Optimization of typhoon track prediction based on neural network ensemble prediction
ZHOU Xiaotian1,2, ZHANG Feng1,2, DU Zhenhong1,2, LIU Renyi1,2
1.Zhejiang Provincial Key Lab of GIS, Zhejiang University, Hangzhou 310028, China 2.Department of Geographic Information Science, Zhejiang University, Hangzhou 310027, China
Abstract：An optimization model of typhoon track prediction based on ensemble prediction of neural network is presented in the paper to overcome the shortcomings of the current typhoon track prediction models, such as inadequate prediction accuracy and high coupling of forecast times. The proposed model uses the idea of ensemble prediction of mixed model and multi-layer feedforward training mechanism of reverse propagation to fully mine data characteristics, and solves the solidification problem of single ensemble prediction model as well as the randomness problem of the single neural network forecasting model, which provides a new idea for the combination of the existing typhoon numerical forecasting method and artificial intelligence technology. The typhoon activity in the Northwest Pacific and South China Sea in 2018 is taken as a sample for comparative experiments. The results show that the forecasting accuracy of the model regarding a period of 60 hours has been improved , which reflects the practical application value of the model to a certain extent.
1 XUB L, LIUC D, WUC D, et al. Typhoon detection of radar improved using ultra low elevation angle[J]. Plateau Meteorology, 2012, 31(1): 251-257. 2 俞兆文，刘健文，钟中，等. 云雨条件下AMSR2微波成像资料同化试验及其在台风预报中的应用[J]. 气象科学，2018, 38(2): 203-211. YUZ W, LIUJ W, ZHONGZ, et al. Assimilation experiment of AMSR2 microwave imaging data under cloudy and rainy condition and its application on the forecast of a typhoon process[J]. Journal of the Meteorological Sciences, 2018, 38(2): 203-211. 3 袁杰颖，陈永平，潘毅，等. 台风路径集合化预报方法的优化[J]. 海洋预报，2017, 34(2): 37-42.DOI:10.1201/9781315116242-26 YUANJ Y, CHENY P, PANY, et al. Improvement of ensemble forecast of typhoon track in the Northwestern Pacific[J]. Marine Forecasts, 2017,34(2):37-42. DOI:10.1201/9781315116242-26 4 WANGC X. Ensemble prediction experiments of typhoon track based on the stochastic total tendency perturbation[J]. Journal of Tropical Meteorology, 2016, 22(3): 305-317. 5 HUOZ H, DUANW S. The application of the orthogonal conditional nonlinear optimal perturbations method to typhoon track ensemble forecasts[J]. Science China Earth Sciences, 2019, 62(2): 376-388.DOI:10.1007/s11430-018-9248-9 6 陈子通，戴光丰，罗秋红，等. 模式动力过程与物理过程耦合及其对台风预报的影响研究[J]. 热带气象学报，2016, 32(1): 1-8. DOI:10.16032/j.issn.1004-4965.2016.01.001 CHENZ T, DAIG F, LUOQ H, et al. Study on the coupling of model dynamics and physical processes and its influence on the forecast of typhoons[J]. Journal of Tropical Meteorology, 2016, 32(1): 1-8.DOI:10.16032/j.issn.1004-4965.2016.01.001 7 黄文锋，周焕林，孙建鹏. 应用台风风场经验模型的台风极值风速预测[J]. 哈尔滨工业大学学报，2016, 48(2): 142-146. HUANGW F, ZHOUH L, SUNJ P, et al. Prediction typhoon design wind speed with empirical typhoon wind field model[J]. Journal of Harbin Institute of Technology, 2016, 48(2): 142-146. 8 CHUNY , MINJ Z, LIUZ Q, et al. The impact of AMSR2 radiance data assimilation on the analysis and forecast of typhoon Son-Tinh[J]. Chinese Journal of Atmospheric Sciences, 2017,41(2):372-384. 9 王向阳，颜琼丹，鄢志波. 热带气旋路径集成预报在业务中使用[C]//中国气象学会年会. 广州：中国气象学会， 2007:127-130. WANGX Y, YANQ D, YANZ B. Application of integrated tropical cyclone track forecasting in operations [C]//Annual Conference China Meteorological Society. Guangzhou：China Meteorological Society ，2007:127-130. 10 黄小燕，金龙. 基于主成分分析的人工智能台风路径预报模型[J]. 大气科学，2013, 37(5): 1154-1164. HUANGX Y, JINL, et al. An artificial intelligence prediction model based on principal component analysis for typhoon tracks[J]. Chinese Journal of Atmospheric Sciences, 2013, 37(5): 1154-1164. 11 邵利民，傅刚，曹祥村，等. BP神经网络在台风路径预报中的应用[J]. 自然灾害学报，2009, 18(6): 104-111. DOI:10.13577/j.jnd.2009.0618 SHAOL M, FUG, CAOX C, et al. Application of BP neural network in typhoon track forecasting[J]. Journal of Natural Disasters, 2009, 18(6): 104-111. DOI:10.13577/j.jnd.2009.0618 12 谭燕, 梁旭东, 黄伟. 集合预报技术在台风路径预报中的应用[C]//海峡两岸气象科学技术研讨会. 北京：中国气象学会，2008: 37-45. TANY, LIANGQ D, HUANGW. Application of ensemble forecasting technology in typhoon track forecasting[C]//Seminar on Meteorological Science and Technology across the Taiwan Strait. Beijing：Chinese Meteorological Society 2008: 37-45. 13 EPSTEINE S. Stochastic dynamic prediction[J]. Tellus, 1969, 21(6): 739-759 .DOI:10.3402/tellusa.v21i6.10143 14 LEITHC E. Theoretical skill of Monte Carlo forecasts[J]. Monthly Weather Review, 1974, 102(6): 409-418. DOI:10.1175/1520-0493(1974)102<0409:tsomcf>2.0.co;2 15 MOLTENIF, BUIZZAR, PALMERT N, et al. The ECMWF ensemble prediction system: Methodology and validation[J]. Quarterly Journal of the Royal Meteorological Society, 2010, 122(529): 73-119. DOI:10.1002/qj.49712252905 16 TOTHZ, KALNAYE. Ensemble forecasting at NCEP and the breeding method[J]. Monthly Weather Review, 1997, 125(12): 3297-3319.DOI:10.1175/1520-0493(1997)125〈3297:efanat〉2.0.co;2 17 TOTHZ. Meeting summary: Ensemble forecasting in WRF[J]. Bulletin of the American Meteorological Society, 2010, 82(4): 695-698.DOI:10.1175/1520-0477(2001)082〈0695:msefiw〉2.3.co;2 18 KRISHNAMURTIT N, KISHTAWALC M, LAROWT E, et al. Improved weather and seasonal climate forecasts from multi-model superensemble[J]. Science, 1999, 285(5433): 1548-1550. 19 ATGERF. Spatial and interannual variability of the reliability of ensemble-based probabilistic forecasts: consequences for calibration[J]. Monthly Weather Review, 2003, 131(8): 1509-1523.DOI:10.1175//1520-0493(2003)131<1509:saivot>2.0.co;2 20 杨学胜. 业务集合预报系统的现状及展望[J]. 气象, 2001, 27(6): 3-9. YANGX S. Current situation and prospect of operational ensemble forecasting system[J]. Meteorological Monthly, 2001, 27(6): 3-9. 21 HANS , LIUX Y, MAOH Z, et al. EIE: Efficient inference engine on compressed deep neural network[J]. ACM SIGARCH Computer Architecture News, 2016, 44(3): 243-254. DOI:10.1109/isca.2016.30 22 SUNW , XUY. Using a back propagation neural network based on improved particle swarm optimization to study the influential factors of carbon dioxide emissions in Hebei province, China[J]. Journal of Cleaner Production, 2016, 112(part 2): 1282-1291. DOI:10.1016/j.jclepro.2015.04.097 23 张少华. 基于深度卷积神经网络的人脸基准点定位研究[D]. 武汉：华中科技大学，2016. ZHANGS H. Research on Face Baseline Location Based on Deep Convolution Neural Network[D]. Wuhan: Huazhong University of Science and Technology, 2016. 24 IOFFES, SZEGEDYC. Batch normalization: Accelerating deep network training by reducing internal covariate shift[C]// Proceedings of the 32nd International Conference on Machine Learning. Lille： ACM, 2015, 37:448-456. 25 KIMY , JERNITEY, SONGTAGD , et al. Character-aware neural language models[C]//30th AAAI Conference on Artificial Intelligence. Phoenix: AAAI Press, 2016. 26 RAYNAUDL, BOUTTIERF. Comparison of initial perturbation methods for ensemble prediction at convective scale[J]. Quarterly Journal of the Royal Meteorological Society, 2016, 142(695): 854-866.DOI:10.1002/qj.2686 27 王皘, 钱传海, 张玲. 2017年西北太平洋和南海台风活动概述[J]. 海洋气象学报, 2018, 38(2): 1-11.DOI:10.19513/j.cnki.issn2096-3599.2018.02.001 WANGQ, QIANC H, ZHANGL, et al. Summary of typhoon activities in the northwest pacific and south china sea in 2017[J]. Journal of Marine Meteorology, 2018, 38(2): 1-11.DOI:10.19513/j.cnki.issn2096-3599.2018.02.001
2020, Vol. 47 
