WANG Chengxing.Convective Initiation Identification based on Machine Learning[J].Journal of Chengdu University of Information Technology,2025,40(04):493-502.[doi:10.16836/j.cnki.jcuit.2025.04.014]
基于机器学习的初生对流识别
- Title:
- Convective Initiation Identification based on Machine Learning
- 文章编号:
- 2096-1618(2025)04-0493-10
- 分类号:
- TN957
- 文献标志码:
- A
- 摘要:
- 通过识别对流初生过程,能有效监测对流发展情况,准确预报强对流,对中国预防对流灾害具有重要意义。传统对流初生识别算法虽然命中率较高(probability of detection,POD),但误报率(false-alarm ratio,FAR)也往往较高,且对预报人员的气象背景及预报经验有较高要求。利用2018-2019年风云4号A星(记为FY4-A,下同)的先进静止轨道辐射成像仪(advanced geosynchronous radiometer for imaging,AGRI)数据以及山东省的天气雷达数据,通过AGRI单通道、组合通道差以及时间变化率计算得到特征参量,结合雷达反射率因子大于35 dBZ的条件,建立客观的初生对流单体标记方法。通过使用决策树(decision tree,DT)、支持向量机(support victor machine,SVM)和人工神经网络(aArtificial neural network,ANN)3种机器学习方法,对山东省内对流初生过程进行建模与评估。经计算,SVM的效果最好,POD为91.2%,FAR为10.1%,临界成功指数(critical success index,CSI)为82.7%; 其次是DT,POD为84.5,FAR为10.9%,CSI为76.6%; ANN的评分最差,POD为75.8%,FAR为16.7%,CSI为65.8%。实验结果表明采用机器学习方法可以有效识别对流初生,在强对流预报中能发挥很好的作用。
- Abstract:
- Through the identification of the initial development process of convection,it is possible to effectively monitor the development of convection and make accurate predictions of severe convective weather.This is of great significance for preventing convective disasters in China.Although traditional algorithms for identifying the initial development of convection often have a high probability of detection(POD),thefalse-alarm ratio(FAR)is also frequently high.Additionally,traditional methods require a high level of meteorological background and forecasting experience from forecasters.In this study,data from the Advanced Geosynchronous Radiometer for Imaging(AGRI)on Fengyun-4A satellite from 2018 to 2019,as well as weather radar data from Shandong Province,were utilized.Characteristic parameters were calculated using AGRI single-channel,combination channel differences,and time rate of change.An objective method for marking the initial convection cells was established by combining these parameters with the condition of radar reflectivity factor greater than 35 dBZ.Three machine learning methods,namely Decision Tree(DT),Support Vector Machine(SVM),and Artificial Neural Network(ANN)were employed to model and evaluate the initial convection process in Shandong Province.The calculations showed that SVM achieved the best performance,with a POD of 91.2%,a FAR of 10.1%,and a Critical Success Index(CSI)of 82.7%.Secondly,DT showed a POD of 84.5%,a FAR of 10.9%,and a CSI of76.6%.Lastly,ANN performed the worst,with a POD of 75.8%,a FAR of 16.7%,and a CSI of 65.8%.The evaluation results indicate that machine learning methods can effectively identify convection and play a significant role in meteorological forecasting.
参考文献/References:
[1] Wagner T J,Feltz W F,Ackerman S A.The Temporal Evolution of Convective Indices in Storm-Producing Environments[J]. Weather and Forecasting,2008,23(5):786-794.
[2] Houze R A,Rasmussen K L,Zuluaga M D,et al.The Variable Nature of Convection in the Tropics and Subtropics:A Legacy of 16 Years of the Tropical Rainfall Measuring Mission Satellite[J]. Reviews of Geophysics,2015,53(3):994-1021.
[3] Reif D W,Bluestein H B.A 20-Year Climatology of Nocturnal Convection Initiation over the Central and Southern Great Plains during the Warm Season[J]. Monthly Weather Review,2017,145(5):1615-1639.
[4] Reif D W,Bluestein H B.Initiation Mechanisms of Nocturnal Convection without Nearby Surface Boundaries over the Central and Southern Great Plains during the Warm Season[J]. Monthly Weather Review,2018,146(9):3053-3078.
[5] Wilson J W,Trier S B,Reif D W,et al.Nocturnal Elevated Convection Initiation of the PECAN 4 July Hailstorm[J]. Monthly Weather Review,2018,146(1):243-262.
[6] 覃丹宇,方宗义.利用静止气象卫星监测初生对流的研究进展[J]. 气象,2014,40(1):7-17.
[7] 茂宾,白洁,周著华,等.基于 FY-2G 静止气象卫星资料的对流初生预警[J]. 气象水文海洋仪器,2018,35(1):8-13.
[8] 应急管理部官网.应急管理部发布2022年前三季度全国自然灾害情况 Available online[EB/OL]. https://www.ndrcc.org.cn/zqtj/27127.jhtml.
[9] Mecikalski J R,Bedka K M.Forecasting Convective Initiation by Monitoring the Evolution of Moving Cumulus in Daytime GOES Imagery[J]. Monthly Weather Review,2006,134(1):49-78.
[10] Delgado G,Machado L A T,Angelis C F,et al.Basis for a Rainfall Estimation Technique Using IR-VIS Cloud Classification and Parameters over the Life Cycle of Mesoscale Convective Systems[J]. Journal of Applied Meteorology and Climatology,2008,47(5):1500-1517.
[11] 王晓芳,崔春光.暴雨中尺度对流系统研究的若干进展[J]. 暴雨灾害,2011,30(2):97-106.
[12] Browning K A,Atlas D.Initiation of Precipitation in Vigorous Convective Clouds[J]. Journal of the Atmospheric Sciences,1965,22(6):678-683.
[13] Marshall J S,Radhakant S.Radar Precipitation Maps as Lightning Indicators[J]. Journal of Applied Meteorology,1978,17(2):206-212.
[14] Wilson J W,Schreiber W E.Initiation of Convective Storms at Radar-Observed Boundary-Layer Convergence Lines[J]. Monthly Weather Review,1986,114(12):2516-2536.
[15] Wilson J W,Foote G B,Crook N A,et al.The Role of Boundary-Layer Convergence Zones and Horizontal Rolls in the Initiation of Thunderstorms:A Case Study[J]. Monthly Weather Review,1992,120(9):1785-1815.
[16] Wilson J W,Mueller C K.Nowcasts of Thunderstorm Initiation and Evolution[J]. Weather and Forecasting,1993,8(1):113-131.
[17] Mueller C,Saxen T,Roberts R,et al.NCAR Auto-Nowcast System[J]. Weather and Forecasting,2003,18(4):545-561.
[18] Roberts R D,Rutledge S.Nowcasting Storm Initiation and Growth Using GOES-8 and WSR-88D Data[J]. Weather and Forecasting,2003,18(4):562-584.
[19] Roberts R D,Anderson A R S,Nelson E,et al.Impacts of Forecaster Involvement on Convective Storm Initiation and Evolution Nowcasting[J]. Weather and Forecasting,2012,27(5):1061-1089.
[20] Walker J R,MacKenzie W M,Mecikalski J R,et al.An Enhanced GeostationarySatellite-Based Convective Initiation Algorithm for 0-2-h Nowcasting with Object Tracking[J]. Journal of Applied Meteorology and Climatology,2012,51(11):1931-1949.
[21] Mecikalski J R,Feltz W F,Murray J J,et al.Aviation Applications for Satellite-Based Observations of Cloud Properties,Convection Initiation,In-Flight Icing,Turbulence,and Volcanic Ash[J]. Bulletin of the American Meteorological Society,2007,88(10):1589-1607.
[22] Sieglaff J M,Cronce L M,Feltz W F,et al.Nowcasting Convective Storm Initiation Using Satellite-Based Box-Averaged Cloud-Top Cooling and Cloud-Type Trends[J]. Journal of Applied Meteorology and Climatology,2011,50(1):110-126.
[23] Lee J G,Min K H,Park H,et al.Improvement of the Rapid-Development Thunderstorm(RDT)Algorithm for Use with the GK2A Satellite[J]. Asia-Pacific Journal of Atmospheric Sciences,2020,56(2):307-319.
[24] Algorithm Theoretical Basis Document for “ Rapid Development[M]. METEO-FRANCE:Paris,2013.
[25] 姚秀萍,曹晓敏.大气对流初生的研究进展与展望[J]. 大气科学学报,2023,46(6):131-13.
[26] Tebbi M A,Haddad B.Artificial Intelligence Systems for Rainy Areas Detection and Convective Cells’ Delineation for the South Shore of Mediterranean Sea during Day and Nighttime Using MSG Satellite Images[J]. Atmospheric Research,2016,178380-392.
[27] Kim M,Im J,Park H,et al.Detection of Tropical Overshooting Cloud Tops Using Himawari-8 Imagery[J]. Remote Sensing 2017,9(7):685.
[28] Moraux A,Dewitte S,Cornelis B,et al.A Deep Learning Multimodal Method for Precipitation Estimation[J]. Remote Sensing,2021,13(16):3278.
[29] Beusch L,Foresti L,Gabella M,et al.Satellite-Based Rainfall Retrieval:From Generalized Linear Models to Artificial Neural Networks[J]. Remote Sensing,2018,10(6):939.
[30] 李扬.基于深度学习的强对流临近预报方法研究[D]. 南京:南京信息工程大学,2023.
[31] 李冰村,唐晓文,何建新,等.基于机器学习的冰雹天气识别研究[J]. 气象科学,2022,42(5):581-590.
[32] Pedregosa F,Varoquaux G,Gramfort A,et al.Scikit-Learn:Machine Learning in Python[J]. Journal of Machine Learning Research,2011,12(January):2825-2830.
[33] Abiodun O I,Jantan A,Omolara A E,et al.Comprehensive Review of Artificial Neural Network Applications to Pattern Recognition[C]. IEEE Access,2019:7158820-158846.
[34] Hering A M,Morel C,Galli G,et al.Nowcasting Thunderstorms in the Alpine Region Using a Radar Based Adaptive Thresholding Scheme[J]. Proceedings of ERAD,2004,1(6):206-211.
[35] Zhuge X,Zou X.Summertime Convective Initiation Nowcasting over Southeastern China Based on Advanced Himawari Imager Observations[J]. Journal of the Meteorological Society of Japan.Ser.II,2018,96(4):337-353.
[36] Yang J,Zhang Z,Wei C,et al.Introducing the New Generation of Chinese Geostationary Weather Satellites,Fengyun-4[J]. Bulletin of the American Meteorological Society,2017,98(8):1637-1658.
[37] Strabala K I,Ackerman S A,Menzel W P.Cloud Properties Inferred from 8-12-Μm Data[J]. Journal of Applied Meteorology,1994,33(2):212-229.
[38] Baum B A,Kratz D P,Yang P,et al.Remote Sensing of Cloud Properties Using MODIS Airborne Simulator Imagery during SUCCESS:1.Data and Models[J]. Journal of Geophysical Research:Atmospheres,2000,105(D9):11767-11780.
[39] Zou X,Zhuge X,Weng F.Characterization of Bias of Advanced Himawari Imager Infrared Observations from NWP Background Simulations Using CRTM and RTTOV[J]. Journal of Atmospheric and Oceanic Technology,2016,33(12):2553-2567.
[40] Ackerman S A.Global Satellite Observations of Negative Brightness Temperature Differences between 11 and 6.7/Лт[J]. Journal of the Atmospheric Sciences,1996,53(19).
[41] Matthee R,Mecikalski J R.Geostationary Infrared Methods for Detecting Lightning-Producing Cumulonimbus Clouds[J]. Journal of Geophysical Research:Atmospheres,2013,118(12):6580-6592.
相似文献/References:
[1]李宝林,周 坤,李仕伟.一种基于M-Bisearch的最大频繁项集挖掘算法研究[J].成都信息工程大学学报,2016,(05):463.
LI Bao-lin,ZHOU Kun,LI Shi-wei.Research on Mining Algorithm of Maximal Frequent Itemsets based on M-blsearch[J].Journal of Chengdu University of Information Technology,2016,(04):463.
[2]唐雨奇,李则辰,杨东东,等.基于机器学习的面部运动神经传导检查数据的研究及应用[J].成都信息工程大学学报,2020,35(05):519.[doi:10.16836/j.cnki.jcuit.2020.05.007]
TANG Yuqi,LI Zechen,YANG Dongdong,et al.Research and Application of Facial Motor Nerve Conduction Examination Data based on Machine Learning[J].Journal of Chengdu University of Information Technology,2020,35(04):519.[doi:10.16836/j.cnki.jcuit.2020.05.007]
[3]毛开银,赵长名,何 嘉.基于XGBoost的10 m风速订正研究[J].成都信息工程大学学报,2020,35(06):604.[doi:10.16836/j.cnki.jcuit.2020.06.004]
MAO Kaiyin,ZHAO Changming,HE Jia.A Research for 10 m Wind Speed Prediction based on XGBoost[J].Journal of Chengdu University of Information Technology,2020,35(04):604.[doi:10.16836/j.cnki.jcuit.2020.06.004]
[4]李孝涌,陈科艺,李熙晨.卷积神经网络在ENSO预报中的应用[J].成都信息工程大学学报,2022,37(01):81.[doi:10.16836/j.cnki.jcuit.2022.01.014]
LI Xiaoyong,CHEN Keyi,LI Xichen.Application of Convolutional Neural Network in ENSO Prediction[J].Journal of Chengdu University of Information Technology,2022,37(04):81.[doi:10.16836/j.cnki.jcuit.2022.01.014]
备注/Memo
收稿日期:2024-02-19