標題: | 利用人造衛星測高儀資料推算海深 The prediction of the Bathymetry Using Satellite Altimeter Data |
作者: | 黃大山 Dar-Shan Hwang 黃金維 Cheinway Hwang 土木工程學系 |
關鍵字: | 測高儀;微波脈衝;海深;altimeter;pulse;bathymetry |
公開日期: | 1994 |
摘要: | 自從1978年Seasat衛星發射以來,利用衛星測高資料逆推海深的研究即已 展開,本文即使用衛星測高資料推算重力,並據此推算得海深。本文利用 結合Seasat、Geosat、TOPEX/POSEIDON和ERS-1/35天、ERS-1/ 168 天等 衛星測高資料,以最小二乘配置法推算南海地區海上重力值,其與 Sandwell之重力值比較,明顯地精度要高出一倍。本文所推算之重力在平 差前後之均方根值為47.769、13.148 mgals,Sandwell之重力在 差前後 之均方根值為51.02、21.781 mgals。短波長海深將會產生短波長之重力 異常,長波長的海深將會被補償。由於重力異常和海底地形存在一線性轉 換關係,利用傅立葉變換,將理論之反應函數補償模式來測試最密合於實 際之反應模式及地球物理參數。以南海地區之海上重力值,利用去除─回 復法理論,以快速傅立葉變換及其逆變換來求得海深。由於海底地形和重 力之間受到一層沉積物之影響,所以其與船測海深資料比較,均方根值 在200─400公尺。南海為一複雜海域,其精度之提昇與補償模式及地球物 理參數有密切之關係。 Since the lauch of Seasat, the research on the prediction of bathymetry has begun a new era. In this study the satellite -derived gravity is used for the prediction. The satellite gravity over the South China Sea is determined by least-squares collocation using data from Seasat, Geosat, TOPEX/POSEIODN, S-1/35day and ERS-1/168day missions. The satellite gravity has a higher accuracy than that derived by Sandwell et al.(1994) and shows rms errors of 47.76 and 13.14 mgals respectively afterd before the adjustment when compared with the ship data. The gravity from Sandwell et al.(1994) has rms errors of 51.02 and 21.78 mgals with ship gravity. The short-wavelength bathymetry produces the short-wavelength gravity anomalies, while the long-wavelength bathymetry ismpensated. Using the linear relationship between gravity anomaly and bathymetry in frequency domain, this paper employs a theoretical compensation model to determine geophysical parameters. Based on the remove- restore procedure, we compute thethymetry from marine gravity using FFT. The rms difference between the predicted bathymetry and ship bathymetry is on the order of 200-400 meters, owing to the uncertainty in the geophysical parameters, which are affected by the sediment layer. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#NT830015005 http://hdl.handle.net/11536/58694 |
顯示於類別: | 畢業論文 |