利用波形重定改善近岸與內陸小水體之Envisat測高資料

測高衛星於近岸與內陸應用上,受到陸地地形等影響,難以求得準確的觀測量,尤其是近岸5 km內海域與內陸小範圍水體仍是尚待解決之課題。本研究發展新的波形重定方法,以18 Hz Envisat RA-2 Cycle 10 ~ Cycle 90為研究資料,於臺灣周圍近岸實驗區中,利用Brown加上高斯模型擬合測高子波形,將高斯模型擬合之非海洋資訊移除,並自動化篩選出含有海水面資訊的波形。計算成果顯示,本研究發展之波形重定結果於各實驗區域之0-5 km海域內表現皆最為優異,重定後觀測量與驗潮站資料差值之標準偏差(Standard Deviation STD)之最小值為11 cm(和平港站),比其他重定方...

Full description

Bibliographic Details
Main Authors: 楊亭宜, Yang, Ting-Yi
Other Authors: 測量及空間資訊學系, 郭重言, Kuo, Chung-Yen
Format: Thesis
Language:Chinese
English
Published: 2014
Subjects:
Online Access:http://ir.lib.ncku.edu.tw/handle/987654321/150581
http://ir.lib.ncku.edu.tw/bitstream/987654321/150581/-1/index.html
Description
Summary:測高衛星於近岸與內陸應用上,受到陸地地形等影響,難以求得準確的觀測量,尤其是近岸5 km內海域與內陸小範圍水體仍是尚待解決之課題。本研究發展新的波形重定方法,以18 Hz Envisat RA-2 Cycle 10 ~ Cycle 90為研究資料,於臺灣周圍近岸實驗區中,利用Brown加上高斯模型擬合測高子波形,將高斯模型擬合之非海洋資訊移除,並自動化篩選出含有海水面資訊的波形。計算成果顯示,本研究發展之波形重定結果於各實驗區域之0-5 km海域內表現皆最為優異,重定後觀測量與驗潮站資料差值之標準偏差(Standard Deviation STD)之最小值為11 cm(和平港站),比其他重定方法之STD小4~11 cm。然而,於離岸5~10 km處,因測高波形多為典型的海洋波形,故各演算法之表現差異並不大。此外,比較各演算法經自動化篩選前後之精度,於距海岸線0~5 km中,測高觀測精度皆有提升之情形,其最大幅度之改善可將未重定觀測量與EGM2008差值之STD由5 119 m降至1 815 m(B-1區)。於內陸小範圍水體區(濁水溪下游及嘉義縣曾文水庫),本研究利用連續小波轉換測高波形,以轉換結果之過零值獲得真正的前緣中點位置,並自動化篩選出內陸水位之波形。濁水溪研究成果顯示,本研究發展之波形重定結果可將未重定觀測量與水位站資料差值之STD從1 409 m降至0 624 m,比其他波形重定演算法小7~165 cm。於曾文水庫之結果顯示,因測高觀測量嚴重受到水庫周圍地形影響,導致水庫內所有測高觀測量中,僅有約25 %(88筆)為標準的內陸水體波形,各波形重定技術也無法顯著改善測高觀測量。其中,Ice-1與Sea-Ice演算法甚至有不合理之改正情形,故於地形起伏較大之內陸區域,即使測高衛星距離觀測量已利用波形重定演算法處理過,其應用及觀測精度仍需格外小心。 Satellite altimeters provide many kinds of accurate ocean measurements and become an irreplaceable tool to address wide range of scientific questions over the open oceans However the accuracy of observations is quickly degenerated as the altimeters are close to coasts or on non-ocean surfaces due to improper geophysical corrections and complex returned waveforms Although several algorithms have been developed to retrack the waveforms over the coastal oceans or non-ocean areas the performance and accuracy are not as good as expected In this study we develop a novel strategy to derive accurate sea surface heights (SSHs) near the Taiwan coasts from Envisat RA-2 waveforms by using the Ocean plus Gaussian Models We also automatically filter the improper ocean waveforms and validate the experimental results using tide gauge records and EGM2008 The results show that our retracking strategy has substantially improved the accuracy of the altimeter derived SSHs within 0~5 km from the coastlines Furthermore a wavelet-based algorithm is used to retrack the Envisat waveforms in Tsengwen reservoir and Zhoushui river The retracked water heights are evaluated using water level gauge data In Zhuoshui river the STD of the developed retracked water heights is 0 624 m compared to the unretracked water heights with the standard deviation (STD) of 1 409 m However in Tsengwen reservoir waveform retracking cannot significantly improve the accuracy of the altimetry data due to the surrounding terrain and automatic gain control (AGC) excessive adjustment