Ensemble prediction has been an important tool for weather forecasting, sub-seasonal to seasonal prediction, seasonal prediction, interannual prediction and even simulation of climate change, which has garnered widespread attention in the field of meteorology. This paper introduces the ensemble prediction scheme of China Meteorological Administration Climate Prediction System version 3 (CMA-CPSv3). In this scheme, we adopt the approach of combining stochastic perturbations of physical process tendencies in the atmosphere and the air-sea flux with time-lagged initial value perturbation. Based upon the version 2 of High-Resolution Beijing Climate Centre Climate System Model (BCC-CSM2-HR), we have developed a multi-layer random perturbation ensemble prediction system with relatively good ensemble sample dispersion, stability, and reliability. Results of evaluation for hindcasts over the past 20 years show that this ensemble prediction system significantly improves the prediction of precipitation and 2 m air temperature over China, as well as the El Niu00F1o-Southern Oscillation (ENSO), Indian Ocean Dipole (IOD) and Asian Monsoon. In particular, the random perturbation of air-sea flux shows a positive effect on improving the prediction skills of ENSO and Southeast Asian Monsoon (SEAM) and Western North Pacific Summer Monsoon (WNPSM) indices. This study offers useful insights for further characterizing uncertainty in other component models of the climate system.
This study explores the direct assimilation of Level-1 observations from the Hyperspectral Infrared Atmospheric Sounder (HIRAS-u2161) aboard the Fengyun-3E (FY-3E) satellite into a global real-time atmospheric analysis system. A dedicated operational processing chain for HIRAS-u2161 data has been developed within the Gridpoint Statistical Interpolation (GSI) assimilation framework, incorporating data preprocessing, quality control, bias correction, cloud detection and variational assimilation. By combining the CrIS-FSR channel selection method with Jacobian sensitivity analysis of HIRAS-u2161, a subset of 28 temperature-sensitive channels is identified from the long-wave infrared band. Assimilation experiments demonstrate that this channel subset significantly improves temperature assimilation and effectively suppresses errors in humidity. This approach contributes to better temperature analyses in the mid-to-upper troposphere while preserving positive effects in the lower troposphere. A maximum reduction of 2% in the Root Mean Square Error (RMSE) of temperature analyses is observed over the northern Hemisphere. This study marks the first successful operational assimilation of FY-3E/HIRAS-u2161 data into the global real-time atmospheric analysis system, providing reliable technical support for the operational assimilation of Fengyun satellite hyperspectral infrared data.
To reveal the environmental conditions, triggering mechanisms, and predictability of extreme short-term heavy rainfall in urban areas under weak synoptic-scale forcing, a comprehensive analysis of an abrupt heavy precipitation event that occurred in the main urban area of Zhengzhou on 22 July 2024 is conducted. Data from dense automatic weather stations, ERA5 reanalysis data, short-term forecast products from various numerical models, and multi-source novel observational data including FY-4B geostationary satellite imagery, dual-polarization radar, wind profile radar, cloud radar, and GNSS/PWV are used. The results indicate that this event occurred at the edge of the western Pacific subtropical high, where the synoptic forcing was weak. Westerlies prevailed at 500 hPa, while warm and moist southerlies prevailed in the lower levels. The surface was controlled by a warm low-pressure system. Prior to the heavy rainfall, the environment in Zhengzhou was characterized by extreme high-temperature and high-humidity, with the Convective Available Potential Energy (CAPE) exceeding 5000 J/kg and the Precipitable Water Vapor (PWV) above 80 mm, indicating persistent enhancement of atmospheric instability. FY-4B infrared imagery revealed that the heavy precipitation was caused by the development and southward movement of a Mesoscale Convective System (MCS) originating from northern Henan. During its mature stage, the MCS exhibited a significant expansion of cold cloud, with the heavy rainfall concentrated in the region of large brightness temperature gradient at the leading edge of the MCS. Further analysis suggests the rainfall was triggered by boundary-layer processes in a phased manner, i.e., the urban heat island effect initially induced a meso-u03B3-scale surface low pressure over the main urban area, triggering the initial convection. The outflow from the cold pool of this convective system converged with the ambient warm, moist flow, triggering new convections accompanied by persistent small-scale vortices. The thunderstorm high pressure from northern Henan moved southward, forcing the rise of the leading-edge warm, moist air and triggering a widespread convective outbreak. Short-term forecasts from several numerical models failed to predict this event, reflecting their limitations in simulating the evolution of meso- and micro-scale systems under weak synoptic-scale forcing. In contrast, high spatiotemporal resolution observations demonstrate significant value for nowcasting. Wind profile radar detected the establishment of a westerly jet stream at 4u20145 km altitude 2u20143 h before the precipitation, and the 0u20146 km vertical wind shear increased from weak to moderate, favoring convective organization in the high-CAPE environment. GNSS/PWV data showed a sharp increase of over 4 mm about 1 h before the rainfall, indicating rapid low-level moisture convergence and uplift. Cloud radar observations captured a sharp decrease in cloud base height, signaling the critical phase of microphysical processes transitioning into heavy precipitation formation. In conclusion, under weak synoptic-scale forcing, the multi-stage triggering mechanism that involves urban heat island, cold pool outflow, and thunderstorm high pressure was the key to the occurrence of this extreme short-duration heavy rainfall. The continuous evolutionary signals revealed by multi-source novel observations provide crucial information for improving the nowcasting capability of such kinds of abrupt local heavy precipitation events.
Northern China is characterized by steep terrain, complex circulations and strong land-atmosphere interaction, which make it highly susceptible to frequent droughts. In recent years, extreme drought events have increased in this region under the background of global warming, posing a threat to socioeconomic sustainable development. To promote the prediction of extreme drought events, it is imperative to improve the performance of regional numerical models for the simulation of drought factors such as precipitation and temperature. Therefore, RegCM5.0 is localized and improved in this study to develop a drought prediction system. The improvements include surface type data update, thermal conductivity parameterization modification and cumulus convection scheme optimization. Results show that precipitation bias is reduced by 50% and 2 m air temperature bias is reduced by 1u2103. The correlation coefficients of precipitation and 2 m air temperature with observations increase from 0.61 to 0.72 (u03B1=0.001) and from 0.83 to 0.88 (u03B1=0.001), respectively. The correlation coefficients of simulated drought indexes such as SPI (Standardized Precipitation Index) and SPEI (Standardized Precipitation Evapotranspiration Index) with those from observations can reach above 0.50 (u03B1=0.05). A case study shows that the prediction system can accurately forecast the extreme drought event in the summer of 2020 and those extreme heat and drought events in 2024.
Based on the monthly Extended Reconstructed Sea Surface Temperature dataset (ERSST v5) provided by the National Oceanic and Atmospheric Administration (NOAA), this study investigates how the longitudinal position of La Niu00F1a affects summer precipitation in Southeast China during its developing phase from 1950 to 2023. The underlying physical mechanisms are also explored. The results show that an eastward shift of the La Niu00F1a cold center tends to induce a meridional dipole precipitation anomaly pattern over Southeast China, characterized by increased rainfall over the Yangtze River Basin and decreased rainfall over South China. In contrast, no significant precipitation anomalies are observed in Southeast China when La Niu00F1a shifts westward. Further analysis reveals that the eastward-shifted La Niu00F1a events are often accompanied by positive Sea Surface Temperature Anomalies (SSTA) in the equatorial western Pacific, which strengthen the zonal SSTA gradient and enhance equatorial easterly wind anomalies. These changes are favorable for the development of an anomalous anticyclonic circulation near the South China Sea, which enhances moisture transport to the Yangtze River Basin while suppressing convection over South China via its subsiding branch. Conversely, westward-shifted La Niu00F1a events are featured by a weaker zonal SSTA gradient and negligible atmospheric circulation responses, thus exerting little influence on precipitation in Southeast China. This study highlights the crucial role of La Niu00F1a's longitudinal position in modulating summer rainfall patterns over China and provides new physical insights into the ENSO-precipitation relationship. These findings offer important implications for improving flood-season precipitation prediction.
Based on ERA5 reanalysis data and monthly atmospheric circulation indices provided by the Climate Prediction Center (CPC), this study employs composite analysis and dynamical diagnosis methods to comparatively analyze the impacts of Eastern Pacific (EP) and Central Pacific (CP) El Niu00F1o events on winter near-surface temperature in high-latitude regions and their underlying physical mechanisms. Results show that the two types of El Niu00F1o events influence high-latitude near-surface temperature by triggering different teleconnection wave trains: EP El Niu00F1o primarily induces warming over Canada through a positive-phase Pacific-North American (PNA) teleconnection but has weaker effects on the polar region. CP El Niu00F1o, in contrast, triggers a negative-phase North Atlantic Oscillation (NAO), leading to warming over Greenland and cooling in the Arctic. In both cases, near-surface temperature anomalies are predominantly driven by temperature advection processes. Further analysis reveals distinct wave propagation mechanisms. During EP El Niu00F1o events, robust waves triggered in the central-eastern equatorial Pacific propagate zonally to the North Pacific, forming a classic PNA wave train along the westerly waveguide. CP El Niu00F1o events generate weaker waves in the eastern tropical Pacific that cannot penetrate the North Pacific but instead propagate meridionally into the Atlantic, where they amplify under the influence of negative Potential Vorticity (PV) gradients, forming a negative-phase NAO-like response. Notably, CP El Niu00F1o is accompanied by pronounced Sea Surface Temperature (SST) anomalies in the subtropical Atlantic. These SST anomalies reinforce the negative NAO phase via transient eddy vorticity feedback, sustaining the anomalous circulation. This study enriches the understanding of how El Niu00F1o diversity modulates high-latitude climate variability and provides a theoretical framework for improving seasonal-to-interannual climate predictions in polar regions.
Effective representation of model uncertainty is crucial for improving the forecast skill of convection-permitting ensemble prediction system. The Stochastic Perturbed Parameterization Tendencies (SPPT) scheme is one of the primary approaches used to represent model uncertainty, and its effect is controlled by three parameters: Perturbation magnitude, temporal correlation scale, and horizontal perturbation scale. There have been few studies on the optimization of these three parameters for the operational 3 km CMA-REPS (Regional Ensemble Prediction System of China Meteorological Administration) v4.0. Based on CMA-REPS, this study selects 13 heavy rainfall cases in North China in 2024 to conduct SPPT parameter sensitivity experiments. The forecast skill for upper-air and surface variables, precipitation, and perturbation energy growth are analyzed. First, using a smaller magnitude (with a standard deviation of 0.35) and dropping attenuation of the perturbations within the boundary layer most effectively enhances the forecast skill of the variables. Second, a 3 h temporal scale is conducive to improving the forecast skill within the initial 12 h, whereas a 6 h time scale performs better after 24 h of integration. A horizontal scale of 500 km yields the best overall performance. Compared with a 1000 km scale, it improves the spread and consistency for most variables. Further reducing the scale to 200 km can improve light and moderate rain forecasts within the initial 12 h but leads to a decline in forecast skill after 18 h. Third, spatiotemporal scales significantly influence the perturbation energy growth. The 3 h temporal scale promotes the perturbation energy growth across scales within the initial 12 h, while the 6 h scale is more favorable for perturbation growth after 18 h. The 500 km horizontal scale is most beneficial for the development of difference kinetic energy and difference latent energy. Although the 200 km horizontal scale can initially enhance low-level perturbation energy and promote smaller-scale perturbation growth during convectively active periods, it results in the minimal development of larger- and meso-scale components, as well as perturbation in the middle and late periods of integration. In conclusion, a 0.35 standard deviation with unattenuated boundary layer perturbations, a 6 h temporal scale, and a 500 km horizontal scale are recommended.
Using a nearly 1000-year pre-industrial control simulation from the CESM2 Earth System Model in the CMIP6 archive, we identify positive Indian Ocean Dipole (pIOD) events and categorize them into single-year and consecutive multi-year (two- and three-year) types based on their duration. Direct comparisons between different types are conducted to reveal their impacts on Antarctic spring sea ice and atmospheric circulation. Results show that the sea ice anomaly pattern triggered by pIOD events is strongly duration-dependent. Single-year events lead to a dipole pattern in the West Antarctic, whereas the pattern induced by multi-year events is characterized by eastward shift and meridional expansion of the anomaly centers. The lower tropospheric circulation anomalies, guided by Rossby wave trains, are the key mechanism underlying these differences. Sea ice anomalies are modulated by dynamic processes (Ekman transport and thermal advection) and are closely associated with thermodynamic processes (downwelling shortwave and longwave radiation). Compared with single-year events, multi-year events tend to generate more persistent and stronger Rossby wave trains, which maintain and intensify the pIOD impacts on Antarctic sea ice. Furthermore, quantitative sea ice mass budget analysis reveals that changes in sea ice mass are mainly attributed to dynamic processes, basal melting, lateral melting and snow-to-ice conversion.
To address the lack of the assimilation capability for Fengyun satellite microwave imager observations in the CMA-RA v1.5 development system and to enhance the application value of domestic satellite observations, this study processes L1 observations from the Fengyun-3D (FY-3D) MicroWave Radiation Imager (MWRI), constructs an assimilation module based on the Gridpoint Statistical Interpolation (GSI) system, and performs a one-month batch experiment and evaluation using the Hybrid four-dimensional ensemble-variational (Hybrid-4DEnVar) assimilation method. The results show that the adopted quality control and bias correction schemes are reasonable and reliable, which can effectively screen high-quality clear-sky over-ocean observations and correct systematic biases. Assimilation of MWRI observations improves the quality of specific humidity analysis, i.e., the Root Mean Square Error (RMSE) of 800 hPa specific humidity analysis is reduced by up to 1.05% (passing the significance test), and the dry bias near 800 hPa in the tropics and the wet bias below 700 hPa in the southern Hemisphere are corrected. The 0u201472 h specific humidity forecasts show significant improvements in the lower and middle troposphere (950u2014600 hPa), with particularly notable improvements occurring near 800 hPa during the first 24 h of forecast (the maximum error reduction reaches 8.74 mg/kg). The RMSEs of 3 h, 6 h, and 9 h forecasts are reduced by up to 1.4%, and the anomaly correlation coefficients of 700 hPa specific humidity forecasts are increased by 0.001u20140.01. The forecasts of wind and temperature fields exhibit a neutral-to-positive effect, and the geopotential height forecast is improved at certain time steps of medium-range forecasts. The core contribution of FY-3D MWRI data assimilation focuses on the improvement of humidity analysis and forecasts, while its overall impact on other meteorological elements is relatively neutral.
In the research on direct assimilation of land surface brightness temperature data, conventional Radiative Transfer Models (RTM) used as observation operators often require extensive auxiliary information. However, the complexity and spatiotemporal variability of land surface parameters can easily lead to significant errors in auxiliary data, which severely degrade the accuracy of brightness temperature simulations and affect assimilation performance. To enhance the effect of direct assimilation of observations from the MicroWave Radiation Imager (MWRI) aboard the FY-3D satellite into land surface models, this study develops a data-driven observation operator based on the Multi-Layer Perceptron (MLP) architecture, which obviates the explicit parameterization of surface emissivity. This operator is grounded in adjoint sensitivity analysis, which identifies primary sources of simulation errors in conventional RTM. Given the complex spatiotemporal variability of land surface radiation, a modeling strategy with independent models for different surface types, day/night conditions, and seasons is adopted to reduce data dimensionality and further enhances the accuracy of the observation assimilation operator. Evaluation results demonstrate that the MLP operator significantly outperforms the Radiative Transfer for TOVS (RTTOV) model in brightness temperature simulation across most surface types. The most remarkable improvement is achieved over barren areas in summer: The Mean Absolute Error (MAE) decreases from 7.297 to 4.021 K (a reduction of 44.9%), and the Root Mean Square Error (RMSE) declines from 9.029 K to 5.721 K (a reduction of 36.6%). Additionally, MAE improvements of 38.6% and 37.3% are observed over grasslands and broadleaf forests, respectively. The MLP operator exhibits the most significant accuracy enhancement under daytime conditions, while it still maintains an advantage at night, the magnitude of error reduction is relatively smaller. The season-specific modeling approach ensures that the MLP operator achieves substantial improvements in brightness temperature simulation across all seasons, with the most notable gains in autumnu2014RMSE reductions of approximately 4.0 K are observed for various vegetation types. Quantitative analysis using the Shapley additive explanations (SHAP) method confirms that the MLP operator can effectively replicate the physical mechanisms of land surface radiation, highlighting its promising potential for practical applications.
During Juneu2014July 2023, North China (NC) experienced a record-breaking Extreme High Temperature (EHT) event with temperatures exceeding 40u2103 at many meteorological stations. This event consisted of three relatively distinct processes, occurring during 14u201418 June (P1), 21u201425 June (P2), and 30 Juneu20143 July (P3), respectively. Using hindcast data from the European Centre for Medium-Range Weather Forecasts (ECMWF) and the China Meteorological Administration (CMA) within the Subseasonal-to-Seasonal (S2S) Prediction Project and diagnostic analysis of the local temperature budget, surface energy budget, and large-scale atmospheric circulation, this study evaluates the performance of ECMWF and CMA dynamic models in forecasting the spatial distribution and intensity of the three EHT processes, and further reveals the sources of forecast errors and predictability. The results indicate that: (1) ECMWF and CMA models can predict the spatial distribution of Surface Air Temperature (SAT) anomalies over NC for P1, P2 and P3 with lead time of 10u201411, 12u201414, and 3u20146 d, respectively. However, both models underestimate the amplitude of SAT anomalies, especially during P3; (2) when the forecast lead time exceeds 15 d for P1 and P2, and 5 d for P3, both models fail to capture features of the mid- to high-latitude Rossby wave train over Eurasia, resulting in prediction biases in both the location and intensity of localized high-pressure anomalies over NC; (3) prediction biases of the localized high-pressure anomalies over NC further induce biases in the forecast of local physical processes and SAT anomalies. During P1 and P3, the models' underestimation of both shortwave and longwave radiation leads to local negative temperature tendency biases, with diabatic heating being the primary contributor. In contrast, during P2, the underestimation of adiabatic heating is primarily responsible for the cold bias in predicted SAT. This study highlights the mid- to high-latitude subseasonal Rossby wave train teleconnection as a critical source of predictability for subseasonal EHT in NC, suggesting that better representation of this wave pattern is a key to improve EHT prediction skills over NC.
Traditional ensemble forecasts are based on deterministic models, and their performance is directly affected by the forecast quality of the deterministic models they rely on. The China Meteorological Administration Global Forecast System (CMA-GFS) was upgraded from v3.3 to v4.0 in 2023, resulting in significant improvements in both its horizontal resolution and forecast performance. This upgrade also provides a crucial foundation for optimizing the CMA Global Ensemble Prediction System (CMA-GEPS) with the CMA-GFS as the forecast model. To investigate the impact of the forecast model update on CMA-GEPS forecasts in wintertime, the operational CMA-GEPS v1.3 is taken as the baseline in this paper, and two ensemble forecast experimental systems are constructed using CMA-GFS v3.3 and v4.0, respectively, with the only difference lying in the forecast model. Two groups of 31 d ensemble forecast comparative experiments for the winter season of 2024 are carried out, and analyses are implemented from three perspectives, i.e., perturbation characteristics, statistical scores, and a representative case study. Results show that after the forecast model is updated from CMA-GFS v3.3 to v4.0, the ability of ensemble perturbations from CMA-GEPS to capture the forecast errors is slightly improved, and the perturbations grow faster. Moreover, the ensemble spread (SPD) of most elements verified increases significantly. Except that the ensemble mean Root Mean Square Errors (RMSE) of upper-level temperature in the tropical region deteriorate, the RMSEs of other variables remain nearly unchanged or improved. Overall, the gap between SPD and the ensemble mean RMSE becomes narrow, suggesting an enhanced ensemble reliability. Apart from the upper-level temperature in the tropical region, both the Continuous Ranked Probability Score (CRPS) and Outlier primarily decline, indicating improved probabilistic forecast skills. Over China, the forecast skill of light rain is nearly unchanged, and forecasts of moderate rain and heavy rain are improved to some extent. In summary, the forecast model upgrade generally improves the wintertime forecast performance of CMA-GEPS, which is further confirmed through the analysis of a representative cooling weather process in the northeastern region of China in late November. However, the increased SPD introduced by the new forecast model exacerbates the over-dispersive problem of the geopotential height forecast from CMA-GEPS during the early forecast period. Therefore, even if the initial perturbation technology, model perturbation strategy, horizontal resolution, and ensemble size of the ensemble prediction system are kept unchanged, it is still necessary to optimize the perturbation parameters after updating the forecast model to achieve a comprehensive improvement of the CMA ensemble forecast skill. In the future, impacts of the forecast model upgrade on CMA-GEPS in other seasons will be further explored.
Melting Layers (ML) during winter precipitation events, characterized by low-altitude, high temporal variability and spatially heterogeneous melting, pose significant challenges to conventional dual-polarization radar Hydrometeor Classification Algorithms (HCA), resulting in substantially degraded performance in rain-snow transition regions. To address these challenges, we develop an improved HCA scheme using data from 115 dual-polarization radars and vertical profiles from 116 radiosonde stations across China (July 2023u2014July 2024). The scheme integrates three key components: Optimized spatiotemporal matching with radiosonde observations, real-time ML monitoring through Quasi-Vertical Profile (QVP) analysis, and three-dimensional ML identification using the Melting Layer Detection Algorithm (MLDA). These improvements enhance both spatiotemporal precision of ML detection and regional applicability of HCA, ultimately improving the discrimination between various hydrometeor types. The improved scheme effectively resolves issues related to rapid ML variability and spatial heterogeneity in winter, reducing the ML detection update interval from 6u201412 h to 6 min and achieving spatial resolution of 1 km in range, 0.1 km in altitude, and 1u00B0 in azimuth. Sensitivity experiments using wintertime datasets from seven radars around Nanjing demonstrate that the improved scheme can accurately identify the rain-snow boundary, increasing overall classification accuracy by 11.91% and mixed-phase precipitation accuracy by over 50%. Validation against Present Weather Sensor (PWS) observations confirms that the near-surface classification accuracy exceeds 77% within 100 km. Statistics on winter algorithm activation periods from 115 radars nationwide reveal that the winter-specific algorithm operates for 18%u201451% of the year at plain sites in Northeast, North, and Central China, as well as at high-altitude mountain sites, confirming the effectiveness of the improved scheme for wintertime dual-polarization radar hydrometeor classification.
Based on the ECMWF ERA5(European Centre for Medium-Range Weather Forecasts Reanalysis Version 5)atmospheric reanalysis dataset and daily precipitation data from the Climate Prediction Center(CPC)along with the extended reconstructed global monthly mean sea surface temperature(SST)from the National Oceanic and Atmospheric Administration(NOAA)during 1981-2024,two rare freezing rain events that occurred over Jianghuai region in early 2010 and 2024(hereafter Event 1 and Event 2)are analyzed and compared.The analysis is focused on differences in observed characteristics,atmospheric circulation patterns and temperature stratification associated with the two freezing rain events.The association of the two events with tropical ocean-atmosphere forcings and stratospheric polar vortex anomalies as well as their possible effects are further explored.Results are as follows:(1)The range and duration of Event 2 are respectively larger and longer than that of Event 1.In Event 1,freezing rain mainly occurred over Huaibei region during the daytime of 10 February 2010,whereas in Event 2,it occurred on 20-21 February 2024 and extended southward,reaching the Yangtze River Basin.These two freezing rain events occurred under similar circulation patterns,which are characterized by the convergence of cold air from north and warm moist air transported by southwesterlies in front of the trough and the periphery of the Western Pacific subtropical high.The stable and sustained transport of warm moist air is crucial for the establishment and maintenance of the melting layer and the inversion layer during both events.(2)The concurrent El Niño event in the tropical Pacific is favorable for the establishment of a vertical temperature configuration with cooling in the middle levels and warming in the lower levels over the Jianghuai region.Meanwhile,the intensified western North Pacific anticyclone anomalies in the lower levels enhanced moisture transport,providing a large-scale background favorable for the two freezing rain events.(3)The deep convection associated with the active Madden-Julian Oscillation(MJO)over the Western Pacific triggered strong anomalous anticyclonic circulations over the Northwestern Pacific,providing abundant moisture transport to Jianghuai region for the occurrences of the two freezing rain events.This facilitated the establishment and maintenance of the melting layer and inversion layer in the lower levels.(4)In the early stages of the two freezing rain events,the sudden stratospheric warming(SSW)events resulted in the deformation and southward displacement of the stratospheric polar vortex to the Asia.This led to cold air outbreaks by modulating tropospheric circulations in the mid-high latitudes.The cold air strongly converged with warm and moist air over Jianghuai region.As a result,a stable and sustained inversion structure was established,which,accompanied by intense convergence of low-level horizontal wind fields and upward motion,facilitated the occurrence of the two freezing rain events.In general,during the two unexpectable freezing rain events over Jianghuai region,the joint effects of tropical ocean-atmosphere forcing and the deformation of the stratospheric polar vortex played a significant role in the occurrence of these events.Therefore,MJO and SSW events should be considered for short-term predictions of freezing rain processes in the Jianghuai region.
Focusing on the forecasting and early warning of heavy precipitation in the key areas along the Sichuan-Xizang Railway,a heavy precipitation classification forecast model has been constructed for various subregions based on high spatiotemporal resolution Fengyun-4(FY-4)satellite data and ERA5 reanalysis product for the summers of 2020-2024,combined with the Light Gradient Boosting Machine algorithm.The model's interpretability is analyzed using Shapley additive explanation(SHAP),and the distribution characteristics of key forecasting factors are analyzed.Results show that the model achieves a critical success index(CSI)of 0.41 for heavy precipitation forecast in the key regions with a probability of detection(POD)reaching 0.76,and demonstrates a strong forecasting capability.Regional model analysis indicates that the POD for heavy precipitation is 0.83 and the CSI is 0.53 in western Sichuan.The POD is 0.69 and the CSI is 0.33 in the southeastern region of the Qingzang Plateau,indicating significant regional differences in the forecasting.SHAP and statistical analysis show that heavy precipitation in the southeastern Qingzang Plateau is mainly dominated by satellite brightness temperature difference(BTD)factors(such as BTD6.25-7.1 and BTD13.5-10.7)that reflect variations in mid-and upper-level water vapor and cloud top height,while thermal instability parameters(such as CAPE(convective available potential energy),K index)and low-level vertical motion are the main indicators for heavy precipitation in western Sichuan.60 min prior to the occurrence of heavy precipitation,key satellite parameters and physical parameters already exhibited statistically significant differences.Satellite parameters begin to show notable evolutionary characteristics as early as 150 min prior to the precipitation,providing quantitative reference for early warning.Based on interpretable machine learning methods,it is possible to not only enhance the objective forecasting ability of heavy precipitation,but also deepen the understanding of regional heavy precipitation mechanisms.This study provides scientific support for disaster prevention and reduction along the key areas of the Sichuan-Xizang Railway.
To improve the analysis and forecast ability of severe thunderstorm gusts(speed≥24.5 m/s)under the background of Northeast China Cold Vortex(NCCV),it is necessary to study the distribution and environmental characteristcs of severe thunderstorm gusts under the background of NCCV.Based on 8 a observations collected at 17677 automatic weather stations,1717 severe thunderstorm gust events and 49803 ordinary gust events(17.2 m/s≤speed＜24.5 m/s)under the background of NCCV are identified.The thunderstorm gust events are classified into four quadrant groups according to their relative position to the center of the NCCV.Observational characteristics of severe thunderstorm gusts in different quadrants of the NCCV and comparisions of environmental conditions for thunderstorm gusts of different intensities as well as key environment parameters associated with severe thunderstorm gust occurrences in different quadrants are studied.The results show that severe thunderstorm gusts occur most frequently in the southeast quadrant,followed by the southwest,northeast and northwest quadrants.Severe thunderstorm gusts occur predominantly in plain areas.Heilongjiang is the province with the highest frequency of severe thunderstorms gusts under the background of NCCV.Severe thunderstorm gusts occur most frequently in the southwest quadrant in Henan,Jiangsu and Anhui,while they are most frequent in the southeast quadrant in other provinces.Most severe thunderstorm gusts occur in July and during the afternoon.Monthly distributions of severe thunderstorm gusts in different quadrants exhibit the characteristics of a main peak in the summer and a secondary peak in the spring.All severe thunderstorm gust events are analyzed by establishing a new coordinate system with the center of the NCCV as the origin and the radius of NCCV as the unit distance.Major results are as follows.(1)48.7%of severe thunderstorm gusts are concentrated in the bottom of NCCV with the azimuth angle of 135°—210° and within 1.2-2.5 times of the NCCV radius.This area is a high occurrence zone of severe thunderstorm gusts at the bottom of the NCCV.Severe thunderstorm gusts in Northeast China and North China mainly occur on the southeast side of the high occurrence area at the bottom of NCCV,while severe thunderstorm gusts in East China largely occur on the southwest side.Compared to ordinary thunderstorm gusts,especially in the southwestern quadrant,0-6 km wind vector difference,0-3 km storm helicity and average wind speed in the storm bearing layer of severe thunderstorm gusts are larger in spring and autumn,and the convective available potential energy and downdraft convective available potential energy of severe thunderstorm gusts are higher in summer.The above-mentioned physical prameters are key environment parameters used to distinguish thunderstorm gusts of different intensities.The areas of large-value key environment parameters are predominantly concentrated in the high occurrence region at the bottom of the NCCV,which accounts for the frequent occurrence of severe thunderstorm gusts in this region.(2)There is another high occurrence area with the azimuth angle of 235°—245° and within 1.5-2 times of the NCCV radius in the southwest quadrant,which is smaller than the high occurrence area at the bottom of the NCCV.Severe thunderstorm gusts in summer in Northeast China mainly occur in this area,where there is a low-level convergence line accompanied by a dryline.Compared to ordinary thunderstorm gusts,the convergence characteristics and dew point gradient in the convergence line corresponding to severe thunderstorm gusts are stronger.The observational characteristics and environmental conditions of severe thunderstorm gusts in different quadrants of NCCV can provide a reference for improving the forecast ability of severe thunderstorm gusts.
In the past two decades,both numerical weather prediction(NWP)models and AI-based large meteorological models have significantly improved the accuracy of medium-range weather forecasting.However,due to inherent model uncertainties and inadequate simulations over complex terrain areas,these models systematically underestimate extreme weather intensity in topographically challenging regions like the Qingzang Plateau.This study evaluates the performance of traditional NWP models,large meteorological models,and multi-model ensemble forecasts in predicting near-surface air temperature based on the case study of the 14 December 2023 cold wave event.Results indicate that while traditional NWP and large meteorological models effectively capture spatial patterns of temperature anomalies,they consistently underestimate extreme cold intensity.Although the multi-model ensemble mean can improve the spatial correlation coefficient to some extent,its performance in predicting the scope and intensity of extreme low temperatures still needs improvement.To address these limitations,we propose a swin transformer fusion(STF)model that incorporates positional encoding.This framework enables synergistic optimization of multi-model forecasts by systematically extracting and integrating the strengths of NWP and large meteorological models at specific spatiotemporal scales.During the cold wave's peak phase,STF reduces the forecast root mean square error by up to 39.62%,with notable improvements particularly in error-sensitive regions.The model's dynamic preference-error hedging mechanism effectively combines the multi-model advantages,enhancing both forecast accuracy and operational robustness for extreme weather events.This work advances cold wave early warning systems for high-altitude regions,introduces novel methodologies for extreme weather prediction,and demonstrates promising practical applications.
Sea surface temperature(SST)is a pivotal driver of weather and climate.Across the vast oceans,satellite-derived retrievals provide the primary source of global SST data.Yet frequent cloud cover can result in poor or invalid SST data if the retrievals are not well designed and work under all weather conditions.In addition,the inherent imager gaps between polar-orbiting satellite scan swaths leave extensive SST voids.To fill these SST data voids,the present work proposes a contour-aware reconstruction network(CARNet)specifically designed for improving spatial SST distributions.To achieve this,the continental contour is first utilized to eliminate the land-sea mixing field of views.The morphological dilation operators are then developed to enhance edge-gradient learning capabilities for extensive missing regions.The zonal and meridional gradient constraints are integrated into the loss function to preserve latitudinal-longitudinal SST distribution patterns.This algorithm is applied to FY-3D MWRI(Microwave Radiation Imager)-derived SST products from January to December 2023 to reconstruct a gap-free,twice-daily SST dataset spanning 60°S—60°N with global coverage.Evaluation demonstrates that the proposed reconstruction method accurately restores mesoscale eddy features in dynamic current systems(e.g.,tropical instability waves,Gulf Stream)while reducing positive biases in tropical cold tongues.Specifically,spring SST daily mean errors decrease from 0.8℃to approximately 0.5℃,with a concurrent reduction of 0.1℃in global daily SST standard deviation.The reconstructed dataset characterizes the evolutionary dynamics of the 2023 El Niño event,with computed Nino3.4 indices demonstrating high consistency with reanalysis benchmarks.This approach significantly enhances the operational usability of SST products retrieved from FY-3D MWRI data,providing reliable data support for ocean-atmosphere research.
This study analyzes the relationship of spatiotemporal configuration between lightning activities and ZDR and KDP columns during two squall lines over Guangzhou on 4 and 8 May 2017.The observations are from a S-band dual-polarization radar and the Low-Frequency Electric Field Detection Array.The results show that:(1)During the early stage of lightning activities(the intracloud(IC)lightning frequency is less than 50/(6 min),and the cloud-to-ground(CG)lightning frequency is less than 10/(6 min)),IC lightning and CG lightning mainly occurred within the plane spatial coverage of ZDR and KDP columns,and IC lightning tended to occur within the area of KDP columns(the maximum percentage can exceed 40%).As lightning activities became more active,IC lightning and CG lightning activities mainly occurred outside the planar coverage of the ZDR and KDP columns.(2)Changes in ZDR and KDP column volume led the trend of the IC lightning and CG lightning activity by about 30 min,and the correlation coefficient approached or exceeded 0.9.In addition,the peak of ZDR column volume was a good predictor of the occurrence of IC lightning and CG lightning peak.The results of this study have important scientific significance for understanding the relationship between lightning activities and the dynamic structure of the complex mesoscale weather systems.It also provides new perspectives for early warning and forecasting of lightning activities via dual-polarization weather radar data.
In recent years,ensemble forecasting has become a vital tool for major global weather forecasting centres.However,ensemble forecasts often exhibit underdispersion and systematic biases,making the application of statistical post-processing methods essential.The standardized anomalies model output statistics(SAMOS)is a commonly used post-processing technique that provides a complete description of the forecast distribution.Nevertheless,SAMOS typically relies only on predictors either directly related to the forecast variable or selected based on subjective judgment,potentially overlooking other valuable predictors.Moreover,directly incorporating too many predictors into SAMOS may lead to overfitting.Therefore,effectively selecting key predictors from numerous variables provided by forecast models remains a significant challenge.Boosting-based variable selection and optimization algorithms have proven to be effective in mitigating overfitting and identifying the most important predictors.This study proposes the standardized anomalies gradient boosting(SABST)method by integrating the strength of SAMOS with a boosting-based variable selection algorithm.SABST is applied to calibrate ensemble forecasts of 2 m air temperature,2 m relative humidity,and 10 m wind speed.The SABST model is developed using the European Centre for Medium-Range Weather Forecasts(ECMWF)high-resolution ensemble forecast(ensemble prediction system,ENS)products during 2019-2020 and is systematically compared with SAMOS,focusing on its calibration performance and bias correction ability.Results show that compared to ENS and SAMOS,SABST performs better in addressing underdispersion in probabilistic forecasts and improving the accuracy of deterministic forecasts.Based on the continuous ranked probability skill score(CRPSS),SABST improves the average CRPSS by 9.5%,15.3%,and 4.6%across all forecast lead time compared to SAMOS.These findings demonstrate the advantage of introducing additional potential predictors in the SABST framework and highlight the method's potential for application in ensemble forecast post-processing.
