocking mechanism (+0.65 σ) and net accumulation (+1.57 σ) act as multipliers of the thermodynamic signal. We conclude that the increase in torrentiality does not respond linearly to a warmer atmosphere, but rather to the dynamic consolidation of an accumulation mode that concentrates available vapor more efficiently.
Particulate nitrate (NO3−) in coastal air affects the nitrogen cycle and cloud condensation nuclei, yet its chemical formation processes, especially with respect to air mass transport and local emissions, are poorly understood. To clarify NO3− sources and formation mechanisms, nitrogen and oxygen isotopic compositions (δ15N and Δ17O) for NO3− and Δ17O of ozone (O3) were measured at a mountain site located in Tai Mo Shan (640 m a.s.l.) and δ15N and Δ17O of NO3− were measured at an urban site located in Tsim Sha Tsui (60 m a.s.l.) in Hong Kong, a coastal megacity of southern China. Average δ15N-NO3−, Δ17O-NO3−, and Δ17O-O3 at the mountain site were −1.1 ± 2.1‰, 22.4 ± 1.1‰, and 24.1 ± 1.4‰, respectively, while urban δ15N-NO3− and Δ17O-NO3− were 0.8 ± 1.3‰ and 21.1 ± 0.9‰. Bayesian modeling identified the NO2 + OH reaction as the dominant nitrate formation pathway at both sites, highlighting the role of photochemical processes in coastal environments. N2O5 hydrolysis was more prevalent at the humid mountain site, and HC/DMS/XNO3 pathway influenced urban areas through hydrocarbon emissions and marine air masses. Natural gas and coal combustion emerged as the predominant contributors to nitrate aerosols at Tai Mo Shan, where the absence of local emissions underscored regional transport of these aerosols, while NOx emissions from ships and vehicles dominated urban NO3− sources. Transitioning to clean fuels and electric vehicles is vital for reducing urban NOx emissions and associated health risks from nitrate particles.
Regional air-quality models in tropical Southeast Asia commonly transport PM2.5 as a passive scalar, but the particle-size threshold at which this assumption breaks down has not been quantified for tropical surface-layer turbulence. This study derives a physics-based critical particle-size threshold for inertial behaviour in Thai surface-layer conditions and uses PM2.5 as a fine-aerosol verification case. Across the deterministic density envelope (ρp=1100–2100 kg m−3), the resulting mean transition diameters ranged from 27.39 to 37.85μm in Bangkok and from 45.76 to 63.22μm in Chiang Mai, well above the PM2.5 range. ERA5-derived turbulence estimates, site-specific moist-air properties, deterministic particle-size–density cases, Monte Carlo uncertainty propagation, station-level volume-fraction analysis, and surface-wind validation were combined to evaluate Kolmogorov microscales, Stokes number (St), settling velocity parameter (Sv), and critical transition diameter.The transition scale remained consistently above the PM2.5 range. ERA5-based Kolmogorov length scales over Thailand were in the sub-millimetre to millimetre range, confirming that PM2.5 remains far below the smallest dynamically relevant turbulent length scale. Across all 375 deterministic PM2.5 cases, the maximum upper-tail St was 6.366×10−3, below the strict passive-tracer bound of St=0.01, and no case entered the inertially active regime, although some upper-tail cases occupied the near-tracer range as the settling parameter approached, but did not exceed, its near-tracer bound. Particle volume fractions also remained far below the one-way coupling threshold, with a broader station worst-case value of 1.75×10−10.These results indicate that unresolved particle inertia is unlikely to be a limiting source of uncertainty for Thai PM2.5 simulations compared with emissions, atmospheric transport, boundary-layer dynamics, chemistry, deposition, and data assimilation. Coarse particles approaching the 30–60μm transition range, including road dust, agricultural debris, pollen fragments, and large biomass-burning aggregates, may require explicit inertial or gravitational treatment in atmospheric aerosol-transport models. The workflow provides a transferable, dynamics-based method for identifying aerosol–turbulence regime boundaries in monsoonal tropical surface-layer environments.
The Northeast China cold vortex (NECV) is a pivotal synoptic-scale system in East Asia, frequently triggering severe convective weather such as thunderstorm gales (TSG), short-duration heavy rainfall (SDHR) and mixed convective events (MIX). However, forecasting these events remains challenging because different convective types exhibit distinct spatial distributions and cloud development characteristics, while their satellite-observed cloud properties have not been systematically investigated. This study utilizes FY-4 A multi-spectral observations combined with ground-based measurements to analyze the spatial distributions of different types of severe convective events and the corresponding cloud characteristics under NECV conditions during 2021–2023. The results show that SDHR events mainly occur within 10–20° latitude of the vortex center, where the interaction between mid-level dry cold air and moist inflow from the subtropical high favors the development of larger and deeper convective clouds. In contrast, TSG events are primarily located closer to the vortex center, particularly at dry intrusion edges and steep geopotential height gradients, featuring rougher cloud-top texture, lower heights, and smaller spatial extent. During the cloud growth phase, SDHR and MIX related clouds exhibit larger variations in brightness temperature difference between the 6.25 and 10.8 μm channels (BTD6.25–10.8) and between the 12.0 and 10.8 μm channels (BTD12.0–10.8), signaling greater optical thickness, deeper cloud structures, and a higher proportion of ice-phase particles. In contrast, TSG-related clouds show limited cloud-top microphysical development. Although SDHR and MIX have similar multispectral attributes, they differ in their horizontal area expansion patterns. SDHR demonstrates concurrent expansion of the convective cloud area (≤ − 32 °C) and deep core (≤ − 52 °C) with the former being more vigorous, whereas MIX shows more consistent growth between the two regions at a smaller spatial scale.
The impact of cloud–radiation forcing (CRF) on tropical cyclone (TC) intensity under vertical wind shear (VWS) is investigated using Typhoon Lekima (2019), which went through rapid intensification (RI). Numerical simulations are conducted using the WRF model, including the control experiment, an experiment with all CRF excluded, and a suite of sensitivity experiments where CRF associated with individual hydrometeor species is selectively excluded.The control experiment realistically captures the symmetrization processes and RI of Lekima. Cloud–radiation forcing moistens the mid-troposphere by forcing local ascent, particularly on the upshear side. This radiatively driven moistening reduces the drying effect of the mid-level ventilation, thereby supporting convective development and enabling the expansion of the upshear outflow. The strengthened outflow carries more hydrometeors into the upper-level layer, where their radial spreading enhances both the intensity and extent of CRF. Together, these processes establish a positive feedback that promotes the intensification of Lekima.When total CRF is removed, Lekima's intensification is delayed. A comparable delay occurs in the experiment that removes only the CRF from ice and snow particles, indicates that these hydrometeors play a dominant role in the CRF-related impact on RI. Without mid-level moistening induced by the CRF of ice and snow particles, enhanced radial ventilation occurs on the upshear side, hindering the development and maintenance of downwind convection. The radiative effects of liquid hydrometeors, by contrast, exert only a minimal influence on the intensification.
The Freshwater Lake model (FLake) has recently been integrated into the land surface scheme of ERA5 and subsequently downscaled to ERA5-Land, enabling the investigation of global-scale lake thermodynamics at an enhanced spatial resolution. Although ERA5-Land's capability in simulating lake surface water temperature (Ts) has been recognized, its performance in reproducing other critical lake variables remains inadequately explored. This study comprehensively evaluated ERA5-Land atmospheric forcings and lake products against long-term observations at Lake Taihu, a large, subtropical shallow lake. Results show that ERA5-Land captures the surface microclimates reasonably well at Lake Taihu across diurnal to annual scales. Satisfactory performance is also achieved in simulating monthly and annual latent heat fluxes (λE) with a positive bias of 1.0 W m−2 (or 1% of the annual mean). However, water stratification simulated by ERA5-Land is excessively strong in spring and summer. ERA5-Land systematically overestimates downward shortwave radiation (+16 W m−2) and underestimates downward longwave radiation (−13 W m−2) due to cloud cover underestimation. These intrinsic radiative biases only introduce a marginal error in simulated Ts and λE due to the compensation effect between downward shortwave and longwave components. Our proposed correction algorithm can reduce monthly radiative biases both for an independent period at Lake Taihu and at geographically independent Poyang Lake. Notably, despite lacking lake-specific parameter optimization, the default ERA5-Land configuration still outperforms two other FLake simulations using satellite-calibrated parameters in reproducing the multi-year mean Ts. Furthermore, Flake model in ERA5-Land simulates an identical lake warming trend (0.26 °C decade−1 from 1981 to 2020) and a comparable evaporation changing rate (2.9 W m−2 decade−1 from 1979 to 2013) to other independent simulations with tuned parameters, highlighting its reliability for investigating long-term trends in lake thermodynamics.
China's 10-m wind speed (WS10) is widely considered to have recovered after decades of decline, but this inference is based mainly on regional mean series that may obscure substantial local heterogeneity. Using observations from 303 meteorological stations during 1980–2024, we identify divergent transitions in seasonal WS10 trends across China. Significant trend reversals are detected at nearly 90% of stations and can be broadly classified into two categories: Type A, characterized by shifts towards more positive trends, and Type B, characterized by shifts towards more negative trends. Type-A stations predominate in all seasons, accounting for 60.0%–70.0% of stations, whereas Type-B stations account for about 20%. Turning years at Type-A stations are concentrated from the late 1990s to the early 2000s, while those at Type-B stations cluster around 2000 and again during 2010–2020. At Type-A stations within the 0–2000 m elevation band, turning years advance approximately five years per 1000 m elevation increase in MAM, SON, and DJF, with correlations remaining significant after controlling for latitude via partial correlation. Type-B stations exhibit only weak positive elevation correlations within 0–1000 m, but these lose statistical robustness after accounting for latitude. Further analysis indicates that WS10 reversals at Type-A stations are associated with strengthened 925-hPa winds and a deeper boundary layer, whereas WS10 trend shifts at Type-B stations are linked to weakened boundary-layer mixing processes, accompanied by lower-level circulation weakening. These results challenge the narrative of spatially uniform decline–recovery and underscore the importance of station-scale evidence for wind-energy assessment and climate adaptation.
Near-surface 10-m wind speed in the Weather Research and Forecasting model (WRF) over complex terrain exhibits a speed-dependent, two-sided error, characterized by overprediction at weak winds and underprediction at strong winds. The weak-wind overprediction is consistent with insufficient parameterized drag from unresolved terrain. Using WRF simulations and dense surface observations over southwestern China, we show that the residual bias is organized more strongly by the bulk Richardson number than by subgrid terrain standard deviation. The largest diagnosed errors occur mainly at small absolute values of the model-diagnosed bulk Richardson number, although this signal is partly sensitive to modeled wind speed in its denominator. A revised turbulent orographic form drag (TOFD) scheme is developed that replaces the static terrain-variance-based amplitude with a stability-dependent function of the bulk Richardson number. Across four validation windows, the revised scheme reduced RMSE by 12.0–13.5% (weighted mean 12.8%) relative to the default TOFD, with the largest gains under near-neutral, weak-to-moderate wind conditions.
Rainwater serves as an important pathway for the transport of microorganisms from the atmosphere to surface, yet the concentration, viability, and community composition of rainwater bacteria remain insufficiently understood. We quantified bacterial concentration and viability in rainwater from coastal Qingdao, China (2021–2022 and 2024–2025), and the Northwest Pacific Ocean (NWPO) in summer 2022. Bacterial community composition and its environmental drivers across climate zones were examined based on our results and reported global rainwater samples. Bacterial concentrations in Qingdao rainwater, sampled synchronously with the NWPO samples, were significantly higher than those in rainwater over the NWPO, with mean concentrations of (4.42 ± 3.14) × 103 and (0.12 ± 0.03) × 103 cells·mL−1, respectively. In Qingdao, bacterial concentrations were higher in spring and autumn (4.71 × 103 and 3.83 × 103 cells·mL−1) than in summer (2.58 × 103 cells·mL−1). Concentrations increased by 3.08-fold in snow samples and by 4.74-fold during dust-affected events. Bacterial viability averaged 11.0 ± 9.3% in Qingdao but reached 27.1 ± 7.6% over the NWPO. Community composition also shifted markedly by season, from Proteobacteria dominance in spring (52.9%) to Bacteroidota enrichment in summer (37.5%), autumn (43.5%), and winter (51.5%). At the global scale, Proteobacteria decreased from 76.4% in tropical samples to 34.1% in polar samples, whereas Actinobacteriota increased to 24.6% in polar regions. Partial least squares path modeling (PLS-PM) showed that natural emissions were strongly associated with bacterial concentration (0.350) and α-diversity (0.462), whereas bacterial viability was jointly associated with natural emissions (0.503), human activities (0.629), and hydrometeorological stress (−0.337). Overall, these findings show that different characteristics of rainwater bacteria were associated with distinct combinations of environmental factors: natural emissions were mainly linked to bacterial concentration and α-diversity, whereas viability was jointly associated with natural emissions, human activities, and hydrometeorological stress.
Understanding the drivers of North American summer surface air temperature (SAT) variability and persistent extreme heat days (PEHDs) is critical for climate research. Building on prior findings of zonal intraseasonal waves over the North Pacific that disperse energy into North America, this study examines their influence on 10–30-day SAT variations and PEHDs over North America. Results reveal that there are two primary SAT modes: a “west-east” dipole pattern and a “southwest-northeast” monopole pattern, with the dipole pattern showing a tendency to evolve into the monopole by about 3–4 days. Further analysis suggests that these two SAT patterns are modulated by intraseasonal waves over the North Pacific, which cause cyclonic and anticyclonic circulation anomalies over North America through energy dispersion. In addition, the findings also show that the wave train can significantly increase (decrease) the frequency of PEHDs in the region affected by the anticyclonic (cyclonic) circulation anomalies of the waves. These findings offer fresh insights into the mechanisms driving extreme summer heat, advancing the accuracy of prediction.
Extreme hydrometeorological events, including freezing rain and ice pellets, pose severe threats to countries across the Asian monsoon region. This was underscored by the unprecedented 2008 freezing disaster and recent extreme freezing events in 2025/2026 over southern China. While previous studies have highlighted the role of atmospheric rivers (ARs) in modifying precipitation types in other regions, research focusing on the populous Asian monsoon region, a hotspot of such high-impact events, remains insufficient. Here, based on an AR detection tool (ARDT) named ARIA-Asia v1.8 validated against seven ARDTs, we demonstrate that during 1961–2023, ARs account for over one-third of the liquid rain frequency in key coastal areas of East Asia (e.g., southern Japan, the Yangtze River Estuary), account for up to 50% of freezing rain events in eastern China, and account for 26%–44% of ice pellet events over the Yangtze River basin and southern China. Trends in AR activity have reduced liquid rain frequency across mid-latitude East Asia while significantly increasing AR-associated solid and mixed-phase precipitation in southern China, particularly during the cold season. Using five machine-learning algorithms, feature importance analyses reveal that trends in AR frequency contribute approximately 53% of the total feature importance for AR-associated liquid rain, whereas trends in solid and mixed-phase precipitation are jointly governed by thermodynamic (e.g., sensible/latent heat fluxes) and microphysical (e.g., cloud ice, snow water content) processes. Our findings demonstrate that ARs are key weather systems reshaping the regional distribution of hydrometeors and high-impact precipitation types in Asia. These findings provide critical insights for improving the prediction of hydrometeor-related disasters and informing mitigation strategies under climate change.
Utilizing ground-based aerosol observations collected from 19 August 2024 to 19 August 2025, we analyzed the characteristics of aerosol particle size distributions (PSDs) at the summit of Mount Liupan during non-precipitation periods. The average aerosol number concentration spectrum exhibited a trimodal distribution, characterized by the presence of coarse particles exceeding 1000 nm. Distinct diurnal and monthly variations were observed in the number concentrations across the nucleation (5–20 nm), Aitken (20–100 nm), accumulation (100–1000 nm), and coarse (>1000 nm) modes. Generally, diurnal variations were characterized by higher concentrations during the daytime and lower levels at night, although the specific evolutionary patterns varied among different modes. The nucleation mode particles exhibited a unimodal distribution; their number concentration increased rapidly after sunrise and peaked at 14:00 Beijing Time (BJT). The peak for the Aitken and accumulation modes lagged behind that of the nucleation mode. Furthermore, the changes in aerosol PSD under varying precipitation intensities were analyzed, revealing that the reductions in aerosol particle number concentrations generally became more pronounced as precipitation intensity increased. Finally, aerosol PSDs across different times, months, and precipitation stages (before, during, and after precipitation) were fitted using multiple log-normal distributions. Overall, the multiple log-normal fits showed good performance. Except for the during- and post-heavy-rain spectra, which yielded R2 values of 0.8258 and 0.8879, respectively, the R2 values of all other fitted spectra ranged from 0.9121 to 0.9997. This provides localized aerosol spectral parameters as foundational references for numerical simulations in this region.
