ISSN 1000-0526
CN 11-2282/P
CN 11-2282/P
Volume 51,Issue 4,2025 Table of Contents
SHENG Jie
,
JIN Ronghua
,
ZHANG Xiaowen
,
DAI Kan
,
ZHANG Xiaoling
,
GUAN Liang
,
YANG Bo
,
ZHANG Yuchen
,
XING Lanxiang
,
LONG Mingsheng
,
WANG Jianmin
2025, 51(4):389-399.
DOI: 10.7519/j.issn.1000-0526.2025.032801
Abstract:
Traditional extrapolation techniques, such as the optical flow method, are the main objective methods currently used for nowcasting severe convective weather. These methods fail to represent the generation, dissipation, and evolution of convective systems, resulting in limited forecast validity periods. In 2024, the China Meteorological Administration released China’s first AI-based meteorological nowcasting model “Fenglei” V1 (hereafter referred to as “Fenglei”). “Fenglei” can generate 3 h extrapolation forecasts based on composite radar reflectivity. The results of quantitative verification on the 2023 data show that “Fenglei” outperforms the traditional optical flow extrapolation algorithms in objective verification scores, with more significant advantages for the forecasts exceeding a lead time of 1 h. Its verification scores decline relatively slowly and flatly, having relatively small Biases within the 3 h forecast lead time. Its TS score for severe and hazardous echo systems has been improved by 33% compared to the optical flow extrapolation algorithm. Case evaluations on the 2024 severe convective events of different scales reveal that “Fenglei” can accurately forecast the generation, dissipation, and evolution of convective systems within a certain forecast lead time. It shows the forecasting capability that traditional methods lack for thunderstorm trend evolution, effectively extending the extrapolation lead time. Thus, “Fenglei” can provide reliable AI-based objective forecast products for the nowcasting of severe convection.
Traditional extrapolation techniques, such as the optical flow method, are the main objective methods currently used for nowcasting severe convective weather. These methods fail to represent the generation, dissipation, and evolution of convective systems, resulting in limited forecast validity periods. In 2024, the China Meteorological Administration released China’s first AI-based meteorological nowcasting model “Fenglei” V1 (hereafter referred to as “Fenglei”). “Fenglei” can generate 3 h extrapolation forecasts based on composite radar reflectivity. The results of quantitative verification on the 2023 data show that “Fenglei” outperforms the traditional optical flow extrapolation algorithms in objective verification scores, with more significant advantages for the forecasts exceeding a lead time of 1 h. Its verification scores decline relatively slowly and flatly, having relatively small Biases within the 3 h forecast lead time. Its TS score for severe and hazardous echo systems has been improved by 33% compared to the optical flow extrapolation algorithm. Case evaluations on the 2024 severe convective events of different scales reveal that “Fenglei” can accurately forecast the generation, dissipation, and evolution of convective systems within a certain forecast lead time. It shows the forecasting capability that traditional methods lack for thunderstorm trend evolution, effectively extending the extrapolation lead time. Thus, “Fenglei” can provide reliable AI-based objective forecast products for the nowcasting of severe convection.
ZENG Xiaotuan
,
TAN Zhao
,
SHEN Yuwei
,
FAN Jiao
,
HUANG Rongcheng
,
ZHOU Hongyuan
,
LIANG Xiao
,
HUANG Dajian
2025, 51(4):400-416.
DOI: 10.7519/j.issn.1000-0526.2025.012001
Abstract:
To improve the accuracy of precipitation nowcasting, this paper proposes an adversarial neural network model named CastNet that combines deep neural networks. This model utilizes the recurrent neur-al network to capture the spatio-temporal features of radar echo data, employs the adversarial neural network to simulate the generation and dissipation changes of cloud clusters, and then integrates the optical flow constraint into the neural network to guide the model training. This accelerates the learning process of the neural network and enhances the spatio-temporal consistency of the model, effectively solving the problem of forecast ambiguity and significantly improving the accuracy of precipitation intensity and location. Tests are conducted on 9 major precipitation processes in Guangxi and its surrounding areas from May to October 2023. The results show that under various precipita-tion intensities (≥0.1, ≥2, ≥7, ≥15, ≥25, ≥40 mm·h-1), the average TS scores of SWAN 2.0 are 0.458, 0.270, 0.085, 0.034, 0.014 and 0.003， respectively; the average TS scores of SWAN 3.0 are 0.452, 0.402, 0.225, 0.129, 0.085 and 0.048， respectively; and the average TS scores of the CastNet model are 0.439, 0.397, 0.225, 0.139, 0.104 and 0.073， respectively. It can be seen clearly that the TS scores by the CastNet are higher than those of SWAN 2.0 and SWAN 3.0 under high-intensity precipita-tion of ≥7 mm·h-1 and above, except for few data points that are flat. In addition, as the forecast lead time extends, the relative advantage of CastNet becomes more obvious.
To improve the accuracy of precipitation nowcasting, this paper proposes an adversarial neural network model named CastNet that combines deep neural networks. This model utilizes the recurrent neur-al network to capture the spatio-temporal features of radar echo data, employs the adversarial neural network to simulate the generation and dissipation changes of cloud clusters, and then integrates the optical flow constraint into the neural network to guide the model training. This accelerates the learning process of the neural network and enhances the spatio-temporal consistency of the model, effectively solving the problem of forecast ambiguity and significantly improving the accuracy of precipitation intensity and location. Tests are conducted on 9 major precipitation processes in Guangxi and its surrounding areas from May to October 2023. The results show that under various precipita-tion intensities (≥0.1, ≥2, ≥7, ≥15, ≥25, ≥40 mm·h-1), the average TS scores of SWAN 2.0 are 0.458, 0.270, 0.085, 0.034, 0.014 and 0.003， respectively; the average TS scores of SWAN 3.0 are 0.452, 0.402, 0.225, 0.129, 0.085 and 0.048， respectively; and the average TS scores of the CastNet model are 0.439, 0.397, 0.225, 0.139, 0.104 and 0.073， respectively. It can be seen clearly that the TS scores by the CastNet are higher than those of SWAN 2.0 and SWAN 3.0 under high-intensity precipita-tion of ≥7 mm·h-1 and above, except for few data points that are flat. In addition, as the forecast lead time extends, the relative advantage of CastNet becomes more obvious.
2025, 51(4):417-430.
DOI: 10.7519/j.issn.1000-0526.2025.010201
Abstract:
According to the specific image feature that the bird echo shows obvious ring shape in the weather radar reflectivity product, this article proposes an improved algorithm based on a lightweight convolutional neural network You Only Look Once Version5 （YOLOv5） and multi-object tracking based on deep learning based simple online and realtime tracking （DeepSort）. The training and test datasets are constructed based on radar volume scanning echo intensity. Data obtained from the Yingkou Weather Radar from 2020 to 2023. The bird echoes are tracked, respectively. Firstly, Shuffle Attention （SA）, a lightweight attention mechanism, is introduced into YOLOv5 algorithm to improve the accuracy and effectiveness of the overall model checking. Secondly, in DeepSort algorithm, the original cross-merge-ratio intersection over union （IOU） matching mechanism is replaced by an improved loss function of object detection, distance-intersection over union （DIOU） matching mechanism. DIOU introduces the distance between the center points of the boundary box on the basis of calculating the overlap degree of the boundary box, so as to provide more accurate positioning. The number of identification （ID） error matching and ID switching caused by partial occlusion overlap is reduced. The test results show that the optimized YOLOv5 algorithm improves the accuracy by 2.6 percentage point, the recall rate by 1 percentage point, and the average accuracy of threshold values greater than 0.5 by 1.2 percentage point. The improved DeepSort algorithm reduces the number of ID switches by 2 times, and multi target tracking accuracy multi-object tracking accuracy （MOTA） increases by 4.5 percentage point, thus improves lightweight of the initial model. Generally, the overall checking performance is significantly improved, and may meet the actual demand for bird echo recognition and tracking.
According to the specific image feature that the bird echo shows obvious ring shape in the weather radar reflectivity product, this article proposes an improved algorithm based on a lightweight convolutional neural network You Only Look Once Version5 （YOLOv5） and multi-object tracking based on deep learning based simple online and realtime tracking （DeepSort）. The training and test datasets are constructed based on radar volume scanning echo intensity. Data obtained from the Yingkou Weather Radar from 2020 to 2023. The bird echoes are tracked, respectively. Firstly, Shuffle Attention （SA）, a lightweight attention mechanism, is introduced into YOLOv5 algorithm to improve the accuracy and effectiveness of the overall model checking. Secondly, in DeepSort algorithm, the original cross-merge-ratio intersection over union （IOU） matching mechanism is replaced by an improved loss function of object detection, distance-intersection over union （DIOU） matching mechanism. DIOU introduces the distance between the center points of the boundary box on the basis of calculating the overlap degree of the boundary box, so as to provide more accurate positioning. The number of identification （ID） error matching and ID switching caused by partial occlusion overlap is reduced. The test results show that the optimized YOLOv5 algorithm improves the accuracy by 2.6 percentage point, the recall rate by 1 percentage point, and the average accuracy of threshold values greater than 0.5 by 1.2 percentage point. The improved DeepSort algorithm reduces the number of ID switches by 2 times, and multi target tracking accuracy multi-object tracking accuracy （MOTA） increases by 4.5 percentage point, thus improves lightweight of the initial model. Generally, the overall checking performance is significantly improved, and may meet the actual demand for bird echo recognition and tracking.
2025, 51(4):431-445.
DOI: 10.7519/j.issn.1000-0526.2024.022003
Abstract:
Based on forecast products of the European Centre for Medium-Range Weather Forecasts-Integrated Forecasting System (ECMWF-IFS) and hourly temperature observation data from the CMA Land Data Assimilation System (CLDAS), an enhanced model named ED-LSTM-FCNN is constructed, with an embedding layer module incorporated to handle high-dimensional spatial and temporal features. A fully connected neural network is utilized to integrate various feature types, achieve regression prediction of temperature, and generate gridded hourly temperature forecast products with a resolution of 0.05°×0.05°. Verification for the forecast products in Hunan Province in 2022 shows that this model exhibits a notable capacity to mitigate forecast errors inherent in the numerical model, and can enhance the overall forecast stability. The root mean square errors (RMSEs) of forecasts with lead time ranging from 1 to 24 hours exhibit a reduction of 25.4%-37.7% when compared to ECMWF-IFS and a decrease of 15.8%-40.0% relative to the National Meteorological Centre forecast (SCMOC). The model can significantly improve the forecast performance of ECMWF-IFS forecast, in spatial scale, particularly in regions characterized by intricate terrain. The RMSEs across most areas vary within the range of 1.2-1.6℃. The forecast accuracy of the model, with an error margin of ±2℃, surpasses 83.0% across various seasons, demonstrating a significant improvement compared to both ECMWF-IFS and SCMOC. The forecasting performance is notably superior, particularly in stable extreme high-temperature weather conditions compared to alternative products. In conclusion, this model has proved to be effective in the high-resolution temperature forecasting operations.
Based on forecast products of the European Centre for Medium-Range Weather Forecasts-Integrated Forecasting System (ECMWF-IFS) and hourly temperature observation data from the CMA Land Data Assimilation System (CLDAS), an enhanced model named ED-LSTM-FCNN is constructed, with an embedding layer module incorporated to handle high-dimensional spatial and temporal features. A fully connected neural network is utilized to integrate various feature types, achieve regression prediction of temperature, and generate gridded hourly temperature forecast products with a resolution of 0.05°×0.05°. Verification for the forecast products in Hunan Province in 2022 shows that this model exhibits a notable capacity to mitigate forecast errors inherent in the numerical model, and can enhance the overall forecast stability. The root mean square errors (RMSEs) of forecasts with lead time ranging from 1 to 24 hours exhibit a reduction of 25.4%-37.7% when compared to ECMWF-IFS and a decrease of 15.8%-40.0% relative to the National Meteorological Centre forecast (SCMOC). The model can significantly improve the forecast performance of ECMWF-IFS forecast, in spatial scale, particularly in regions characterized by intricate terrain. The RMSEs across most areas vary within the range of 1.2-1.6℃. The forecast accuracy of the model, with an error margin of ±2℃, surpasses 83.0% across various seasons, demonstrating a significant improvement compared to both ECMWF-IFS and SCMOC. The forecasting performance is notably superior, particularly in stable extreme high-temperature weather conditions compared to alternative products. In conclusion, this model has proved to be effective in the high-resolution temperature forecasting operations.
QI Haixia
,
PENG Tao
,
ZHI Xiefei
,
JI Yan
,
YIN Zhiyuan
,
SHEN Tieyuan
,
WANG Junchao
,
XIANG Yiheng
,
HU Po
2025, 51(4):446-459.
DOI: 10.7519/j.issn.10000526.2025.031301
Abstract:
Flood forecasting is an effective nonengineering measure to reduce the economic losses brought by floods. Accurately forecasting flood is one of the key technical challenges in hydrological field. There are flood forecasting models based on physical mechanisms used, but the accuracy and efficiency of forecasts need to be improved. At present, forecasting models constructed by using deep learning techniques have been developed rapidly. This article comprehensively reviews the principles and characteristics of deep learning models that have been applied in the field of flood forecasting and summarizes their application progresses and problems in the quantitative and probabilistic flood forecasting. In addition, this article explores the relevance and application prospects of deep learning models combined with flood physics models, particularly in the parameterization of physical processes, interpretability studies, and error correction of flood forecasting models. The results suggest that the deep coupling of deep learning technology with physical models is the developing direction of deep learning models in the future. It will be an important development paradigm for the time series prediction of flood, and also an important research component to achieve intelligent water resource management in the future. Finally, to better apply the deep learning technology in the field of flood forecasting, some thoughts on the difficulties of deep learning in flood prediction are given and corresponding solutions are proposed for the current challenging problems.
Flood forecasting is an effective nonengineering measure to reduce the economic losses brought by floods. Accurately forecasting flood is one of the key technical challenges in hydrological field. There are flood forecasting models based on physical mechanisms used, but the accuracy and efficiency of forecasts need to be improved. At present, forecasting models constructed by using deep learning techniques have been developed rapidly. This article comprehensively reviews the principles and characteristics of deep learning models that have been applied in the field of flood forecasting and summarizes their application progresses and problems in the quantitative and probabilistic flood forecasting. In addition, this article explores the relevance and application prospects of deep learning models combined with flood physics models, particularly in the parameterization of physical processes, interpretability studies, and error correction of flood forecasting models. The results suggest that the deep coupling of deep learning technology with physical models is the developing direction of deep learning models in the future. It will be an important development paradigm for the time series prediction of flood, and also an important research component to achieve intelligent water resource management in the future. Finally, to better apply the deep learning technology in the field of flood forecasting, some thoughts on the difficulties of deep learning in flood prediction are given and corresponding solutions are proposed for the current challenging problems.
2025, 51(4):460-472.
DOI: 10.7519/j.issn.1000-0526.2024.111401
Abstract:
To address the issue of high-concealed abnormal wind directions in automatic weather station (AWS) data, this study establishes an abnormal wind direction identification method based on the density-based spatial clustering of applications with noise (DBSCAN) clustering algorithm. Historical wind direction data from 16 weather events affecting Guangzhou between 2016 and 2022, including cold waves, cold air masses, and typhoons, as well as observed wind direction data from AWSs during the impact of Typhoon Saola (No.2309), are used to detect abnormal wind directions. The analysis results reveal that the proportion of AWSs with suspicious wind directions in historical cases ranges from 0.46% to 5.56%, while the proportion of AWSs with erroneous wind directions varies from 0.25% to 2.05%. During the case of Typhoon Saola, the method identifies 13 AWSs with significantly deviating wind directions from the dominant surface wind direction, which is primarily due to wind direction sensor malfunctions and environmental impacts on AWS observations. Compared to that by the traditional method, the accuracy of wind direction error identification has improved by 20.32 percentage point. The new method provides a novel approach for the quality control of historical wind direction data from AWSs and offers an effective reference for the operational monitoring and on-site verification of AWS equipment.
To address the issue of high-concealed abnormal wind directions in automatic weather station (AWS) data, this study establishes an abnormal wind direction identification method based on the density-based spatial clustering of applications with noise (DBSCAN) clustering algorithm. Historical wind direction data from 16 weather events affecting Guangzhou between 2016 and 2022, including cold waves, cold air masses, and typhoons, as well as observed wind direction data from AWSs during the impact of Typhoon Saola (No.2309), are used to detect abnormal wind directions. The analysis results reveal that the proportion of AWSs with suspicious wind directions in historical cases ranges from 0.46% to 5.56%, while the proportion of AWSs with erroneous wind directions varies from 0.25% to 2.05%. During the case of Typhoon Saola, the method identifies 13 AWSs with significantly deviating wind directions from the dominant surface wind direction, which is primarily due to wind direction sensor malfunctions and environmental impacts on AWS observations. Compared to that by the traditional method, the accuracy of wind direction error identification has improved by 20.32 percentage point. The new method provides a novel approach for the quality control of historical wind direction data from AWSs and offers an effective reference for the operational monitoring and on-site verification of AWS equipment.
FU Zhilong
,
JIANG Shuai
,
LI Guoping
,
CHEN Fangli
,
HUANG Chuxian
,
LUO Rong
,
ZHANG Qiuming
,
LIANG Huilong
2025, 51(4):473-483.
DOI: 10.7519/j.issn.1000-0526.2024.122702
Abstract:
Based on the precipitation data from automatic weather stations and ERA5 reanalysis data, case seletion and classification of synoptic circulation are carried out in this paper to study the warm-sector heavy rainfall in Huizhou during the first rainy season from April to June of 2003-2022. And a comparative analysis is conducted on characteristics of average synoptic circulation and ambient parameters between different types of warm-sector rainfall events. The results show that a total of 48 warm-sector heavy rainfall events occurred in Huizhou during the first rainy season in 2003-2022. Accroding to synoptic circulation, the selected warm-sector rainfall events are divided into three types, i.e., shear line (the first type), shortwave trough with low-level jet (the second type), and the edge of subtropical high with the entrance of low-level jet (the third type). By further comparing the average synoptic circulation between different types of warm-sector rainfall events, we reveal that Huizhou is under the control of the west wind flow and the southwest flow in the periphery of subtropical high at 500 hPa, except the second type of heavy rainfall. In low levels, there are double low-level jet (southwest low-level jet and boundary layer low-level jet) near Huizhou both in the second and third types of heavy rainfall, while in the first type of heavy rain, there is boundary layer low-level jet only to the south of the Pearl River Estuary at 925 hPa. The analysis of the ambient features indicates that the ambient parameters calculated based on ERA5 reanalysis are reliable and applicable. The second and third types of heavy rainfall are superior to the first type of heavy rainfall in moisture and energy conditions. But in terms of dynamic conditions, the vertical wind shear, and the static instability of the first type of heavy rainfall is considerably stronger than those of the other two types of heavy rainfall.
Based on the precipitation data from automatic weather stations and ERA5 reanalysis data, case seletion and classification of synoptic circulation are carried out in this paper to study the warm-sector heavy rainfall in Huizhou during the first rainy season from April to June of 2003-2022. And a comparative analysis is conducted on characteristics of average synoptic circulation and ambient parameters between different types of warm-sector rainfall events. The results show that a total of 48 warm-sector heavy rainfall events occurred in Huizhou during the first rainy season in 2003-2022. Accroding to synoptic circulation, the selected warm-sector rainfall events are divided into three types, i.e., shear line (the first type), shortwave trough with low-level jet (the second type), and the edge of subtropical high with the entrance of low-level jet (the third type). By further comparing the average synoptic circulation between different types of warm-sector rainfall events, we reveal that Huizhou is under the control of the west wind flow and the southwest flow in the periphery of subtropical high at 500 hPa, except the second type of heavy rainfall. In low levels, there are double low-level jet (southwest low-level jet and boundary layer low-level jet) near Huizhou both in the second and third types of heavy rainfall, while in the first type of heavy rain, there is boundary layer low-level jet only to the south of the Pearl River Estuary at 925 hPa. The analysis of the ambient features indicates that the ambient parameters calculated based on ERA5 reanalysis are reliable and applicable. The second and third types of heavy rainfall are superior to the first type of heavy rainfall in moisture and energy conditions. But in terms of dynamic conditions, the vertical wind shear, and the static instability of the first type of heavy rainfall is considerably stronger than those of the other two types of heavy rainfall.
2025, 51(4):484-495.
DOI: 10.7519/j.issn.1000-0526.2024.121301
Abstract:
The primary source of the bias of atmospheric motion vectors (AMVs) is related to inaccuracies in assigning heights. The cloud top height algorithm employed by China’s new generation geostationary meteorological satellite FY-4A make use of both the infrared window channel and CO2 slicing channel. This leads to an improvement in its accuracy. In this study, cloud top height is used to correct AMVs of FY-4A satellite through spatiotemporal matching between the two products. Representative pixels within the AMVs tracking box are searched, and their average cloud top height is used to replace the original cloud height of the AMVs, achieving height reassignment. Verification using ERA5 reanalysis data shows that after height reassignment, the root mean square error (RMSE) of FY-4A infrared channel AMVs is significantly reduced across high, medium, and low layers. Specifically, the RMSE for the high layer decreases from 4.06 m·s-1 to 3.25 m·s-1, for the medium layer from 4.25 m·s-1 to 3.71 m·s-1, and for the low layer from 2.42 m·s-1 to 2.13 m·s-1. Height reassignment alleviates the problem of AMVs being assigned to overly high altitudes, reducing biases, particularly improving slow motion biases. Case studies of the Northeast China cold vortex and Typhoon Chaba demonstrate that this method can improve the consistency between cloud-derived winds and the background field. Promising applications of this method in numerical weather prediction assimilation and weather process analysis are envisioned.
The primary source of the bias of atmospheric motion vectors (AMVs) is related to inaccuracies in assigning heights. The cloud top height algorithm employed by China’s new generation geostationary meteorological satellite FY-4A make use of both the infrared window channel and CO2 slicing channel. This leads to an improvement in its accuracy. In this study, cloud top height is used to correct AMVs of FY-4A satellite through spatiotemporal matching between the two products. Representative pixels within the AMVs tracking box are searched, and their average cloud top height is used to replace the original cloud height of the AMVs, achieving height reassignment. Verification using ERA5 reanalysis data shows that after height reassignment, the root mean square error (RMSE) of FY-4A infrared channel AMVs is significantly reduced across high, medium, and low layers. Specifically, the RMSE for the high layer decreases from 4.06 m·s-1 to 3.25 m·s-1, for the medium layer from 4.25 m·s-1 to 3.71 m·s-1, and for the low layer from 2.42 m·s-1 to 2.13 m·s-1. Height reassignment alleviates the problem of AMVs being assigned to overly high altitudes, reducing biases, particularly improving slow motion biases. Case studies of the Northeast China cold vortex and Typhoon Chaba demonstrate that this method can improve the consistency between cloud-derived winds and the background field. Promising applications of this method in numerical weather prediction assimilation and weather process analysis are envisioned.
2025, 51(4):496-507.
DOI: 10.7519/j.issn.1000-0526.2025.021901
Abstract:
Based on the NCEP/NCAR reanalysis datasets and climate observation from 2400 stations of China, the characteristics of climatic anomaly over China in autumn 2024 and the possible causes are analyzed. The results denote that, in autumn 2024, the mean temperature in China peaked the record in the same period since 1961 and the average precipitation was slightly more than the climatic mean, with the spatial distribution generally characterized by “more in the north and less in the south”. In autumn, the typhoons generated in the Northwest Pacific and the landing typhoons in China were more than normal, and their paths were mainly in the northwest direction. The circulation in the mid-high latitudes in East Asia showed the feature of “low in the west and high in the east”. The West Pacific subtropical high (WPSH) was stronger and more westward than usual overall, and its ridge line swung frequently in the meridional direction, that is, the ridge line was abnormally further north than usual in September, slightly north than usual in October and close to normal position in November, which led to the abnormal warming in China. The SST in the West Pacific was higher than normal. The South China Sea and its eastern area were the positive vorticity area with small vertical wind shear, and the significantly higher pseudo-equivalent temperature in the lower level from the northern Philippines to the south of the Sea of Japan, were conducive to the formation of abnormally more typhoons in autumn. The strengthening of convection activity from the maritime mainland to the West Pacific brought by MJO eastward spread from the end of August to September was beneficial to the abnormal northward position of the subtropical high, which made the typhoon track by north in autumn. In addition, the abnormally lower SST in the equatorial Middle East Pacific and the unusually late retreat of the summer monsoon in the South China Sea all contributed to the frequent occurrence and northward track of typhoons in autumn 2024.
Based on the NCEP/NCAR reanalysis datasets and climate observation from 2400 stations of China, the characteristics of climatic anomaly over China in autumn 2024 and the possible causes are analyzed. The results denote that, in autumn 2024, the mean temperature in China peaked the record in the same period since 1961 and the average precipitation was slightly more than the climatic mean, with the spatial distribution generally characterized by “more in the north and less in the south”. In autumn, the typhoons generated in the Northwest Pacific and the landing typhoons in China were more than normal, and their paths were mainly in the northwest direction. The circulation in the mid-high latitudes in East Asia showed the feature of “low in the west and high in the east”. The West Pacific subtropical high (WPSH) was stronger and more westward than usual overall, and its ridge line swung frequently in the meridional direction, that is, the ridge line was abnormally further north than usual in September, slightly north than usual in October and close to normal position in November, which led to the abnormal warming in China. The SST in the West Pacific was higher than normal. The South China Sea and its eastern area were the positive vorticity area with small vertical wind shear, and the significantly higher pseudo-equivalent temperature in the lower level from the northern Philippines to the south of the Sea of Japan, were conducive to the formation of abnormally more typhoons in autumn. The strengthening of convection activity from the maritime mainland to the West Pacific brought by MJO eastward spread from the end of August to September was beneficial to the abnormal northward position of the subtropical high, which made the typhoon track by north in autumn. In addition, the abnormally lower SST in the equatorial Middle East Pacific and the unusually late retreat of the summer monsoon in the South China Sea all contributed to the frequent occurrence and northward track of typhoons in autumn 2024.
2025, 51(4):508-516.
DOI: 10.7519/j.issn.1000-0526.2025.040701
Abstract:
The main characteristics of the general atmospheric circulation in January 2025 are as follows. The polar vortex in the Northern Hemisphere exhibited a dipole pattern, the East Asian trough was more eastward, and the southern branch trough was weaker. The monthly average precipitation (9.3 mm) over China was 30.1% less than the average value 13.3 mm in the same period of normal years. Moreover, the spatial distribution of precipitation was uneven, more in the north and less in the south of China. The monthly mean temperature was -3.3℃, which is 1.5℃ higher than in the same period of normal years (-4.8℃). Temperatures fluctuated significantly throughout the month, with one nationwide cold wave event and one large-scale persistent fog-haze weather event. The cold wave and snow-rain event from 23 to 27 was characterized by severe temperature drop, strong wind force, widespread rain and snow with substantial accumulated precipitation, heavy snowfall and deep snow depth. As a result, many new weather records were built in many regions.
The main characteristics of the general atmospheric circulation in January 2025 are as follows. The polar vortex in the Northern Hemisphere exhibited a dipole pattern, the East Asian trough was more eastward, and the southern branch trough was weaker. The monthly average precipitation (9.3 mm) over China was 30.1% less than the average value 13.3 mm in the same period of normal years. Moreover, the spatial distribution of precipitation was uneven, more in the north and less in the south of China. The monthly mean temperature was -3.3℃, which is 1.5℃ higher than in the same period of normal years (-4.8℃). Temperatures fluctuated significantly throughout the month, with one nationwide cold wave event and one large-scale persistent fog-haze weather event. The cold wave and snow-rain event from 23 to 27 was characterized by severe temperature drop, strong wind force, widespread rain and snow with substantial accumulated precipitation, heavy snowfall and deep snow depth. As a result, many new weather records were built in many regions.
Mobile website
® 2025 All Rights Reserved
Supported by:Beijing E-Tiller Technology Development Co., Ltd.
ISSN:1000-0526 CN:11-2282/P