As an important component of the cryosphere， snow cover is an indispensable part of the Earth's system and also an "indicator" of global climate change.The impact of snow cover on the climate system mainly originates from its physical properties such as high albedo and low thermal conductivity.Since the climate system is highly sensitive to changes in surface albedo， slight variations in surface albedo can significantly affect the energy balance of the climate system， thereby rapidly influencing the atmospheric conditions on the Qinghai-Xizang Plateau （QXP） and its downstream areas through the snow albedo feedback mechanism.Given that snow cover on the QXP is relatively thin and characterized by repeated accumulation and ablation， this paper aims to conduct an in-depth study on the variation characteristics of soil hydrothermal processes and surface energy fluxes on the QXP under snow-covered and snow-free conditions.Based on daily-scale data， this study explores the impact of snow cover on surface micrometeorological characteristics under different underlying surface types by setting a criterion that albedo is greater than 0.5.The research results show that the multi-year average surface albedo of the QXP is 0.22， exhibiting a spatial distribution characteristic of "high in the northwest and low in the southeast" and a seasonal variation pattern of "high in winter and spring， low in summer and autumn".The surface albedo of the QXP is affected by snow cover， with significant regional and seasonal differences.The area of perennial snow cover is small， accounting for only 0.55% of the total area.This paper selects Naqu Station， Namors Station， and Yakou Station located in different climate zones to analyze the soil hydrothermal characteristics and surface energy flux characteristics under snow-covered and snow-free conditions.The main conclusions are as follows： （1） When snow cover exists， the surface albedo at noon usually exceeds 0.6， while under snow-free conditions， the albedo at noon is usually lower than 0.3； （2） When snow cover exists， the soil hydrothermal synergistic effect is relatively stable.Specifically， at Naqu Station， although the snow cover state is unstable and the heat preservation effect is weak， the presence of snow cover can reduce the fluctuation of soil temperature.The snow cover states at Namors Station and Yakou Station are stable， and the heat preservation effect is more obvious.Compared with the snow-free conditions， the soil temperature and soil water content are higher， the variation range is smaller， and the soil freezing depth is shallower； （3） Under snow-covered conditions， the closure rate of surface energy is low， and the correlation between turbulent flux and effective energy is strong.Due to the different snow cover conditions at Naqu Station and Yakou Station， there are obvious differences in surface energy distribution between the two stations.Regardless of snow cover， the net radiation at Naqu Station is mainly allocated to sensible heat flux， but shallow snow cover increases the proportion of latent heat flux； at Yakou Station， snow cover is continuous.When there is snow cover， net radiation is mainly allocated to latent heat flux， and when there is no snow cover， net radiation is mainly allocated to sensible heat flux； when there is snow cover， the underlying surface is wet， and the Bowen ratio is mostly below 1.0， while when there is no snow cover， the Bowen ratio is larger； （4） In winter， when there is snow cover， the heat preservation effect of snow cover makes the soil temperature higher than the atmospheric temperature， and the soil heat flux is mainly transmitted to the atmosphere.As the snow gradually melts， the energy received by the surface gradually increases.

The stable isotopes of hydrogen and oxygen in precipitation over the Qinghai-Xizang Plateau and their related proxy data such as ice cores and tree rings are of great significance for a deeper understanding of the climate and hydrological changes on the Qinghai-Xizang Plateau.The introduction of stable water hydrogen and oxygen isotope modules in atmospheric circulation models helps understand the climatic interpretations of precipitation isotopes and related paleoclimate proxy data.In recent years， some studies have applied nudging techniques to these models to reduce model errors by fusing reanalysis data.However， further research is needed to determine the extent to which this method improves the simulated precipitation isotopes.This study used data from the Qinghai-Xizang （Tibetan） Plateau Network Isotope Plan （TNIP） to evaluate four atmospheric models with water isotope modules and compare the results with and without nudging reanalysis data.The results indicate that all four models can well reproduce the climate mean annual precipitation δ ¹⁸O and seasonal variability of precipitation δ ¹⁸O， while the models have lower simulation ability for interannual variability of precipitation δ ¹⁸O.The simulations of the spatial distribution and seasonal cycle changes of precipitation δ ¹⁸O over the Qinghai-Xizang Plateau with nudging reanalysis data is significantly better than that of free running simulations.The differences in model choice have a greater impact on the simulation results than the differences in nudged reanalysis data.Overall， the fusion of reanalysis data can effectively correct the model bias in the atmospheric circulation over and around the Qinghai-Xizang Plateau， thereby improving the model's ability to simulate precipitation and precipitation δ ¹⁸O.This study provides a useful reference for simulating stable isotopes of precipitation over the Qinghai-Xizang Plateau using models and the improvement effect of nudging techniques on the hydrological and climatic simulation of the Qinghai-Xizang Plateau.

The variation in water budget of closed lakes serves as an ideal research subject for studying lake responses to climate change.However， most existing studies focus on the long-term water volume changes of lakes on Qinghai-Xizang Plateau.Due to data scarcity， the intra-annual dynamics of water budgets and the contributions of balance components in these lakes remain unclear， hindering research on water budget dynamics and climate change responses.Base on hydrological-meteorological observation data， reanalysis datasets， water balance equations， and inflow runoff calculation methods， this study analyzes the dynamic characteristics of water balance components ［lake surface precipitation （P）， inflow runoff （ R_s）， lake surface evaporation （E）］， lake level （H） and lake level difference（ΔH）） in Bamu Co during the 2021 -2023 monsoon periods； the relative contributions rates of water balance factors in different months and years； along with the interannual variations in meteorological elements including air temperature， downward radiation， and relative humidity.The results show that： （1） The water budget of Bamu Co exhibits significant interannual variation.In 2021， abundant P and R_s drove in H.In 2022， sharp declines in P and R_s led to a drop in H.In 2023， increases in P and R_s offset the intensified E， resulting in another rise in H.（2） ΔH is mainly regulated by R_s and E.On the monthly scale， the contribution of R_s displays a “V”-shaped， while E shows an opposite trend.On the interannual scale， R_s plays a dominant role in positive variations （2021 and 2023）， whereas E dominates negative variations （2022）.（3） Meteorological elements influenced ΔH through P-E balance adjustments： moderate temperature， high humidity， and medium radiation in 2021 maintained medium E/P ratio with positive ΔH （+231 mm）； elevated temperature， reduced humidity， and enhanced radiation in 2022 increased E/P ratio alongside decreased P and R_s， resulting in negative ΔH （-75 mm）； improved meteorological conditions in 2023 reduced E/P ratio and restored hydrothermal conditions， driving positive ΔH （+350.6 mm）.This study provides crucial insights into unique hydrological mechanisms of monsoon-influenced lakes， supports ecological security barrier construction for the "Asian Water Tower"， and promotes sustainable development on the Qinghai-Xizang Plateau.

Using an objectively identified dataset of Tibetan Plateau Vortex （TPV） based on three reanalysis datasets and observations， combined with the Yearbook of TPV， this study analyzes the spatiotemporal variations of TPV activity from 1979 to 2022 and their possible relation with climate warming over the Tibetan Plateau.Results indicate that the genesis and moving off numbers of TPV have mutations in 1997 and 2002， respectively.In the past 20 years， although genesis and moving off numbers of TPV have significantly decreased， the intensity of individual TPV， presents a significant increasing trend.After removing the effects of climate warming， the genesis number of TPV in the past 20 years has exhibited a significant negative correlation with the daily temperature range and a significant positive correlation with atmospheric heat sources over the entire layer in the Tibetan Plateau.The significant weakening of the averaged atmospheric heat source over the entire layer in the Tibetan Plateau over the past two decades is the main reason for the reduction in the numbers of TPV.In contrast， the significant increase in the intensity of individual TPV is significantly positively correlated with the significant increase of the precursor accumulated atmospheric heat sources before TPV genesis.Results of this study suggest that although the TPVs have decreased under the climate warming， the extremes of individual TPV and their impacts on precipitation over the Tibetan Plateau and its surrounding regions， have been strengthening.

The Qinghai-Xizang Plateau， due to its high altitude， exhibits unique lightning and convective activity compared to lower-altitude regions.Using the ‘Group’ data from the Lightning Mapping Imager （LMI） on the Fengyun-4A satellite， a statistical analysis of the spatiotemporal distribution of lightning activity on the plateau was conducted for the summers （June to August） from 2019 to 2023.The results reveal significant geographical variations in lightning frequency across the Plateau， with high-frequency areas predominantly located in the eastern， central （north of the Himalayas， in and around the city of Nagqu）， and southern （south of Shannan and Linzhi） regions.In terms of temporal distribution， lightning activity is most prevalent in June and July.Specifically， June sees concentrated lightning activity in the eastern and southern regions， while July experiences an increase in the central region and a shift of activity in the eastern region towards the southeast and northeast.By August， lightning activity decreases in the eastern region but increases in the southern and central regions.Regarding the daily distribution， the evening and night hours ［18:00 （Beijing Time， same as after） to 03:00 the next day］ are identified as the peak periods for lightning activity on the Qinghai-Xizang Plateau.During the daytime， the intensity of lightning radiation exhibits a 'double-peak' characteristic， frequently exceeding 400 \mathrm{\mu } \mathrm{J} \bullet {\mathrm{m}}^{-\mathrm{2}} \bullet \mathrm{s} \mathrm{t} \mathrm{e} {\mathrm{r}}^{-\mathrm{1}}.Conversely， at night， the intensity of lightning radiation generally follows a 'single-peak' pattern， usually remaining below 400 \mathrm{\mu } \mathrm{J} \bullet {\mathrm{m}}^{-\mathrm{2}} \bullet \mathrm{s} \mathrm{t} \mathrm{e} {\mathrm{r}}^{-\mathrm{1}}.Notably， the intensity of lightning radiation is significantly higher during the day than at night.Most lightning events recorded were less than 100 \mathrm{\mu } \mathrm{J} \bullet {\mathrm{m}}^{-\mathrm{2}} \bullet \mathrm{s} \mathrm{t} \mathrm{e} {\mathrm{r}}^{-\mathrm{1}}， with a five-year average of 67.13%.In terms of daily trends， lightning activity in the central and eastern parts of the Qinghai-Xizang Plateau shows similar patterns， with peaks occurring around 20:00 -21:00 and 02:00.In contrast， the southern part of the Plateau experiences peak activity around 05:00 -06:00.The highest lightning activity across all three regions was observed in July， with the southern region displaying more activity in August compared to June and July， unlike the central and eastern regions.This study reveals the spatiotemporal distribution of lightning activity on the Qinghai-Xizang Plateau， providing a scientific basis for better understanding climate change and disaster prevention in the region.

In the context of global warming， it is of great significance to clarify the change characteristics and trends of hydrological elements in the mountainous areas of inland river basins to ensure water resource security.In this study， the spatiotemporal variation of hydrological elements under four scenarios， SSP1-2.6， SSP2-4.5， SSP3-7.0 and SSP5-8.5， was predicted with the SWAT model and the collective average data of five global climate models in CMIP6.The results showed that： （1） The evaluation coefficient of the SWAT model was higher in the calibration period （NSE=0.92， R ²=0.93， PBIAS=-7.09%） and validation period （NSE=0.89， R ²=0.91， PBIAS=4.74%）， indicating that SWAT had good applicability in the simulation of runoff in the upper reaches of the Heihe River Basin.（2） Under the four scenarios， the runoff from the mountains in the future will increase by 12.2%， 8.1%， 10.4% and 19.2% respectively compared with the base period， and the runoff will increase significantly in autumn and winter.In the near and far future， the increase in the average total water yield in the basin is between 6.2~25.4 mm （22.2~35.7 mm）， and the increase in the average underground flow is 1.6~7.4 mm （7.4~12.1 mm）， and the increase of each hydrological element is greater in the far future.（3） In terms of the spatial distribution of hydrological elements， the spatial distribution of precipitation， evapotranspiration， surface production and underground runoff increased from northwest to southeast， while the total water yield and lateral flow were high in the middle and low in the north.（4） The spatial distribution pattern of the changes of each hydrological element is quite different， and the spatial distribution difference of the change of hydrological elements between different scenarios in the near future （2021 -2060） period is relatively small， and the difference in the temporal and spatial distribution of the change in the far future （2061 -2100） is even greater.In conclusion， the SWAT model can better describe the temporal and spatial changes of hydrological processes and hydrological elements in the mountainous areas of inland river basins.

To investigate the long-term variation characteristics of actual evapotranspiration （ET_a） in different climatic regions of the Qinghai-Xizang Plateau and its main influencing factors， as well as to enhance the understanding of land-atmosphere interactions and eco-hydrological processes in the plateau region， this study selected MAWORS， BJ， QOMS， and SETORS as representative observational sites.Based on long-term site observations and satellite remote sensing data， the ET_a at different temporal scales was calculated for each site， and its correlation with meteorological factors was analyzed.Furthermore， a path analysis method was employed to quantify the impact of environmental factors on daily ET_a during the monsoon season.The results showed that： （1） In terms of annual ET_a， BJ exhibited the highest value， with a multi-year average of 592.17 mm， followed by SETORS at 521.34 mm.MAWORS ranked third， with an annual ET_a of 422.84 mm， while QOMS had the lowest ET_a of only 206.33 mm， significantly lower than the other sites.（2） Regarding the ratio of annual ET_a to precipitation， MAWORS had the highest value at 3.34 due to its low precipitation.BJ followed with a ratio of 1.11， while QOMS had a ratio of 0.96， indicating a near balance in local water budget， with precipitation being the primary water source for ET_a.SETORS exhibited the lowest ratio of 0.68， suggesting a strong water retention capacity of the underlying surface， with a surplus of precipitation conducive to water resource accumulation and vegetation growth.（3） The interannual variation trends showed that annual ET_a at MAWORS， BJ， and SETORS displayed an increasing trend.The increase at MAWORS was mainly attributed to the continuous enhancement of net radiation， while the rise in BJ was closely related to an increase in soil moisture content.The upward trend at SETORS was primarily driven by rising temperatures.In contrast， QOMS showed a decreasing trend in ET_a， mainly due to a reduction in precipitation during the monsoon season.（4） ET_a at different sites was controlled by various environmental factors.At MAWORS， ET_a was primarily regulated by net radiation and soil moisture content， with glacier meltwater serving as a significant water source.At QOMS， where the terrain acts as a barrier to water vapor transport， ET_a was predominantly controlled by precipitation， showing a clear water-limitation effect.At BJ， ET_a was jointly regulated by net radiation and soil moisture， with a stronger energy-limitation effect.At SETORS， abundant precipitation resulted in weak water limitations on ET_a， making it mainly influenced by net radiation and temperature.（5） The regulatory effect of vegetation cover on ET_a was more evident at BJ and SETORS， whereas at MAWORS and QOMS， the correlation was weaker due to sparse vegetation.（6） During the non-monsoon period， water vapor supply decreased significantly at all sites.Except for SETORS， soil at the other sites remained frozen， with soil moisture content reaching its lowest level of the year.Consequently， the ET_a process was mainly controlled by net radiation.At SETORS， where water vapor transport was reduced， ET_a relied more on precipitation supply.

To investigate the relationship between sensible heat （SH） over the Qinghai-Xizang Plateau （QXP） and TP monsoon dynamics， this study employs station-derived SH data （1979 -2016） alongside ERA5 reanalysis datasets to compute multiple QXP monsoon indexes.Using correlation and composite analyses， the spatiotemporal variability and interactions between SH and monsoon systems were explained.The primary findings are summarized as follows： （1） The two monsoon indexes （IPM_tang and IPM_zhang） exhibit divergent temporal responses to SH variability， attributable to their distinct definitions.Summer SH and summer monsoon intensity show a significant negative correlation with IPM_zhang （r=-0.30， p<0.1） but only a weak， statistically insignificant correlation with IPM_tang.Conversely， May SH correlates positively with summer IPM_tang （r=0.32， p<0.1） and negatively with summer IPM_zhang （r=-0.44， p<0.01）， underscoring May SH’s critical role in preconditioning monsoon intensity.（2） For summer SH and monsoon interactions， IPM_tang reveals an east-west dipole spatial correlation with SH， whereas IPM_zhang displays a north-south contrast.Despite regional disparities， both indexes align with monsoon-precipitation linkages， reflecting soil moisture feedback that suppresses SH through monsoon-driven precipitation.（3） Elevated May SH across the southern QXP initiates an upper-tropospheric anomalous low-pressure system over the central QXP in summer， generating a midlatitude “- + - + -” wave train.Coupled with a low-level ascending vertical structure and anomalous mid-upper tropospheric pressure systems， these dynamics establish a synergistic thermodynamic framework conducive to precipitation in the southeastern QXP.A 600 hPa anomalous low-pressure system over the southeastern QXP and intensified westerlies over the northern QXP further strengthen IPM_tang while weakening IPM_zhang， demonstrating May SH’s pivotal influence on summer monsoon modulation.

Due to its unique geographical environment， Qinghai Province is highly susceptible to frequent hail events.Considering the complex topography of high-altitude regions， particularly the Qinghai Plateau， this study constructs a hail forecasting dataset by integrating hail observations from 52 meteorological stations in Qinghai from 2009 to 2023， corresponding hail disaster records， and the ERA5 atmospheric reanalysis dataset.Based on this dataset， three ensemble decision tree models-Random Forest， XGBoost， and LightGBM-are employed to develop a hail forecasting model， with separate analyses conducted on hail samples with diameters of ≥2 mm and ≥5 mm.Experimental results demonstrate that the LightGBM model consistently outperforms both Random Forest and XGBoost， with particularly superior performance in forecasting large hail events （diameter ≥5 mm）.Specifically， for small hail samples （diameter ≥2 mm）， the LightGBM model achieves a hit rate of 0.923， a false alarm rate of 0.041， a Critical Success Index （CSI） of 0.858， an accuracy of 0.946， and a recall rate of 0.924， while for large hail samples （diameter ≥5 mm）， it attains a hit rate of 0.938， a false alarm rate of 0.038， a CSI of 0.908， an accuracy of 0.960， and a recall rate of 0.964.Further analysis of the hail forecasting model in the complex terrain of the plateau reveals that the most influential meteorological factors for hail forecasting in Qinghai Province include thermodynamic conditions （vertically integrated temperature p54.162， vertically integrated thermal energy p60.162， and 2-meter dew point temperature d2m）， characteristic height layer conditions （100 hPa temperature t100， 400 hPa temperature t400， and 20 hPa geopotential height z20）， and dynamic conditions （500 hPa zonal wind component u500， 200 hPa meridional wind component v200， and 200 hPa zonal wind component u200）.Kernel density estimation analysis indicates that most feature variables exhibit limited separability， suggesting that no single factor alone can determine the occurrence of hail events.A case study demonstrates that the LightGBM-based hail forecasting model exhibits strong spatial forecasting capabilities.Analysis of the 24-hour evolution of key meteorological variables preceding a large-scale hail event at the Chaka station identifies several crucial atmospheric indicators： （1） significant fluctuations in vertically integrated temperature （p54.162）， indicating intense convective activity； （2） persistently high 2-meter dew point temperature （d2m）， reflecting abundant near-surface moisture； （3） strong 500 hPa zonal wind speed （u500）， suggesting enhanced mid-level atmospheric dynamics； and （4） low 100 hPa temperature （t100）， capturing upper-atmosphere characteristics.The coordinated evolution of these atmospheric variables not only reveals key stages in the development of severe convective weather systems but also provides a scientific foundation for improving hail potential forecasting methods in Qinghai Province.

The eastern and northeastern regions of Yunnan province are severe ice-coating areas.This study conducts a diagnostic analysis of a one-week ice-coating event in Zhaotong of Yunnan province during December 2023， utilizing manual observations， power grid sensor data， and ERA5 reanalysis data.The results indicate that the ice-coating area is predominantly located on the northern windward slope of the plateau， with the maximum ice thickness observed at elevations between 1500 and 2000 meters.The types of ice-coating are complex and varied， closely related to the timing of cold air influence.Specifically， the high-temperature and high-humidity type （GG type， t>0 ℃， RH≥75%） occurs during the initial stage of cold air influence.When the cold and warm air masses confront each other and a stationary front is maintained， the low-temperature and high-humidity type （DG type， t≤0 ℃， RH≥75%） occurs most frequently.As the cold air influence near its end， the low-temperature and low-humidity type （DD type， t≤0 ℃， RH<75%） becomes dominated.The high-temperature and low-humidity type（GD type， t>0 ℃， RH<75%）is the least frequent.During the ice formation period， the 500 hPa level is characterized by westerly airflow or short-wave troughs.At the 700 hPa level， southwesterly winds are observed， occasionally reaching the intensity of southwesterly jets， which transport abundant moisture.At 700 hPa， a wind shear line sometimes forms， coinciding with the surface stationary front， providing favorable dynamic uplift conditions for precipitation during icing events.The GG and DG types of icing occur under conditions of abundant moisture and strong uplift.Both of them are mixed icing associated with freezing rain （drizzle）， sleet， snow or fog.In contrast， the DD and GD types occur during the late stage of cold air retreat， characterized by weaker humidity and dynamic conditions.The DD type may be associated with freezing fog and drizzle under the influence of surface radiative cooling at night.In contrast， the formation of GD type is more likely related to fog or weak precipitation induced by local surface thermal or micro-terrain uplift.

Relative to individual extreme weather and climate events， Compound Wind and Precipitation Extremes （CWPE）， which result from extreme winds and extreme precipitation， have a profound impact on the economy and daily life in coastal areas.In this study， we utilized the simulations from 17 models within the Coupled Model Intercomparison Project Phase 6 （CMIP6） for the period 1961-2000 of CWPE in the eastern coastal region of China as a training set， and established a Deep Learning （DL） model employing a multi-layer perceptron neural network.By constructing a loss function suitable for CWPE and optimizing the model accordingly， we have developed a DL model aimed at reducing the simulation bias and uncertainty of CMIP6 models for CWPE.The research results indicate that the majority of CMIP6 models possess a relatively good simulation capability for CWPE in the eastern coastal region of China， with the Multi-Model Ensemble Mean （MME-Mean） and the Multi-Model Ensemble Median （MME-Median） demonstrating better performance in assessments compared to individual models.The DL model constructed with the Mean Squared Error （MSE） function as the loss function performs worse in terms of Taylor Skill Score （TS） and Root Mean Squared Error （RMSE） compared to the statistical results of the Multi-Model Ensemble.Incorporating the Ratio of the Standard Deviation （RSD） from climate evaluation metrics and an underestimation constraint function into the MSE loss function can significantly enhance the performance of the DL model in terms of TS and RMSE.Therefore， the DL model trained with a weighted loss function constructed from MSE， RSD， and underestimation constraint function is defined as DL-MRM， while the DL model trained solely with MSE as the loss function is defined as DL-MSE.By comparing and analyzing the performance of the two DL models in simulating CWPE over the eastern coastal region of China from 2001 to 2014， as well as the performance of DL-MRM relative to multi-model ensemble methods， we conclude： （1） Both DL models exhibit underestimation in their simulation results， but the bias of the DL-MRM is lower than that of the DL-MSE， being closer to the observations.Specifically， in the study area， the relative bias of the DL-MRM is lower than that of the DL-MSE by about 63%， and the average relative bias is reduced by approximately 20%.（2） The DL-MRM has a lower overall bias compared to the MME-Mean and MME-Median， with simulation results that are closer to the observations.In the study area， the DL-MRM has a lower relative bias in 67% and 62% of the area compared to the MME-Mean and MME-Median， respectively， and the average relative bias is reduced by approximately 10% and 20%， respectively.Overall， by integrating the RSD and underestimation constraint functions to construct a weighted loss function for model optimization， a DL model suitable for improving the simulation of CWPE by CMIP6 models was established.This indicates that the combination of deep learning methods can more effectively reduce the biases in CMIP6 model simulations of CWPE compared to traditional multi-model ensemble methods.

The low-temperature rain-snow freezing weather always has multiple precipitation phases， which greatly increases the difficulty of weather forecasting.Based on data from C-band dual-polarization radar and precipitation particle spectrum in Honghe， Yunnan， the characteristics of atmospheric circulation， the vertical structure of cloud microphysical structure and the spectral distribution of precipitation particles for a freezing weather event occurred on February 22， 2022 are investigated.The main results are summarized as following： （1） the strengthen and westward of the Kunming quasi-stationary front and its combination with the southern westerly trough of Tibet Plateau should be responsible for the freezing weather event.The alternative changes of the front and cold-warm atmospheric stratification directly induced the rain-snow phase transition on the ground.（2） when precipitation occurred， at -15 ~ -10 ℃ layer， there were sudden increases in Z _DR and K _DP， and a decrease in CC， showing a significant growth of dendritic ice crystals.There was also obvious bright band property at the melting layer， indicating an increase in melting process.（3） There were apparent differences in spectral distributions for different precipitation-phase particles.The formation of large snowflakes mainly depends on frost and collision growth of small snowflakes.The average particle number concentration of snowflakes reached the highest during snowfall， and corresponded the falling velocity of 1.1 m·s^-1， which was smaller than that of rain or sleet （2.2 m·s^-1）.The average spectra display a single-peak mode， with a spectral width of 0.312~7.5 mm for sleet and snow， which was wider than that of rain （0.312~5.5 mm）.When sleet occurred， the particle size had a turning point at 5.5 mm， as it turned into snow， the concentration of large-sized particles increased significantly.The results of this study can provide an important reference for the nowcasting of precipitation phase in winter.

Unstable energy， strong horizontal environmental wind shear and 0 °C layer height are important factors affecting surface hailfall.To explore the differences in environmental stratification of hail clouds between northern and southern China， this study compares the environmental characteristics before two hail events in Xunyi， Shaanxi， and northwestern Fujian.The three-dimensional hail cloud numerical model was used to simulate the development characteristics of these two events under different disturbance temperatures through sensitivity experiments， while also discussing the impact of strong wind shear on the Fujian hail case.The results show that the Shaanxi case has a large Convective Available Potential Energy （CAPE） but very weak vertical wind shear， whereas the Fujian case exhibits a relatively small CAPE but strong vertical wind shear.The height of the 0 °C layer in both hail clouds is relatively consistent.The atmospheric water vapor content during the formation of the Shaanxi hail cloud is lower， the cloud base height is higher， and the efficiency of CAPE transforming into updraft kinetic energy is higher.In contrast， the southern hail cloud has higher lower atmospheric humidity and a lower cloud base height， making it easier to release unstable energy.As the intensity of convection increases， the peak values of vertical wind speed become greater and occur earlier， the content of hail increases and its duration is prolonged， the intensity of solid-liquid precipitation intensifies， and the precipitation volume intensifies.The strong wind shear in the Fujian case weakens the convective intensity and shortens its lifespan， the weaker the convection is， the more obvious the effect is.Additionally， strong wind shear weakens solid-liquid precipitation.When there is no wind shear， hail exhibits larger specific water content and number concentration， wider range of existence， longer time and earlier occurrence time， but the hail range is smaller.

To enhance the prediction accuracy of the 2-meter maximum temperature in complex terrain areas， this study developed a gradient modeling approach based on the LightGBM （Light Gradient Boosting Machine， LGB） algorithm， applied to the Sichuan Basin and its surrounding regions.By selecting and and analyzing multiple meteorological and topographic factors， an optimized model was constructed.The results demonstrate that： （1） From January to June 2024， the LightGBM model reduced the mean absolute error by 2.48 ℃ and improved the forecast accuracy by 36.97% compared to EC model.Among them， the improvement effect of the west Sichuan Plateau and Panxi area was the most significant， the accuracy rate increased by 67.2% and 57.5%， respectively.（2） Compared with the existing objective forecast products SPCO and SCMOC， the LightGBM model improved prediction accuracy by 5.1% and 10.3%， respectively.Particularly in the Panxi area and the Sichuan Basin， the accuracy at individual stations improved by up to 17.6% and 23.4%， respectively.（3）The LightGBM model reduced the mean absolute error by 2.05~2.78 ℃， and increased the accuracy by 31.1%~41.0%， with the most notable enhancement occurring in April.（4）The LightGBM model exhibits strong scalability.Future work could further improve temperature prediction across Sichuan Province and other regions by incorporating time-lag effects， spatial neighborhood characteristics， and combining zoning modeling and multi-model integration.

This study aims to enhance the accuracy of urban grid temperature forecasts by integrating multi-scale features from both upper-air and surface levels.Utilizing forecast data from the European Centre for Medium-Range Weather Forecasts （ECMWF） model and hourly temperature grid observation data from the China Meteorological Administration's High-Resolution Land Data Assimilation System （HRCLDAS） for the Changde region from April to September 2021 -2024， a high-resolution Multi-Scale U-Net （MU-NET） model was employed.Three sets of experiments were designed to develop a deep learning model capable of predicting hourly temperatures in Changde City for the next 24 hours.The experimental results demonstrate that the MU-NET model， which integrates surface and upper-air features， exhibits the best correction performance within the study area.The mean absolute error （MAE） and root mean square error （RMSE） of the MU-NET model were reduced by 22% and 25%， respectively， compared to the ECMWF model.Additionally， the MU-NET model achieved the lowest MAE values in diurnal variations， particularly excelling in the prediction of daily maximum temperatures， with an average reduction of over 0.4 ℃.On a spatial scale， the MU-NET model showed a 60%~80% improvement in forecast skill scores over the ECMWF model in high-altitude regions， with the most significant improvements observed in plain and Dongting Lake areas.During two critical weather events in 2024， the MU-NET model demonstrated stable forecast performance due to its integration of surface and upper-air features， particularly in handling complex weather phenomena.The findings of this study provide new insights for improving the accuracy of temperature forecasts and offer valuable references for practical meteorological forecasting.

Using the AWOS （Automated Weather Observing System） automatic observation data at Maotai Airport from 2017 to 2023， ERA5 （ECMWF Reanalysis v5） reanalysis data， and the ETOPO2v2 topography data provided by the National Oceanic and Atmospheric Administration （NOAA） of the United States.Using methods such as weather analysis， statistical analysis， and diagnostic analysis， this study investigates the characteristics and influencing factors of low visibility weather at airports， and employs the HYSPLIT backward trajectory model to trace the sources of water vapor in low visibility weather.The results show that Maotai Airport has higher humidity throughout the year， with obvious characteristics such as low temperature， high humidity， and low wind speed in autumn and winter， which is more conducive to the formation and maintenance of low visibility weather.Low visibility weather occurs at the airport every month of the year， with the number of low visibility days being higher than average from October to February of the following year and in late spring and early summer.The frequency of low visibility weather gradually increases from dusk throughout the day.In spring， the frequency reaches its highest point from 06:00 （Beijing Time， same as after） to 08:00 in the morning， in summer and autumn from 05:00 to 07:00， and in winter from 03:00 to 05:00.In addition to low wind speeds， the main factors affecting low visibility weather include terrain， circulation background， inversion， temperature advection， and water vapor conditions.When low visibility weather occurs， the prevailing wind direction at the airport is WNW， which is closely related to the terrain of the Chishui River Valley and the activity of cold air.In spring， autumn and winter， the upper air level above the airport is mainly affected by the westerly airflow， and the low-level airflow from the north and the south converge over the Yunnan-Guizhou Plateau， resulting in the Kunming quasi-stationary front.Frontal inversion is beneficial for water vapor to accumulate near the ground and reach saturation.Low visibility weather at the airport typically appears in the cold zone behind the frontal zone.In summer， the temperature inversion over airports is mostly radiation inversion， and low visibility weather is mainly radiation fog.The HYSPLIT water vapor tracking results show that there are two main sources of water vapor at an altitude of 1500 meters during low visibility weather， namely the Bay of Bengal on the south side of the Qinghai-Tibet Plateau and the Chishui River Valley in the northwest of the airport； the water vapor at an altitude of 3000 meters mainly originates from the Bay of Bengal.

The atmospheric boundary layer height （ABLH） is one of the most important parameters in the study of atmospheric environment， weather and climate.With the development of ground-based remote sensing technology， continuous monitoring of ABLH has become possible.However， the ABLH derived based on ground-based remote sensing depends on the inversion method used and is affected by complex atmospheric conditions.In this study， we use the data from ceilometer and rain gauge， weather records from 27^th August 2020 to 1^st August 2023 and radiosonding records obtained during the 2023 summer extensive observation period at the Pingliang Land Surface Process and Severe Weather Research Station， Chinese Academy of Sciences.The effectiveness of several commonly used algorithms for inverting ABLH based on backscatter profiles from ceilometer are evaluated by comparing with the ABLH identified by potential temperature profiles.A hybrid algorithm that employs different backscatter gradient inversion methods for daytime［08:00（Beijing time， same as after） -19:00］ and nighttime （from 20:00 to 07:00 the next day） is proposed with constrained retrieval heights tailored for the study area.The results reveal notable differences in the inversion results among various algorithms， including the maximum negative gradient method， the three-major negative gradient evaluation method， the percentage fluctuation method， the inflection point method， and the Flamant method.Specifically， the ABLHs derived by the Flamant method， the three-major negative gradient evaluation method， and the maximum negative gradient method correlated well with that determined by the potential temperature profile and give lower mean absolute deviations.In contrast， the ABLHs derived by the inflection point method and the percentage fluctuation method give large absolute deviations.Appropriate smoothing of the backscatter profiles， combined with the hybrid algorithm， significantly improved the accuracy of the derived ABLH.Among the investigated methods， the SG 25/25 smoothing scheme combined with the Flamant and maximum negative gradient hybrid algorithm yielded the best results， achieving a correlation of 0.56 with the ABLH determined by the potential temperature profile and an average absolute deviation of approximately 406 m.The correlation between the ABLHs derived from the hybrid algorithm and that from the ceilometer’s internal algorithm is 0.64.The hybrid algorithm can effectively capture the diurnal variation of ABLH.The proposed hybrid algorithm can be used to obtain continuous， high-resolution ABLH information， serving as a valuable supplementary method for obtaining fundamental data on ABLH and related parameters.

To investigate the evolutionary patterns of atmospheric photochemical parameters and the influence mechanisms of aerosols in Beijing， this study analyzed the spatiotemporal characteristics of nitrogen dioxide photolysis rate J（NO₂） and explored the impacts of aerosols on J（NO₂） based on near-surface observations of J（NO₂） and ultraviolet （UV） radiation in Beijing from September 2018 to August 2019， combined with radiative transfer model TUV simulations and aerosol optical parameter analysis.Additionally， this study reconstructed a long-term photolysis rate dataset from 2013 to 2023 to further reveal the correlation between long-term trends and aerosol characteristics.The results show that the diurnal variation of J（NO₂） exhibits a typical unimodal pattern， with the peak generally occurring during the noon period ［12:00 -13:00，（Local Time， the same as after）］， directly influenced by changes in solar zenith angle.The daytime maxima of J（NO₂） in summer are 1.9 times those in winter， measuring 5.65×10^-3 s^-1 and 2.95×10^-3 s^-1， respectively， indicating significantly enhanced photolysis rates in summer due to higher solar radiation intensity.Seasonal variations follow the order： summer （3.77×10^-3 s^-1） > spring （3.51×10^-3 s^-1） > autumn （2.97×10^-3 s^-1） > winter （2.25×10^-3 s^-1）， driven by seasonal changes in solar radiation intensity and the combined effects of increased summer precipitation.Furthermore， an estimation model for J（NO₂） constructed using the UV clear-sky index （K_UV） and the cosine of solar zenith angle demonstrated a linear correlation coefficient of 0.99 between calculated and observed values， with a mean relative error of 8.8% and a root mean square error of 0.00036.These results validate the model’s high applicability under complex atmospheric conditions， providing an effective predictive tool.The reconstructed long-term dataset reveals a significant upward trend in J（NO₂） in Beijing from 2013 to 2023， with an annual increase rate of 2.73%.The 2023 annual mean （4.20×10^-3 s^-1） increased by 31.3% compared to 2013 （3.20×10^-3 s^-1）， closely linked to the continuous decline in PM_2.5 concentrations （annual reduction rate： 5.51%） and aerosol optical depth （AOD， annual reduction rate： 5.74%）.These changes suggest that reduced particulate matter due to air quality improvement measures has diminished UV radiation attenuation， indirectly promoting photolysis rate enhancement.The study found that J（NO₂） significantly decreases under polluted conditions.When PM_2.5 concentrations exceed 75 μg·m^-³， the J（NO₂） maximum decreases by 22.8% compared to clean conditions， indicating that PM_2.5 concentration is a critical factor influencing photolysis rates.Analysis of aerosol optical properties revealed that J（NO₂） is negatively correlated with AOD and Ångström exponent （AE）， but positively correlated with single-scattering albedo （SSA）.Sensitivity tests demonstrated that increasing AOD from 0.5 to 2.5 reduces the daily maximum J（NO₂） by 45.6%， with a noon attenuation rate of 49.8% at AOD = 2.5.Conversely， increasing SSA from 0.2 to 1.0 enhances aerosol scattering capacity， raising the daily maximum J（NO₂） by 43%.Increasing AE from 0.5 to 2.0 results in only a 3.0% reduction in J（NO₂） maxima， indicating that AE has a weaker influence compared to AOD and SSA.The hierarchy of aerosol impacts on photochemical processes is ranked as AOD > SSA > PM_2.5 > AE， with AOD exerting the most significant influence on J（NO₂） variations， highlighting the dominant role of aerosol extinction in suppressing photochemical processes.Additionally， validation of the TUV model confirmed its reliability in simulating J（NO₂） spatiotemporal variations， with a correlation coefficient of 0.93 between simulated and observed values.This validation provides crucial methodological support for quantifying aerosol-radiation-photochemistry coupling mechanisms.

Radio Occultation data are of great value in numerical weather forecasting and climate monitoring.This study evaluates the deviation characteristics of occultation data from the FY-3E/GNOS and COSMIC-2 satellites with ERA5 reanalysis data.The results indicate that FY-3E/GNOS and COSMIC-2 exhibit high consistency with ERA5 in the altitude range of 5 km to 30 km， with refractivity deviations less than 1% and bending angle deviation is stable within ±0.4%.However， in the lower atmosphere （below 5 km）， COSMIC-2 presents a maximum negative bias of -1.2% at altitudes below 4 km， while FY-3E/GNOS demonstrates a systematic positive bias below 5 km in the bending angle， and the refractivity shows negative deviations below 2 km in the low-latitude region and positive deviations from 2 to 4 km.Research findings indicate that the bending angle profiles of FY-3E/GNOS are smoother than those of COSMIC-2， especially in the lower troposphere.

Vegetation competition and coexistence are fundamental ecological processes in Dynamic Global Vegetation Models （DGVMs）， exerting a direct influence on plant community structure and function， and ultimately shaping global vegetation distribution patterns.Clarifying the differences in how these mechanisms are represented is critical for reducing uncertainty in ecosystem modeling and improving prediction accuracy.This review systematically examines the representations of vegetation competition and coexistence in DGVMs， classifying them into implicit and explicit approaches.Drawing on both model-based analyses and a comprehensive literature survey， the study explore their potential impacts on simulations and compare the advantages and disadvantages of each approach.Three key findings emerge： （1） the primary distinction between implicit and explicit representations lies in assumptions about vegetation structure.Implicit approaches typically assume a homogeneous canopy and treat plant functional types （PFTs） as independent and non-overlapping units.This simplification limits the ability to simulate direct resource competition， relying instead on heuristic strategies such as competitive ranking to adjust PFT cover fractions.Although computationally efficient， this approach often neglects the complexity of plant interactions.In contrast， explicit representations characterize three-dimensional vegetation structure and its environmental interactions， enabling direct simulation of fine-scale competition processes and producing more dynamic outcomes that reflect continuous resource partitioning rather than fixed hierarchies.（2） The representation of competition significantly affects community-level simulations.Three comparative experiments— models with versus without competition， implicit versus explicit schemes， and sensitivity tests of implicit parameters— demonstrate that different formulations can cause substantial differences in spatial distribution and carbon biomass， with discrepancies reaching up to 48.6%.In implicit models based on the Lotka-Volterra framework， the form of the resource-density relationship plays a pivotal role in determining coexistence.Nonlinear formulations better refelct plant "clustering" effects， promoting multispecies coexistence， while linear assumptions tend to eliminate non-dominant PFTs， resulting in unrealistic monocultures.（3） This review also highlights the trade-offs between modeling approaches.Implicit schemes are widely used in large-scale Earth system models due to their simplicity and computational efficiency， but may introduce systematic biases， such as overestimating the productivity of shaded plants.Explicit approaches offer a more mechanistic representation of plant competition and individual-level clustering， aligned more closely with ecological theory， yet are limited by higher computational demands and uncertainty in parameterization.Finally， the review concludes by outlining several future directions for DGVM development， including enhancing model intercomparison projects， constructing benchmark datasets on vegetation structure and distribution， and gradually integrating explicit competition schemes under controlled uncertainty.These efforts could deepen our understanding of plant community dynamics and carbon cycle responses under climate change， thereby supporting more robust scientific foundations for climate change policy.