| DC 欄位 |
值 |
語言 |
| DC.contributor | 大氣科學學系 | zh_TW |
| DC.creator | 姚奕安 | zh_TW |
| DC.creator | I-An Yao | en_US |
| dc.date.accessioned | 2019-6-19T07:39:07Z | |
| dc.date.available | 2019-6-19T07:39:07Z | |
| dc.date.issued | 2019 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=106621007 | |
| dc.contributor.department | 大氣科學學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 此研究使用中央氣象局所提供的整合最大回波資料(Maximum Reflectivity)以及加拿大麥基爾大學(McGill University)所開發的雷達回波外延預報系統(McGill Algorithm for Precipitation nowcasting using Lagrangian Extrapolation, MAPLE)針對16個颱風個案進行分析，以此檢視此外延法在即時預報上應用於臺灣地區颱風天氣系統的預表現。此外為了提升極短期預報(Nowcasting, 0-6小時)能力，將數值模式中所提供的風場資訊與回波外延之移動場進行結合。而本研究中除了使用最常見的連續校驗(continuous verification)以及絕對校驗(categorical verification)，也加入了相鄰法(neighborhood method)進行評估預報結果，避免受到高解析度的影響低估模式預報能力。
首先，為了得到良好的定量降水估計(Quantitative Precipitation Estimation, QPE)，針對數個Z-R關係式進行檢驗。接著，由外延預報的結果顯示MAPLE在定量降水預報上可以提供約2小時的有效預報。但回波外延預報系統隨著時間預報結果會有嚴重的變形現象。藉由結合數值天氣預報的風場資訊來改進此現象，同時可以維持颱風旋轉特性以及能捕捉到雨帶在地形附近的結構。在透過上述之各項校驗分數的評量，整體而言，結合過後的即時預報系統可以將極短期預報的能力提升至3小時，有效的降低在外延預報中產生的位移不確定性。這樣的改善，尤其在颱風登入後至離臺這段時期，改善顯著。對於防災 、減災而言方面可以即時的提供天氣資訊。
| zh_TW |
| dc.description.abstract | In this study, composite radar data from the Central Weather Bureau (CWB) of 16 typhoons are collected to examine the statistical performance of the McGill Algorithm for Precipitation nowcasting using Lagrangian Extrapolation (MAPLE) over Taiwan. In addition, in order to improve the nowcasting system, the information of the numerical model is blended into the system. In order to examine the performance of the nowcasting, continuous and categorical verification is used. However, grid-point verification is strict for high resolution and could underestimate the ability of the prediction system. Therefore, the neighborhood method is also applied for validation.
First, in comparison to the rainfall amount from gauges, the best Z-R relationship is determined. Second, the statistical results of the radar echo extrapolation for 16 typhoon cases show that the quantitative precipitation nowcasting skill could persist for up to 2 hours. However, significant distortion for the rotational system is found after 2 hours. Therefore, the information of the numerical model is blended to capture and maintain the rotation of typhoon rain-band structures. When verifying the performance of the hybrid nowcasting system, whether from the aspect of categorical verification or the neighborhood method, in general, the hybrid scheme of the system further improves the nowcasting for up to 3 hours. Furthermore, the improvement of the hybrid scheme is more significant after the typhoon landed in Taiwan. For disaster prevention and mitigation, the nowcasting system can provide effective weather information immediately.
| en_US |
| DC.subject | 即時預報 | zh_TW |
| DC.subject | 雷達回波外延 | zh_TW |
| DC.subject | nowcasting | en_US |
| DC.subject | echo extrapolation | en_US |
| DC.title | 藉由數值模式水平風場改善雷達回波外延即時預報系統:16個颱風個案統計分析 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Improving radar echo Lagrangian extrapolation by blending numerical model wind information: statistical performance of 16 typhoon cases | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |