Improved Weather Radar Echo Extrapolation Through Wind Speed Data Fusion Using a New Spatiotemporal Neural Network Model

Weather radar echo extrapolation plays a crucial role in weather forecasting.However,traditional weather radar echo extrapolation methods are not very accurate and do not make full use of historical data.Deep learning algorithms based on Recurrent Neural Networks also have the problem of accumulating errors.Moreover,it is difficult to obtain higher accuracy by relying on a single historical radar echo observation.Therefore,in this study,we constructed the Fusion GRU module,which leverages a cascade structure to effectively combine radar echo data and mean wind data.We also designed the Top Connection so that the model can capture the global spatial relationship to construct constraints on the predictions.Based on the Jiangsu Province dataset,we compared some models.The results show that our proposed model,Cascade Fusion Spatiotemporal Network(CFSN),improved the critical success index(CSI)by 10.7%over the baseline at the threshold of 30 dBZ.Ablation experiments further validated the effectiveness of our model.Similarly,the CSI of the complete CFSN was 0.004 higher than the suboptimal solution without the cross-attention module at the threshold of 30 dBZ.

Characteristics of Radar Echo Parameters and Microphysical Structure Simulation of a Short-Time Heavy Precipitation Supercell

By using the conventional observations, radar data, NCEP/NCAR FNL 1°×1° reanalysis data and numerical simulation data and with the construction and calculation of radar echo parameters, this paper presents the structural characteristics and physical processes of a short-time heavy precipitation supercell that occurred in the squall line process in Shanxi Province on 24 June 2020. The results show that this squall line event occurred in front of a surface cold front,combined with infiltration of low-level cold air and continuous increase of near-surface humidity in the afternoon. The surface mesoscale convergence line and mesoscale dew point front contributed to the development and systemization of the squall line by a large degree. The short-time extremely heavy precipitation in Pingshun County was caused by the development of a supercell from thunderstorm cells on the front side of the squall line. The characteristics of sharp increase in vertical integral liquid water content, persistent increase in reflectivity factor and continuous rise in the echo top height appeared about 23 min earlier than the severe precipitation, which has qualitative indicating significance for the nowcasting of short-time heavy precipitation. A quantitative analysis of the radar echo parameters suggests that the“sudden drop”of FV40was a precursor signal of cells’ coalescence and rapid development to the mature stage. The areal change of the echo core at the 6 km height was highly subject to the merging and developing of cells, the rapid change of hydrometeor particles in clouds and the precipitation intensity. Changes in the cross-sectional area of convective cells at different heights can indirectly reflect the changes of liquid particles and ice particles in clouds, which is indicatively meaningful for predicting the coalescing and developing-to-maturing of cells and heavy precipitation 30-45 min earlier.A comprehensive echo parameter prediction model constructed by the random forest principle can predict the magnitude of short-time heavy precipitation 40-50 min in advance. Numerical simulation reveals that large amounts of water vapor existed in the near-surface atmosphere, and that the cells rapidly obtained moisture from the ambient atmosphere and developed rapidly through maternal feeding. The cold cloud zone was narrow, upright and had a high stretch height. The upward motion in clouds was strong and deep, and very rich in liquid water content. The graupel particles had a large vertical distribution range, the coexistence area of graupel and snow was large, the height of raindrops was close to the surface with a wide horizontal scale, and the precipitation efficiency was high. These may be the important elements responsible for the occurrence of the short-time heavy precipitation that exceeded historical extreme values. On the basis of the above analyses, a comprehensive parameter(CP) prediction model is worked out, which can estimate the developing trend of supercells and the intensity of short-time heavy precipitation about 1 h in advance.

A novel method for radar echo simulation based on fast-constructed database

This paper presents a novel method for fast calculation of radar echo in near-field regions after the equivalent source has been computed by method of moments(MoM).An easy-to-access near-field database(NFDB)is established,which is built on the auxiliary tetrahedral meshes surrounding the nearfield regions of interest.The near-fields calculation(NFC)of arbitrary observation points can be expressed explicitly via the NFDB.An efficient matrix compression scheme named random sampling-based butterfly factorization(RS-BF)is proposed to speed up the construction of NFDB.With this approach,each group of O(N)elements in the database can be calculated through one fast matrix-vector multiplication operation that has a computational complexity below O(Nlog~2 N).The proposed method can avoid time-consuming point-by-point NFC of the traditional methods.Several numerical examples are presented to demonstrate the accuracy and efficiency of this method.In particular,the echo simulation of a missile-target encounter example is presented to illustrate its capability for practical applications.

CHARACTERISTICS AND RELATIONS OF LIGHTNING AND RADAR ECHOES FOR STRONG CONVECTIVE RAINSTORMS IN YUNNAN

Based on cloud-ground lightning data and Doppler weather radar echo products, both thecharacteristics and the relations of lightning and radar echoes for strong convective rainstorms over Yunnanare analyzed during the flood season of 2007. The results show that most rainstorms are convective in whichlightning is mostly negative and the negative lightning number accounts for more than 90% of the total.Although the correlation between precipitation and the lightning number is small on the rainstorm day, thelarge day-lightning frequency usually produces heavy precipitation. Hourly evolution of precipitation andlightning frequency shows peak-style characteristics. And their evolution is very coherent in strongrainstorm, but lightning often occurs before precipitation, whose peaks are in phase with or 1-to-2-hourlagged behind that of lightning frequency. Meanwhile the peaks of positive frequency are in phase with orfall behind that of precipitation. When the wind field is heterogeneous in radial velocity, it is conducive toboth the development of convection echoes and occurrence of lightning. Strong lightning-producingconvective rainstorms correspond to strong echo fields and usually result in reflectivity above 30 dBZ andecho top ET of more than 9 km, respectively.

RADAR ECHO DELAY IN THE(Ω,Aμν)-FIELD THEORY

（康谭珠迪）ＲＡＤＡＲＥＣＨＯＤＥＬＡＹＩＮＴＨＥ（Ω，Ａμν）－ＦＩＥＬＤＴＨＥＯＲＹ￥ＨｏｎｇＴｕｍｊｕｄｙ（ＤｅｐａｒｔｍｅｎｔｏｆＡｐｐｌｉｅｄＭａｔｈｅｍａｔｉｃｓ．ＨｏｎｇＫｏｎｇＰｏｌｙｔｅｃｈｎｉｃａｌＵｎｉｖｅｒｓｉｔｙ．ＨｏｎｇＫ...

THE STATISTIC RELATION BETWEEN LIGHTNING AND CINRAD DOPPLER RADAR ECHOES IN CENTRAL GUANGDONG

Based on the CINRAD Doppler radar data in Guangzhou and the lightning data in 2004 by power suppliers of Guangdong, statistical study is done by overlaying lightning’s position on radar’s echo. The result shows the followings. The concentrated period in which more negative lightning occurred at the middle levels (2 – 14 km), where radar echo was moderate (12 – 45 dBz), rather than at the low levels with the weakest echoes or at high levels with the strongest echoes. At levels 3 – 11 km, where the radar echo was between 10 dBz and 35 dBz, the area of negative lightning was much larger in central Guangdong than in the rest of the province. At levels 0.5 – 7 km where the radar echoes were between 44 dBz and 51 dBz, the probability for a point to have negative lightning varies from 0.4 to 0.7.

Algorithm of Vertical Profiles of Radar Echoes between Any Two Points Based on Webpage and Its Application

Based on the puzzle data of radar reflectivity in and around Jiangxi Province,the slope profile method and the polar coordinate method were used to display the vertical profiles of radar echoes between any two points.The technical route of extracting radar echo profile data by the slope method was proposed and applied to the short-term early warning platform of Jiangxi Province.A thunderstorm weather process occurring from Dongxiang to Wannian on September 23,2018 was analyzed by radar echo profiles.At present,the program has been applied to the short-term early warning platform of Jiangxi Province based on B/S architecture.The automatic drawing and display of radar echo profiles of any two points is practical for forecasters to quickly and effectively track short-term strong convection weather.

Analysis on the Radar Echo's Characteristics of One Severe Rainstorm in Central Guangxi

[Objective]The aim was to analyze radar echo characteristic of a heavy rainstorm in central Guangxi in June,2010. [Method] Using conventional observations and automatic stations,Doppler radar and other data,a severe rainstorm in central of Guangxi which occurred on 31 May to 1 June 2010 was analyzed. The characteristics of weather radar data were analyzed to reveal the characteristics of radar echo during the heavy precipitation process. [Result]The heavy rainstorm had gone through many singular wind storms towards MCS variation. The existence of ultra-sin-gular and " train effect" was the main reason of the heavy rainstorm. The results showed that during the heavy rainfall experienced an evolution process from multi-cell storm to MCS,and found that super cell and the " train effect" was the major reason which caused this torrential rain. The echo centroid of multi-cell storm stretching along its moving reverse order was the key factor which maintained the heavy rain. Analysis also found that the average echo intensity had a good relationship with rainfall. Strong echoes were in quasi-stationary state for a long time and was favorable heavy rain. In the diameter velocity diagram,mesoscale convergence line,the adverse wind regions and the mid-cyclone were the important basis to determine producing and keeping heavy rainfall. Radar wind profile can reflect the configuration of layers of wind conditions,and it was an effective tool to determine whether the trough was in transition or not. [Conclusion]The study provided reference for the short-term and nowcasting report.

CEMA-LSTM:Enhancing Contextual Feature Correlation for Radar Extrapolation Using Fine-Grained Echo Datasets

Accurate precipitation nowcasting can provide great convenience to the public so they can conduct corresponding arrangements in advance to deal with the possible impact of upcoming heavy rain.Recent relevant research activities have shown their concerns on various deep learning models for radar echo extrapolation,where radar echo maps were used to predict their consequent moment,so as to recognize potential severe convective weather events.However,these approaches suffer from an inaccurate prediction of echo dynamics and unreliable depiction of echo aggregation or dissipation,due to the size limitation of convolution filter,lack of global feature,and less attention to features from previous states.To address the problems,this paper proposes a CEMA-LSTM recurrent unit,which is embedded with a Contextual Feature Correlation Enhancement Block(CEB)and a Multi-Attention Mechanism Block(MAB).The CEB enhances contextual feature correlation and supports its model to memorize significant features for near-future prediction;the MAB uses a position and channel attention mechanism to capture global features of radar echoes.Two practical radar echo datasets were used involving the FREM and CIKM 2017 datasets.Both quantification and visualization of comparative experimental results have demonstrated outperformance of the proposed CEMA-LSTMover recentmodels,e.g.,PhyDNet,MIM and PredRNN++,etc.In particular,compared with the second-rankedmodel,its average POD,FAR and CSI have been improved by 3.87%,1.65%and 1.79%,respectively on the FREM,and by 1.42%,5.60%and 3.16%,respectively on the CIKM 2017.

Impacts of space-time-frequency synchronization errors onwide-band target echo characteristics of bistatic/multistatic radar

Bistatic/multistatic radar has great potential advantages over its monostatic counterpart. However, the separation of a transmitter and a receiver leads to difficulties in locating the target position accurately and guaranteeing space-timefrequency synchronization of the transmitter and the receiver.The error model of space-time-frequency synchronization in a motion platform of bistatic/multistatic radar is studied. The relationship between the space synchronization error and the transmitter platform position, receiver platform position, moving state, and beam pointing error, is analyzed. The effect of space synchronization error on target echo power is studied. The target scattering characteristics are restructured by many separate scattering centers of the target in high frequency regions. Based on the scattering centers model of the radar target, this radar target echo model and the simulation method are discussed. The algorithm of bistatic/multistatic radar target echo accurately reflects the scattering characteristics of the radar target, pulse modulation speciality of radar transmitting signals, and spacetime-frequency synchronization error characteristics between the transmitter station and the receiver station. The simulation of bistatic radar is completed in computer, and the results of the simulation validate the feasibility of the method.

The Echo Modelling and Simulation of the Semi-Active Radar Seeker against a Sea Skimming Target

This paper has proposed a new modelling and simulating technique for the echo of the semi-active radar seeker against the sea skimming target. The echo modelling is based on the electromagnetic scattering mechanisms. A modified Four-path model based on the radar detection scene is used to describe the multipath scattering between the target and rough sea surface. A Facet-based Small Slope Approximation (FBSSA) method is employed to calculate the scattering from the sea surface. The Physical Optics (PO) and the Equivalent Edge Current (EEC) Method is used to calculate the target scattering. In the echo simulations. The results present the original echo and the echo processed by the signal processing procedures, where the clutter and multipath effect can be observed.

Identification and Removal of Non-meteorological Echoes in Dual-polarization Radar Data Based on a Fuzzy Logic Algorithm

A major issue in radar quantitative precipitation estimation is the contamination of radar echoes by non-meteorological targets such as ground clutter,chaff,clear air echoes etc.In this study,a fuzzy logic algorithm for the identification of non-meteorological echoes is developed using optimized membership functions and weights for the dual-polarization radar located at Mount Sobaek.For selected precipitation and non-meteorological events,the characteristics of the precipitation and non-meteorological echo are derived by the probability density functions of five fuzzy parameters as functions of reflectivity values.The membership functions and weights are then determined by these density functions.Finally,the nonmeteorological echoes are identified by combining the membership functions and weights.The performance is qualitatively evaluated by long-term rain accumulation.The detection accuracy of the fuzzy logic algorithm is calculated using the probability of detection(POD),false alarm rate(FAR),and clutter–signal ratio(CSR).In addition,the issues in using filtered dual-polarization data are alleviated.

Aspect sensitivity of polar mesosphere summer echoes observed with the EISCAT VHF radar

The European Incoherent Scatter Scientific Association(EISCAT) Very High Frequency(224 MHz) Radar has been used to investigate the aspect sensitivity of polar mesosphere summer echoes(PMSE) in the period 13–15 July 2010. The aspect sensitivity of PMSE using this radar and at such a high frequency has not been previously reported. Data concerning the aspect sensitivity of PMSE were collected by traversing the antenna beam from the zenith direction, and comparing the received power. Surprisingly, as the intensity received by the oblique beam was often larger than that of the vertical beam, suggesting the presence of tilted dusty plasma layers as a potential cause, a theoretical model was developed to confirm the existence of these layers and their formation process. The experimental results and theoretical model presented help elucidate the structural properties of the possible generation mechanism of strong radar echoes in the polar summer mesosphere region.

Analysis on Echo Characteristics of Weather Radar of One Hailstorm Process in the First Division

Based on weather radar observation data,development and evolution characteristics of radar intensity echoes PPI and RHI of one ha shooting weather process in the Fifth Regiment of the First Division on May 24,2011 were analyzed. The results showed that local thermal forcin effect was obvious. The Fifth Regiment is close to mountain region,and the supplement of downward cold air from the mountain region was th most direct,and the supplement efficiency was the highest,which was main cause for the hail shooting generated by local severe convective weath er in west area of south Xinjiang. In the hail weather development process,vault( weak echo zone),front suspending echo and echo wall of supe cell on radar echo showed that airflow activity was strong in windstorm,which had an important indication significance for discrimination of ha cloud. The research could provide reference basis for conducting artificial hail suppression and proximity prediction of severe convective weather i reclamation area in the future,and had certain indication significance and better application value.

Kinematic Features of a Bow Echo in Southern China Observed with Doppler Radar

A bow echo is a type of mesoscale convective phenomenon that often induces extreme weather and appears with strong reflectivity on radar images.A strong bow echo that developed from a supercell was observed over Foshan City in southern China on 17 April 2011.The intense gusty winds and showers caused huge losses of property and severely a?ected human lives.This paper presents an analysis of this strong meso-β-scale convective system based on Doppler radar observations.The isolated bow echo exhibited a horizontal scale of about 80 km in terms of reflectivity above 40 dBZ,and a life span of 8 hours.The system originated from the merging of a couple of weakly organized cells in a shear line,and developed into an arch shape as it moved through the shear zone.Su?cient surface moisture supply ensured the convective instability and development of the bow echo.The low-altitude winds retrieved from single Doppler radar observations showed an obvious rear-inflow jet along the notch area.Di?erent from the conventional definition,no bookend anticyclone was observed throughout the life cycle.Very strong slantwise updrafts and downdrafts were recognizable from the retrieved winds,even though the spatial scale of the bow echo was small.Strong winds and induced damage on the surface are considered to have been caused by the mid-level rear-inflow jet and intense convective downdrafts.

天气雷达空中生态监测

每年数以亿计的昆虫、候鸟和蝙蝠等空中动物在全球范围内远距离迁飞,显著影响空中生态系统的结构和功能。随着全球气候和环境的不断变化,空中生态环境也在迅速恶化,对人类的生命和健康构成了严重威胁。为了保护空中生态环境,首要任务是对其进行有效监测。天气雷达具有大范围、全天时、全天候监测能力,可覆盖百千千米空间尺度,是监测和研究宏观空中生态的有效工具。十余年来,天气雷达空中生态监测领域取得了突破性的进展,极大提高了对空中迁飞动物的探测能力,可有效弥补传统监测手段存在的覆盖范围小和时间分辨率低等技术局限。天气雷达已在全球范围内的生物多样性保护、迁飞性虫害监测、疫源疫病传播防控等重大空中生态议题中发挥着日益重要的作用,有效填补了宏观生态研究空白。目前,美国、欧洲和中国分别建立了全球三大天气雷达网,为空中生态研究提供了重要的基础设施和数据支持。本文详细阐述了当前全球天气雷达空中生态监测的科学研究进展,以及通过天气雷达揭示的空中动物迁飞科学规律。然后,对天气雷达生态监测中迁飞动物的电磁散射建模、雷达回波提取、群体参数反演、雷达观测验证实验,以及宏观迁飞模型等关键技术进行了梳理和总结,并详细介绍了中国天气雷达空中生态监测的进展情况。最后,对未来发展方向进行了展望,指出针对中国当前农业迁飞性害虫和鸟类宏观迁飞等重大监测需求,需要在扩维探测、多参反演、多源预测等方面深入研究和探索,以进一步解读重大动物迁飞事件、分析宏观迁飞规律、揭示空中生态对气候变化的响应。

空天目标雷达智能识别仿真系统设计与实现

雷达智能识别方法训练、测试与评估通常需要构建大样本的数据集,如何快速获得有效的数据集是亟待解决的问题。分析了空天目标雷达智能识别系统应用场景,并结合应用需求给出了空天目标雷达智能识别仿真系统建设思路;采用组件化、模块化设计方法,设计空天目标雷达智能识别仿真系统框架,构建空天目标智能识别仿真系统,快速准确模拟空天目标电磁特性和动态回波,具有可扩展、可重构能力。仿真结果和初步应用表明,该系统可用于不同应用场景下空天目标雷达智能识别方法的测试与评估。

基于自注意力和门控循环神经网络的雷达回波外推算法研究

为提升现有神经网络对雷达回波序列的时、空特征提取能力,建立外推性能更优的时、空序列预测模型,开展雷达回波外推算法改进研究。基于深圳市气象局与中国香港天文台共同建立的雷达回波数据集,在数据处理层面,通过改进对雷达回波图像序列归一化的方法,提升了常用的5种时、空序列预测模型对强回波的预测水平;在模型算法层面,将两个联立的自注意力结构引入ST-LSTM结构,组成新的循环门控单元,并将这些循环门控单元进行堆叠,建立ST-SARNN模型。选用CSI和POD作为精度评价指标,进行模型对比分析得到:(1)改进的归一化方法提升了近几年内常用的5种时、空序列预测模型对强回波的预测水平。(2)加入自注意力的ST-SARNN模型对雷达回波的预测性能显著优于ConvLSTM、PredRNN和MIM等模型。改进的归一化方法能改变样本数据分布,并在一定程度上提升模型外推性能;自注意力结构能够有效挖掘雷达回波序列的时、空特征,进而改进神经网络的外推表现。

Analysis of a Type II Snowstorm Process in the Early Spring of 2022 in Ulanqab

Based on the observation data of meteorological stations,Doppler radar observation data of Ulanqab City,and ERA-5 reanalysis data,a snowstorm process in Ulanqab City from March 17 to 18,2022 was analyzed.The results show that this was a type II snowstorm process generated under the joint influence of upper trough and ground low inverted trough and frontal cyclone.The main period of snowfall can be divided into two time stages,and the total snowfall was more in the south and less in the north,which was consistent with that of average specific humidity field.Water vapor conditions provided by strong water vapor transport and convergence,strong upward movement shown by large vertical velocity field,and the suction action of high-and low-layer divergence and convergence were the reasons for the hourly heavy snowfall on the 18 th.During the process,radar echoes were mainly sheet-shaped,and composite reflectivity was 15-25 dBZ in most areas.The zero speed line in the first period was positively"S"-shaped,and there was warm advection and southwest wind.On the morning of the 18 th,after the cold front transited the city,Ulanqab City was gradually controlled by northwest wind,and the snow tended to end.

基于分层生成对抗网络的短临降水预报方法研究

本文使用深度学习方法中的生成对抗网络(GAN)来提升短临降水预报的准确率,提出了一个基于历史雷达回波图序列预测未来雷达回波图序列的分层生成对抗网络(HGAN)方法。HGAN方法由全局生成器和局部鉴别器两部分组成,全局生成器以多子网的层次结构构建,采用上采样过程训练模型,捕捉雷达回波的演变趋势,有利于生成清晰的未来雷达回波图。局部鉴别器根据局部区域将预测的雷达回波图与观测的雷达回波图区分开,并引入缓冲区机制,保存历史预测序列,使最终预测的结果更加符合时序性。两者以对抗的方式加以训练,得到的模型能够生成足够清晰且接近真实的未来雷达回波序列,对于回波强度极值和范围的刻画更为准确。对HGAN和GAN进行测试集检验及个例分析,分析结果验证了HGAN对雷达回波预测的有效性。同时在检验反射率阈值相同的情况下,HGAN的临界成功指数命中率高于GAN,而虚警率低于GAN,且在相同预测时长下,HGAN结构相似性指数(SSIM)优于GAN。