Sing-Wen ChenInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, Taiwan, Joyce JuangCentral Weather Administration, Taiwan, Charlotte WangInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, TaiwanMaster Program of Public Health, College of Public Health, National Taiwan University, Taiwan, Hui-Ling ChangCentral Weather Administration, Taiwan, Jing-Shan HongCentral Weather Administration, Taiwan, Chuhsing Kate HsiaoInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, TaiwanMaster Program of Public Health, College of Public Health, National Taiwan University, Taiwan
{"title":"基于卷积神经网络的热带气旋降水预报数据增量集合后处理技术","authors":"Sing-Wen ChenInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, Taiwan, Joyce JuangCentral Weather Administration, Taiwan, Charlotte WangInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, TaiwanMaster Program of Public Health, College of Public Health, National Taiwan University, Taiwan, Hui-Ling ChangCentral Weather Administration, Taiwan, Jing-Shan HongCentral Weather Administration, Taiwan, Chuhsing Kate HsiaoInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, TaiwanMaster Program of Public Health, College of Public Health, National Taiwan University, Taiwan","doi":"arxiv-2409.09607","DOIUrl":null,"url":null,"abstract":"Heavy precipitation from tropical cyclones (TCs) may result in disasters,\nsuch as floods and landslides, leading to substantial economic damage and loss\nof life. Prediction of TC precipitation based on ensemble post-processing\nprocedures using machine learning (ML) approaches has received considerable\nattention for its flexibility in modeling and its computational power in\nmanaging complex models. However, when applying ML techniques to TC\nprecipitation for a specific area, the available observation data are typically\ninsufficient for comprehensive training, validation, and testing of the ML\nmodel, primarily due to the rapid movement of TCs. We propose to use the\nconvolutional neural network (CNN) as a deep ML model to leverage the spatial\ninformation of precipitation. The proposed model has three distinct features\nthat differentiate it from traditional CNNs applied in meteorology. First, it\nutilizes data augmentation to alleviate challenges posed by the small sample\nsize. Second, it contains geographical and dynamic variables to account for\narea-specific features and the relative distance between the study area and the\nmoving TC. Third, it applies unequal weights to accommodate the temporal\nstructure in the training data when calculating the objective function. The\nproposed CNN-all model is then illustrated with the TC Soudelor's impact on\nTaiwan. Soudelor was the strongest TC of the 2015 Pacific typhoon season. The\nresults show that the inclusion of augmented data and dynamic variables\nimproves the prediction of heavy precipitation. The proposed CNN-all\noutperforms traditional CNN models, based on the continuous probability skill\nscore (CRPSS), probability plots, and reliability diagram. The proposed model\nhas the potential to be utilized in a wide range of meteorological studies.","PeriodicalId":501172,"journal":{"name":"arXiv - STAT - Applications","volume":"72 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A Convolutional Neural Network-based Ensemble Post-processing with Data Augmentation for Tropical Cyclone Precipitation Forecasts\",\"authors\":\"Sing-Wen ChenInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, Taiwan, Joyce JuangCentral Weather Administration, Taiwan, Charlotte WangInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, TaiwanMaster Program of Public Health, College of Public Health, National Taiwan University, Taiwan, Hui-Ling ChangCentral Weather Administration, Taiwan, Jing-Shan HongCentral Weather Administration, Taiwan, Chuhsing Kate HsiaoInstitute of Health Data Analytics and Statistics, College of Public Health, National Taiwan University, TaiwanMaster Program of Public Health, College of Public Health, National Taiwan University, Taiwan\",\"doi\":\"arxiv-2409.09607\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Heavy precipitation from tropical cyclones (TCs) may result in disasters,\\nsuch as floods and landslides, leading to substantial economic damage and loss\\nof life. Prediction of TC precipitation based on ensemble post-processing\\nprocedures using machine learning (ML) approaches has received considerable\\nattention for its flexibility in modeling and its computational power in\\nmanaging complex models. However, when applying ML techniques to TC\\nprecipitation for a specific area, the available observation data are typically\\ninsufficient for comprehensive training, validation, and testing of the ML\\nmodel, primarily due to the rapid movement of TCs. We propose to use the\\nconvolutional neural network (CNN) as a deep ML model to leverage the spatial\\ninformation of precipitation. The proposed model has three distinct features\\nthat differentiate it from traditional CNNs applied in meteorology. First, it\\nutilizes data augmentation to alleviate challenges posed by the small sample\\nsize. Second, it contains geographical and dynamic variables to account for\\narea-specific features and the relative distance between the study area and the\\nmoving TC. Third, it applies unequal weights to accommodate the temporal\\nstructure in the training data when calculating the objective function. The\\nproposed CNN-all model is then illustrated with the TC Soudelor's impact on\\nTaiwan. Soudelor was the strongest TC of the 2015 Pacific typhoon season. The\\nresults show that the inclusion of augmented data and dynamic variables\\nimproves the prediction of heavy precipitation. The proposed CNN-all\\noutperforms traditional CNN models, based on the continuous probability skill\\nscore (CRPSS), probability plots, and reliability diagram. The proposed model\\nhas the potential to be utilized in a wide range of meteorological studies.\",\"PeriodicalId\":501172,\"journal\":{\"name\":\"arXiv - STAT - Applications\",\"volume\":\"72 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-15\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - STAT - Applications\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.09607\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - STAT - Applications","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.09607","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
热带气旋(TC)带来的强降水可能导致洪水和山体滑坡等灾害,造成巨大的经济损失和人员伤亡。基于机器学习(ML)方法的集合后处理程序预测热带气旋降水因其建模的灵活性和管理复杂模型的计算能力而受到广泛关注。然而,当将 ML 技术应用于特定区域的 TC 降水时,可用的观测数据通常不足以对 ML 模型进行全面的训练、验证和测试,这主要是由于 TC 的快速移动造成的。我们建议使用卷积神经网络(CNN)作为深度 ML 模型,以充分利用降水的空间信息。所提出的模型有三个显著特点,有别于气象学中应用的传统 CNN。首先,它利用数据扩增来缓解小样本带来的挑战。其次,它包含地理和动态变量,以考虑特定区域的特征以及研究区域与移动 TC 之间的相对距离。第三,在计算目标函数时,它采用了不等权重以适应训练数据中的时间结构。然后,用苏迪罗风暴对台湾的影响来说明所提出的 CNN 全模型。苏迪罗是 2015 年太平洋台风季最强的热带气旋。结果表明,加入增强数据和动态变量可改善强降水预测。根据连续概率技能分数(CRPSS)、概率图和可靠性图,所提出的 CNN 均优于传统 CNN 模型。所提出的模型具有广泛应用于气象研究的潜力。
A Convolutional Neural Network-based Ensemble Post-processing with Data Augmentation for Tropical Cyclone Precipitation Forecasts
Heavy precipitation from tropical cyclones (TCs) may result in disasters,
such as floods and landslides, leading to substantial economic damage and loss
of life. Prediction of TC precipitation based on ensemble post-processing
procedures using machine learning (ML) approaches has received considerable
attention for its flexibility in modeling and its computational power in
managing complex models. However, when applying ML techniques to TC
precipitation for a specific area, the available observation data are typically
insufficient for comprehensive training, validation, and testing of the ML
model, primarily due to the rapid movement of TCs. We propose to use the
convolutional neural network (CNN) as a deep ML model to leverage the spatial
information of precipitation. The proposed model has three distinct features
that differentiate it from traditional CNNs applied in meteorology. First, it
utilizes data augmentation to alleviate challenges posed by the small sample
size. Second, it contains geographical and dynamic variables to account for
area-specific features and the relative distance between the study area and the
moving TC. Third, it applies unequal weights to accommodate the temporal
structure in the training data when calculating the objective function. The
proposed CNN-all model is then illustrated with the TC Soudelor's impact on
Taiwan. Soudelor was the strongest TC of the 2015 Pacific typhoon season. The
results show that the inclusion of augmented data and dynamic variables
improves the prediction of heavy precipitation. The proposed CNN-all
outperforms traditional CNN models, based on the continuous probability skill
score (CRPSS), probability plots, and reliability diagram. The proposed model
has the potential to be utilized in a wide range of meteorological studies.