摘要
针对基于沿海GPS站点坐标时间序列反演海潮负荷特征信息研究较少的问题,提出利用FFT与HHT提取海潮负荷信息,两种频谱方法原理各异,但都可将目标信息从原始数据中分离出。基于沿海站点的坐标时间序列,经小波去噪后进行海潮分潮特征信息提取试验,最后与3种全球海潮模型FES2004、NAO.99b和GOT4.7的模型值比较,评估两种方法的反演精度,结果显示:HHT、FFT都能够准确探测到海潮分潮信号,FFT全日潮周期反演结果优于HHT反演结果,半日潮周期反演精度相当;HHT边际谱的谱值为对应频率出现的可能程度,无法与模型值进行比较,而FFT振幅值为绝对值,其结果与模型值符合情况良好;在探测未知信号时,FFT可能会探测出虚假信号,而HHT更能真实反映信号的组成。
This paper lays emphasis on the lack of the research of deriving the ocean tide loading(OTL)character,which based on the costal GPS sites,using FFT and HHT to drive OTL character.Two types of spectrum analysis are different,but both can derive the target information from original data.Based on the position time series of China costal GPS sites time series,after wavelet filtering,we use two spectrum analysis methods to derive OTL character and comparing with the result from three global ocean loading models to assess the precision.The study shows that:Both FFT and HHT can detect the ocean sub-tides precisely,FFT is better than HHT of the diurnal tides and the precision of semi-diurnal they are the same;The amplitude from HHT is relative value that represents the possibility of some frequency,they can't be compared with the models,while the amplitude derived by FFT is absolute value,FFT/model has a good agreement;When detecting unknown signal,using FFT may detect the some fake signal while HHT can show the truly constitution of the signal.
出处
《测绘科学》
CSCD
北大核心
2017年第1期113-118,共6页
Science of Surveying and Mapping
基金
国家自然科学基金项目(41374044
41274011)
国家海洋局预报减灾司项目(1511311700008)