標題: 以三維大氣模式分析颱風時期超導重力儀資料:以尼伯特颱風為例
Analysis of superconducting gravimeter data during typhoon period using 3-D atmospheric model: case study of typhoon Nepartak
作者: 余明晏
黃金維
Yu, Ming-Yen
Hwang, Cheinway
土木工程系所
關鍵字: 大氣改正;歐洲再分析場;逆氣壓反應;超導重力儀;颱風;Atmospheric Gravity Effect;ERA-Interim;Inverted Barometer Response;Superconducting Gravimeter;Typhoon
公開日期: 2017
摘要: 颱風尼伯特在2016年7月對東台灣造成極大規模的風災,其中在台東縣成功測站所測得最強風速73 m/s更打破過去20年紀錄、中央氣象局公布之最低氣壓達到905 hPa。目前擺放在陽明山衛星追蹤站的超導重力儀SG-T49在近四年提供長期且連續的高頻率且高精度的相對重力資料。由於尼伯特的路徑從南台灣經過,測站周圍並未受到劇烈降雨擾動,在颱風侵台期間完整記錄了其所造成的重力變化。本研究主要使用超導重力資料,利用2013-2016四年的資料進行調和分析重新計算得固體潮與海潮負載效應,與極移等重力環境改正移除後得僅大氣剩餘重力。大氣改正部分透過三維模式取代經驗值方法之大氣改正,利用歐洲中期氣象預報中心ECMWF(European Center of Medium-Range Weather Forecast)提供之ERA-Interim(European Reanalysis Interim)歐洲再分析場全球模式的三維大氣資料,計算出大氣中壓力、溫度、濕度變化對應之牛頓萬有引力變化,搭配超導資料討論颱風對超導重力之特殊影響。利用三維模式計算尼伯特颱風氣團質量引力對於超導重力影響之絕對量可達500 μGal以上,相對變化量約達9 μGal;其中溫度與濕度變化可造成約6 μGal的影響;在測站海拔高度以下之大氣質量在長周期變化可造成約1 μGal的影響、高頻變化可造成0.8 μGal。將三維大氣模式與經驗值方法比較,在颱風期間日夜溫差變化消失,大氣太陽潮改變使僅大氣剩餘重力高頻變化與三維模式較符合;在造成最大影響量上,三維模式較經驗值方法多約3 μGal,改正後之剩餘重力標準偏差為±0.62 μGal較原經驗值方法之±1.51 μGal更佳,確認經驗值方法利用-0.35 μGal∙〖hPa〗^(-1)作為氣壓-重力導納係數在颱風時期之不適用。在三維模式改正後不存在大氣負載的訊號,反映了尼伯特時期的海面仍為逆氣壓反應。為了探討颱風時期非逆氣壓理論,本研究再以2015年8月之天鵝颱風作分析,在經三維吸引力模式改正剩餘重力中發現以0.14 μGal〖∙hPa〗^(-1)計算之大氣負載訊號,發現非逆氣壓反應的發生。我們將尼伯特與天鵝颱風做比較,發現雖然兩者在超導附設的單點氣壓計紀錄的氣壓變化一致,且前者三維影響量變化較大,但後者在氣壓因子的三維影響量變化較大2 μGal。
Typhoon Nepartak caused a significant wind hazard in July, 2016. Weather station Cheng-Kung in Taitung recorded a highest wind speed of about 73 m/s, which broke the record in the last 20 years. The Central Weather Bureau(CWB) recorded lowest atmospheric pressure of 905 hPa. Superconducting gravimeter (SG) T49, located at satellite tracking station Yang-Ming-Shan (YMSM), provides long term, continual, high frequency and high precision gravity records. Because the track of Nepartak passed through southern Taiwan, T49 was able to detect gravity changes during typhoon period without disturbance of heavy rain. Using the gravity records of T49from 2013 to 2016, we calculated tidal parameters for the solid earth tide and ocean loading effects for correcting such effects. The polar motion effect was removed in the raw gravity. With all such effects removed, the gravity residual T49 contains only the atmospheric gravity effect which is computed using a 3-D model (Newtonian part) rather than the empirical, admittance-based method. We used 3-D atmospheric data from European Reanalysis Interim (ERA-Interim), established by European Center of Medium-Range Weather Forecast (ECMWF), to calculate the Newtonian attractions due to atmospheric mass change resulting from changes in pressure, temperature and humidity. With the SG data and the 3-D atmospheric gravity change model, we identified the specific gravity effects induced by typhoons. Based on the 3-D model, the air mass of typhoon Nepartak can cause gravity changes of up to 500 μGal in absolute value and 9 μGal in relative value. Temperature and humidty changes can cause gravity variations of up to 6 μGal. The air mass under the T49 gravity station can cause gravity changes of about 1 μGal over a long period and up to 0.8 μGal in high-frequency gravity oscillation. Compared to the empirical method, the 3-D model produces an atmospheric effect that is more consistent with the expected gravity change due to atmospheric solar tide change caused by day and night temperature fluctuations during Nepartak. The 3-D result was 3 μGal larger than that from the empirical method. The gravity residual of T49 corrected with the 3-D model results in a standard deviation of 0.61 μGal, which is smaller than the standard deviation of 1.51 μGal based on the empirical method. This suggests that the empirical method, based on a -0.35 μGal〖∙hpa〗^(-1) atmosphere-to-gravity admittance, is not appropriate for explaining gravity changes during Nepartak. The gravity residual corrected by the 3-D model did not show any sign of gravity change due to atmospheric loading. This suggests the inverted barometer (IB) effect occurred at sea near T49 during Nepartak. In order to see how non-IB could occur during typhoon period, over different typhoons we added a discussion using a second typhoon, Goni, occurring in August, 2015. After correcting for the 3-D atmospheric effect, the remaining gravity of T49 during Goni shows an empirical admittance of 0.14 μGal〖∙hpa〗^(-1). This suggests that the non-IB response has occurred. During typhoons Nepartak and Goni, the ranges of pressure change were almost the same. Compared to Goni, Nepartak had a lower lowest pressure change and larger 3-D gravity effects (2 μGal more).
URI: http://etd.lib.nctu.edu.tw/cdrfb3/record/nctu/#GT070451277
http://hdl.handle.net/11536/142847
Appears in Collections:Thesis