Wavelet Analysis of Satellite Images for Sea Ice Motion

利用小波理論分析衛星影像來推估海冰的飄移

劉安國(Antony K. Liu) ； 趙雲鶴(Yun-He Zhao)

《航測及遙測學刊》 7卷3期 (2002/09) Pp. 59-74

https://doi.org/10.6574/JPRS.2002.7(3).6

摘要

本研究利用小波理論，分析QuickSCAT衛星和DMSP SSM/I衛星所收集之兩種不同性質(QuickSCAT衛星是微波波段的背向散射，SSM/I衛星是微波波段的輻射)的影像資料，來推估每天海冰在南北極的飄移。由推算的結果發現，兩顆衛星雖然資料性質不同，但所推估之北極海冰在秋天、冬天及春天的飄移和海上浮標的觀測十分吻合，而且兩者是互補的。也就是說，若將三者的資料合併，便可更加完整地了解，北極地區海冰每日飄移的狀況。由南極地區的一個研究案例發現，海冰的飄移主要是風所推動的，而移動的方向大小也與周圍的風場?合。根據1999年12月所製作北極海冰的日飄移圈，發現海冰運行的模式每4到7天就有相當大的變化，而這些改變主要是受到北極氣壓的影響。

關鍵字

小波轉換 ； QuickSCAT ASSM/I ； 海冰運動 ； 特徵追蹤

並列摘要

QuikSCAT backscatter and DMSP SSMII radiance data are used to derive sea ice motion for both the Arctic and Antarctic region using the wavelet analysis method. Sea ice motion of the Arctic for fall/winter/ spring time derived from NSCAT, QuikSCAT and SSM/I data agree well with ocean buoy observations. Both comparisons of result from QuikSCAT and SSMII with buoys are compatible, and their ice tracking results complement each other. Sea-ice drift daily results from QuikSCAT, SSM/I, and buoy data can be merged to generate composite sea ice motion maps for more complete coverage of sea ice motion. In the Antarctic, a case study shows that sea ice motion derived from NSCAT/QuikSCAT data is predominantly forced by and consistent with the surrounding wind field. Based on a series of daily Arctic sea-ice drift maps in December 1999, it is found that the major circulation patterns are changing and shifting significantly within every four to seven days and they are dominated by wind forcing.

並列關鍵字

Wavelet transform ； QuikSCAT ； SSM/I ； sea ice motion ； feature tracking