使用群集分析分類綜觀尺度天氣型態以探討台灣北部地區午後熱對流系統局部環流結構與系統發展特性;Classifying Synoptic Weather Type to Investigate the Relationship between Local Circulation and Characteristics of Afternoon Thunderstorm in Northern Taiwan

NCU Institutional Repository > 地球科學學院 > 大氣物理研究所 > 博碩士論文 > Item 987654321/72853

jsp.display-item.identifier=請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/72853

题名:	使用群集分析分類綜觀尺度天氣型態以探討台灣北部地區午後熱對流系統局部環流結構與系統發展特性;Classifying Synoptic Weather Type to Investigate the Relationship between Local Circulation and Characteristics of Afternoon Thunderstorm in Northern Taiwan
作者:	蘇凱翊;Su, Kai-Yi
贡献者:	大氣科學學系
关键词:	群集分析
日期:	2017-01-25
上传时间:	2017-05-05 17:09:15 (UTC+8)
出版者:	國立中央大學
摘要:	摘要午後雷陣雨又稱為對流雨以及熱雷雨，主要降水特徵為雨時短、雨區小、強度大，並且常發生在午後，為台灣北部地區夏季主要降水來源之一。本研究主要目的為使用群集分析進行綜觀尺度天氣形態分類，藉以探討台灣北部地區午後雷陣雨氣象特性。群集分析所使用之觀測資料，包含台灣北部地區2013年到2015年夏季(7、8及9月份)氣象局地面測站觀測資料，選取的氣象參數包含溫度、濕度、風速、風向、氣壓、降雨量以及比濕。研究方法，首先將這些氣象變數進行不同的組合，並進一步將每種組合的分類結果進行分析，找出各群集的特性，並在每個群集內製作降雨時序圖比對，再找出所有組合的最佳分類，最後進行分析與討論，以上方法可以找出午後雷陣雨和綜觀天氣型態的關係；而後將上述觀測資料重新篩選，單純挑選發生午後雷陣雨的事件，再重新進行群集分析，並與前一個群集分析分類結果進行比對。首先，使用風速和風向觀測資料進行群集分析，分類結果顯示北部地區夏季主要盛行風場為西南、東南以及東北風，主要降雨集中在東北風、西北西風以及東南風這三個群集內。東北風的降雨主要由三種類型的天氣型態所組成，分別為颱風、冷鋒鋒面以及冷高壓所產生的東北風。西北西風和東南風相關的群集，則是包含數個颱風個案，因此降雨量較高。若於群集分析方法排除風場變數，僅使用氣壓、溫度以及相對濕度來進行分類，分類結果可分為五類，第一類屬於弱綜觀天氣型態，降雨幾近於零；第二類為溫度較低的東北風天氣型態，降雨較少且濕度較低，屬於冷高壓影響的天氣型態；第三類為低壓系統所主宰的天氣型態，主要包含華南以及南方雲雨帶、鋒面、颱風外圍環流以及熱帶低壓，這類型的降雨較多且濕度也相對較高；第四類為颱風類型，颱風中心位於台灣本島上的天氣系統；第五類屬於午後雷陣雨的天氣型態，降雨主要集中在午後，屬於弱綜觀天氣型態，整體溼度和溫度較高；第六類為上述五類少數個案的集結，同時具有上述五類所有特性。再者，將三年夏季的觀測資料進行篩選，挑選午後雷陣雨的天數，並進行群集分析，群集分類使用的資料為U、V風場。分類結果共有三類，第一類為受到副熱帶高壓影響的東南風天氣型態，天數較少；第二類的降雨量為三個群集內最多的群集，降雨分布範圍也較廣，屬於太平洋高壓勢力減弱及南方低壓帶北移的天氣型態，北部地區主要吹東北風，發生的天數和第一類相同；第三類的天氣型態和第一類較為接近，綜觀天氣型態皆是受到副熱帶高壓影響，但是主要的風向為西南風，且發生天數的比例超過一半，降雨量和第一類相差無幾，而分類結果屬於西南風的天氣型態較容易產生午後雷陣雨。最後，挑選一個午後雷陣雨的事件(2014年8月19日)進行個案分析，並使用中央大學移動式雷達Team-R以及文化大學為此觀測施放的探空資料進行分析，了解此個案午後雷陣雨的熱力以及動力結構。 ;Abstract The afternoon thunderstorm system is associated with short rainfall time period, confined rainfall area and strong rainfall intensity. The objective of this study is to understand the synoptic weather patterns that are prone for the development of the afternoon thunderstorm system through cluster analysis method and at the same time, to investigate the meteorological characteristic including the local circulations and thermal structures of the afternoon thunderstorm system . The observation data from weather stations of Central Weather Bureau (CWB) is used for cluster analysis. The data period covers the summer season (July, August and September) from 2013 to 2015. The selected meteorological variables are temperature, humidity, wind speed, wind direction, pressure, rainfall, and specific humidity. The cluster analysis using these surface meteorological variables are tested in different combinations to find out the best weather classification that can identify the characteristics of the afternoon thunderstorm system. In first experiment, the cluster analysis was applied using U- and V- component wind to classify the weather pattern. The classification results exhibit three types of prevailing winds in northern Taiwan including the northeasterly flow, southwesterly flow, and southeasterly flow. In second experiment, the cluster analysis was applied using pressure, temperature, and relative humidity. Five clusters are found using the above method. Cluster one and cluster five exhibit similar synoptic weather pattern for which Taiwan is under influence of sub-tropical high-pressure system. Cluster one is associated with southeasterly wind in northern Taiwan and with less rainfall. Cluster five is associated with southwesterly wind and exhibits the afternoon thunderstorm rainfall pattern. The occurrence of cluster two is less and is associated with the northeasterly wind due to the influence of the continental high-pressure system. Cluster two has less rainfall. In Cluster three, the prevailing wind is from northeast and the strength of the Pacific high-pressure system weakens. Taiwan is influence by the low-pressure system. Cluster four is associated with the highest rainfall due to the typhoons. Cluster six doesn’t have apparent characteristics. Lastly, only cases associated with the afternoon thunderstorm are selected from three-year summertime observation data and the cluster analyzed was applied using U- and V-component wind field. Three clusters are found using the above method. The cluster one is with less occurrence and due to the influence of the sub-tropical high pressure system the dominant wind is southeasterly. The second cluster has the highest rainfall amounts and the precipitation occurs in wider area among three clusters. The dominant wind is northeasterly in northern Taiwan when Pacific high-pressure system weakens. The cluster three has similar weather pattern as the cluster one. The synoptic weather pattern is under the influence of sub-tropical high-pressure system but the dominant wind is southwesterly. The occurrence of cluster three is more than 50 percent of the days examined. Furthermore, CWB observation data is employed to analyze an afternoon thunderstorm case (19 August 2014) and the Team-R observation data from National Central University (NCU) and sounding data from Private Chinese Culture University (PCCU) are utilized to analyze the thermal and dynamic structure of the system.
显示于类别:	[大氣物理研究所 ] 博碩士論文

文件中的档案:

档案	描述	大小	格式	浏览次数
index.html		0Kb	HTML	341	检视/开启

在NCUIR中所有的数据项都受到原著作权保护.

社群 sharing

数据加载中.....