###
DOI:
本文二维码信息
相似台风智能检索与可视化分析研究
宋文彬1, 郑卫江2, 薛峰2
(1.国家气象中心;2.国家气象中心)
Research of Intelligent Retrieval and Visualization of Similar Typhoons
Song Wenbin1, Zheng Weijiang2, Xue Feng2
(1. National Meteorological Center;2.National Meteorological Center)
摘要
相似文献
本文已被:浏览 13次   下载 1
投稿时间:2019-09-24    修订日期:2020-08-15
中文摘要: 利用空间分析及可视化技术,对1949-2018年间的台风时空变化规律进行相似性智能检索,形成基于空间位置、特征要素、时间规律三个关键指标的综合定量化判定体系,初步建立了统一的判定标准。其中,“路径”相似性判定提出了多项动态拟合与相似离度结合的方法。研究结果表明:综合定量化判定体系可有效降低单一指标的不确定性。同时,“路径”的动态判定法有效解决了观测点不匹配的问题。以台风Mangkhut(1822)为例,首先通过数据清洗、特征提取和构建快速索引等检索方法,缩小判定范围,检索出10个可能与目标相似的台风,然后利用综合定量化判定方法得到7个相似台风。判定过程中实现了基于定制条件的多维度时空数据逐级下钻、定量化综合智能检索以及相似性结果可视化三个目标,为台风台风预报预警业务提供重要参考。
Abstract:In this study, using a spatial analysis and visualization technology to intelligently and similarly search for regular with the temporal and spatial variation of landing typhoons from 1949 to 2018, to form a comprehensive quantitative judgment system based on three key indicators as spatial location, feature elements and temporal rule, to establish initially a unified judgment standard. According to path similarity determination, the method based on the combination of multiple dynamic fitting curves and similar deviation has been proposed. The research results show that the comprehensive quantitative judgment system can effectively reduce the uncertainty of single index. And the dynamic judgment method can effectively solve the problem of observation point mismatch. Taking typhoon Mangkhut (1822) as an example, firstly using retrieval methods as data cleaning, feature extraction and quick retrieval creation to narrow searching range, to retrieve 10 possible typhoons with similarity to the target, then using the comprehensive quantitative judgment method to confirm 7 similar typhoons. During the judgment process, there are three targets can be achieved, the multi-dimensional spatiotemporal data based on customized conditions are drilled down step by step, the quantitative comprehensive intelligent retrieval and the visualization of similarity results, to provide important reference for typhoon forecast and early warning operation.
文章编号:201909240349     中图分类号:    文献标志码:
基金项目:高时空分辨率气象预报分布式计算能力建设
引用文本:
宋文彬,郑卫江,薛峰,0.[en_title][J].Meteor Mon,():-.
Song Wenbin,Zheng Weijiang,Xue Feng,0.Research of Intelligent Retrieval and Visualization of Similar Typhoons[J].Meteor Mon,():-.