基于3种机器学习算法的台风频数预测-海洋预报

基于3种机器学习算法的台风频数预测

单位：1. 广西民族大学电子信息学院, 广西南宁 530000;
2. 广西气候中心, 广西南宁 530022

分类号：P444

出版年·卷·期（页码）：2023·40·第五期（1-9）

摘要：

为了提高影响广西台风频数的年度预测准确率，利用中国气象局上海台风研究所提供的1951—2020年影响广西的台风样本数据、国家气候中心提供的88项大气环流特征量和26项海温指数资料，使用相关方法找出高影响因子。针对影响台风物理因素的复杂性，为了获取更综合的预测因子信息，使用随机森林对影响因子进行二次筛选，建立基于随机森林、支持向量回归和循环门单元（GRU）3种机器学习算法的影响广西台风频数气候预测模型。实验结果表明：使用随机森林二次筛选得到的因子的建模预测效果明显提高，机器学习算法预测效果整体高于岭回归方法，其中GRU预测效果最好，绝对误差较岭回归方法减少10.30%，其次为随机森林和支持向量回归，误差分别减少9.44%和7.47%。

In order to improve the prediction accuracy of annual number of typhoons affecting Guangxi, this paper uses related methods to find high impact factor based on the sample data of typhoons affecting Guangxi from 1951 to 2020 provided by Shanghai Typhoon Institute of China Meteorological Administration, the 88 atmospheric circulation feature quantities and 26 SST index data provided by the National Climate Center. In view of the complexity of physical factors in typhoon number forecasting, in order to obtain more comprehensive factor information, the random forest is used to screen the factors, and a prediction model for annual number of typhoons affecting Guangxi utilizing three machine learning algorithms, i. e. Random Forest, Support Vector Regression and Gate Recurrent Unit （GRU）, is established. The results show that the prediction ability of using factors selected by Random Forest screening is significantly improved, and the prediction ability of using machine learning algorithms is higher than that of Ridge Regression method. Among them, GRU has the best prediction, and the absolute error is reduced by 10.30% compared with Ridge Regression method, followed by Random Forest and Support Vector Regression, with errors reduced by 9.44% and 7.47%, respectively.

参考文献：

[1] CHAND S S, WALSH K J E. Forecasting tropical cyclone formation in the fiji region: a probit regression approach using Bayesian fitting[J]. Weather and Forecasting, 2011, 26(2): 150-165.
[2] CHU P S, ZHAO X. A Bayesian regression approach for predicting seasonal tropical cyclone activity over the central north pacific[J]. Journal of Climate, 2007, 20(15): 4002-4013.
[3] MCDONNELL K A, HOLBROOK N J. A Poisson regression model of tropical cyclogenesis for the Australian-southwest Pacific Ocean region[J]. Weather and Forecasting, 2004, 19(2): 440-455.
[4] MCDONNELL K A, HOLBROOK N J. A Poisson regression model approach to predicting tropical cyclogenesis in the Australian / southwest Pacific Ocean region using the SOI and saturated equivalent potential temperature gradient as predictors[J]. Geophysical Research Letters, 2004, 31(20): L20110.
[5] KE F. New predictors and a new prediction model for the typhoon frequency over western North Pacific[J]. Science in China Series D: Earth Sciences, 2007, 50(9): 1417-1423.
[6] MESTRE O, HALLEGATTE S. Predictors of tropical cyclone numbers and extreme hurricane intensities over the north Atlantic using generalized additive and linear models[J]. Journal of Climate, 2009, 22(3): 633-648.
[7] LI X, YANG S, WANG H, et al. A dynamical-statistical forecast model for the annual frequency of western Pacific tropical cyclones based on the NCEP Climate Forecast System version 2[J]. Journal of Geophysical Research: Atmospheres, 2013, 118(21): 12061- 12074.
[8] WAHIDUZZAMAN M, CHEUNG K, LUO J J, et al. Impact assessment of Indian Ocean Dipole on the North Indian Ocean tropical cyclone prediction using a Statistical model[J]. Climate Dynamics, 2022, 58(3): 1275-1292.
[9] NONG J F, JIN L. Application of support vector machine to predict precipitation[C]//20087th World Congress on Intelligent Control and Automation. Chongqing: IEEE, 2008: 8975-8980.
[10] SRIVASTAVA S, ANAND N, SHARMA S, et al. Monthly rainfall prediction using various machine learning algorithms for early warning of landslide occurrence[C]//2020 International Conference for Emerging Technology. Belgaum: IEEE, 2020: 1-7.
[11] 甄亿位, 郝敏, 陆宝宏, 等. 基于随机森林的中长期降水量预测模型研究[J]. 水电能源科学, 2015, 33(6): 6-10. ZHEN Y W, HAO M, LU B H, et al. Research of medium and long term precipitation forecasting model based on random forest [J]. Water Resources and Power, 2015, 33(6): 6-10.
[12] 覃卫坚, 陆虹, 黄志, 等. 粒子群—神经网络法在广西寒露风日数预报中的应用[J]. 气象与环境学报, 2015, 31(6): 158-162. QIN W J, LU H, HUANG Z, et al. Application of forecasting cold dew wind day based on PSO-Fuzzy Neural Network in Guangxi province[J]. Journal of Meteorology and Environment, 2015, 31(6): 158-162.
[13] 覃卫坚, 李耀先, 陈思蓉, 等. 粒子群-神经网络在华南夏季降水短期气候预测中应用研究[J]. 气象研究与应用, 2015, 36(2): 1- 7. QIN W J, LI Y X, CHEN S R, et al. Application on the prediction of the summer precipitation in South China basing on PSOArtificial Neutral Network[J]. Journal of Meteorological Research and Application, 2015, 36(2): 1-7.
[14] GHAMARIADYAN M, IMTEAZ M A. Prediction of seasonal rainfall with one-year lead time using climate indices: a wavelet neural network scheme[J]. Water Resources Management, 2021, 35(15): 5347-5365.
[15] 罗芳琼, 吴建生, 金龙. 基于最小二乘支持向量机集成的降水预报模型[J]. 热带气象学报, 2011, 27(4): 577-584. LUO F Q, WU J S, JIN L. Rainfall forecasting model based on least square support vector machine regression ensemble[J]. Journal of Tropical Meteorology, 2011, 27(4): 577-584.
[16] CHEN R, WANG X, ZHANG W M, et al. A hybrid CNN-LSTM model for typhoon formation forecasting[J]. Geoinformatica, 2019, 23(3): 375-396.
[17] 高珊, 刘峻. 基于LSTM的台风强度预测模型分析[J]. 信息与电脑, 2021, 33(11): 30-32. GAO S, LIU J. Typhoon intensity prediction model analysis based on LSTM[J]. China Computer & Communication, 2021, 33(11): 30-32.
[18] 徐光宁. 基于深度学习的台风路径与强度预测方法研究[D]. 哈尔滨: 哈尔滨工业大学, 2020. XU G N. Research on typhoon movement and intensity forecasting based on deep learning[D]. Harbin: Harbin Institute of Technology, 2020.
[19] HAGHROOSTA T, ISMAIL W R. Comparing typhoon intensity prediction with two different artificial intelligence models[J]. Evolving Systems, 2015, 6(3): 177-185.
[20] GAO S, ZHAO P, PAN B, et al. A nowcasting model for the prediction of typhoon tracks based on a long short term memory neural network[J]. Acta Oceanologica Sinica, 2018, 37(5): 8-12.
[21] SONG H J, HUH S H, KIM J H, et al. Typhoon track prediction by a support vector machine using data reduction methods[C]// Proceedings of International Conference on Computational and Information Science. Xi'an: Springer, 2005: 503-511.
[22] LIU H X, ZHANG D L, CHEN J W, et al. Prediction of tropical cyclone frequency with a wavelet neural network model incorporating natural orthogonal expansion and combined weights [J]. Natural Hazards, 2013, 65(1): 63-78.
[23] TAN J K, LIU H X, LI M Y, et al. A prediction scheme of tropical cyclone frequency based on lasso and random forest[J]. Theoretical and Applied Climatology, 2018, 133(3): 973-983.
[24] 覃卫坚, 黄志, 李耀先. 基于海温、雪盖的影响广西热带气旋频数的气候预测模型研究[J]. 气象研究与应用, 2013, 34(3): 1- 5, 32. QIN W J, HUANG Z, LI Y X. A short-term climatic forecast model for the frequency of tropical cyclone affecting Guangxi based on SST and snow data[J]. Journal of Meteorological Research and Application, 2013, 34(3): 1-5, 32.
[25] ÜLKER E D, ÜLKER S. Modelling the currency exchange rates using support vector regression[C]//Science and Information Conference. London: Springer, 2020: 326-333.
[26] BREIMAN L. Random forests[J]. Machine Learning, 2001, 45(1): 5-32.
[27] MEENA L, CHAURASIYA V K, PUROHIT N, et al. Comparison of SVM and random forest methods for online signature verification[C]//Proceedings of the 12th International Conference on Intelligent Human Computer Interaction. Daegu: Springer, 2021: 288-299.
[28] 李光华, 李俊清, 张亮, 等. 一种融合蚁群算法和随机森林的特征选择方法[J]. 计算机科学, 2019, 46(S2): 212-215. LI G H, LI J Q, ZHANG L, et al. Feature selection method based on ant colony optimization and random forest[J]. Computer Science, 2019, 46(S2): 212-215.
[29] SAHA S, SINGH N, MOHAN B R, et al. A combined model of ARIMA-GRU to forecast stock price[C]//Proceedings of the International Conference on Paradigms of Computing, Communication and Data Sciences. Singapore: Springer, 2021: 987-998.

服务与反馈：

【文章下载】【发表评论】【查看评论】【加入收藏】