A deep learning objective forecasting solution for severe convective weather(SCW) including short-duration heavy rain(HR), hail, convective gusts(CG), and thunderstorms based on numerical weather prediction(NWP) data ...A deep learning objective forecasting solution for severe convective weather(SCW) including short-duration heavy rain(HR), hail, convective gusts(CG), and thunderstorms based on numerical weather prediction(NWP) data was developed. We first established the training datasets as follows. Five years of severe weather observations were utilized to label the NCEP final(FNL) analysis data. A large number of labeled samples for each type of weather were then selected for model training. The local temperature, pressure, humidity, and winds from 1000 to 200 h Pa, as well as dozens of convective physical parameters, were taken as predictors in our model. A six-layer convolutional neural network(CNN) model was then built and trained to obtain optimal model weights. After that, the trained model was used to predict SCW based on the Global Forecast System(GFS) forecast data as input. The performances of the CNN model and other traditional methods were compared. The results show that the deep learning algorithm had a higher classification accuracy on HR and hail than support vector machine, random forests, and other traditional machine learning algorithms. The objective forecasts by use of the deep learning algorithm also showed better forecasting skills than the subjective forecasts by the forecasters. The threat scores(TSs) of thunderstorm, HR, hail, and CG were increased by 16.1%, 33.2%, 178%, and 55.7%, respectively. The deep learning forecast model is currently used in the National Meteorological Center of China to provide guidance for the operational SCW forecasting over China.展开更多
In view of the fact that the atmospheric motion isan irreversible process, a memory function which can recall observational data in the past is suggested. In terms of defining an inner product in Hilbert space, a new ...In view of the fact that the atmospheric motion isan irreversible process, a memory function which can recall observational data in the past is suggested. In terms of defining an inner product in Hilbert space, a new concept of self-memorization in the atmospheric motion is proposed, thus the traditional atmospheric motion equation is generalized to a self-memorization equation in inclusion of multi-time observations. Self-memorization equations of a barotropic nondivergent model and a barotropic primitive equation model are exemplified.It is proved that some existing difference schemes can be derived from the self-memorization equation by giving particular values to memory function. However, it demonstrates that multi-time numerical prediction models can be unified into a framework of self-memorization equation. If a stochastic method in solving the memory function is taken, the selfmemorization equation will be transformed to a sort of dynamic-stochastic prediction model.展开更多
This article reviews the advances in severe convection research and operation in China during the past several decades.The favorable synoptic situations for severe convective weather(SCW),the major organization modes ...This article reviews the advances in severe convection research and operation in China during the past several decades.The favorable synoptic situations for severe convective weather(SCW),the major organization modes of severe convective storms(SCSs),the favorable environmental conditions and characteristics of weather radar echoes and satellite images of SCW and SCSs,and the forecasting and nowcasting techniques of SCW,are emphasized.As a whole,Chinese scientists have achieved a profound understanding of the synoptic patterns,organization,and evolution characteristics of SCW from radar and satellite observations,and the mechanisms of different types of convective weather in China.Specifically,in-depth understanding of the multiple types of convection triggers,along with the environmental conditions,structures and organization modes,and maintenance mechanisms of supercell storms and squall lines,has been obtained.The organization modes and climatological distributions of mesoscale convective systems and different types of SCW,and the multiscale characteristics and formation mechanisms of large hail,tornadoes,downbursts,and damaging convective wind gusts based on radar,satellite,and lightning observations,as well as the related features from damage surveys,are elucidated.In terms of operational applications,different types of identification and mesoanalysis techniques,and various forecasting and nowcasting techniques using methods such as the"ingredients-based"and deep learning algorithms,have been developed.As a result,the performance of operational SCW forecasts in China has been significantly improved.展开更多
基金Supported by the National Key Research and Development Program of China(2017YFC1502000 and 2018YFC1507504)National Natural Science Foundation of China(41375051)partially supported by AGS-1602845 and DMS-1830312 of the US National Science Foundation
文摘A deep learning objective forecasting solution for severe convective weather(SCW) including short-duration heavy rain(HR), hail, convective gusts(CG), and thunderstorms based on numerical weather prediction(NWP) data was developed. We first established the training datasets as follows. Five years of severe weather observations were utilized to label the NCEP final(FNL) analysis data. A large number of labeled samples for each type of weather were then selected for model training. The local temperature, pressure, humidity, and winds from 1000 to 200 h Pa, as well as dozens of convective physical parameters, were taken as predictors in our model. A six-layer convolutional neural network(CNN) model was then built and trained to obtain optimal model weights. After that, the trained model was used to predict SCW based on the Global Forecast System(GFS) forecast data as input. The performances of the CNN model and other traditional methods were compared. The results show that the deep learning algorithm had a higher classification accuracy on HR and hail than support vector machine, random forests, and other traditional machine learning algorithms. The objective forecasts by use of the deep learning algorithm also showed better forecasting skills than the subjective forecasts by the forecasters. The threat scores(TSs) of thunderstorm, HR, hail, and CG were increased by 16.1%, 33.2%, 178%, and 55.7%, respectively. The deep learning forecast model is currently used in the National Meteorological Center of China to provide guidance for the operational SCW forecasting over China.
文摘In view of the fact that the atmospheric motion isan irreversible process, a memory function which can recall observational data in the past is suggested. In terms of defining an inner product in Hilbert space, a new concept of self-memorization in the atmospheric motion is proposed, thus the traditional atmospheric motion equation is generalized to a self-memorization equation in inclusion of multi-time observations. Self-memorization equations of a barotropic nondivergent model and a barotropic primitive equation model are exemplified.It is proved that some existing difference schemes can be derived from the self-memorization equation by giving particular values to memory function. However, it demonstrates that multi-time numerical prediction models can be unified into a framework of self-memorization equation. If a stochastic method in solving the memory function is taken, the selfmemorization equation will be transformed to a sort of dynamic-stochastic prediction model.
基金Supported by the National Key Research and Development Program of China(2018YFC1507504 and 2017YFC1502000)National Natural Science Foundation of China(41775044 and 41375051)Strategic Research Projects on Medium-and Long-term Development of Chinese Engineering Science and Technology(2019-ZCQ-06)。
文摘This article reviews the advances in severe convection research and operation in China during the past several decades.The favorable synoptic situations for severe convective weather(SCW),the major organization modes of severe convective storms(SCSs),the favorable environmental conditions and characteristics of weather radar echoes and satellite images of SCW and SCSs,and the forecasting and nowcasting techniques of SCW,are emphasized.As a whole,Chinese scientists have achieved a profound understanding of the synoptic patterns,organization,and evolution characteristics of SCW from radar and satellite observations,and the mechanisms of different types of convective weather in China.Specifically,in-depth understanding of the multiple types of convection triggers,along with the environmental conditions,structures and organization modes,and maintenance mechanisms of supercell storms and squall lines,has been obtained.The organization modes and climatological distributions of mesoscale convective systems and different types of SCW,and the multiscale characteristics and formation mechanisms of large hail,tornadoes,downbursts,and damaging convective wind gusts based on radar,satellite,and lightning observations,as well as the related features from damage surveys,are elucidated.In terms of operational applications,different types of identification and mesoanalysis techniques,and various forecasting and nowcasting techniques using methods such as the"ingredients-based"and deep learning algorithms,have been developed.As a result,the performance of operational SCW forecasts in China has been significantly improved.
