부산대학교 도서관

The domain of Matched-field processing (MFP) has been comprehensively studied in temperate oceans. But MFP source localization in under-ice shallow water environment has rarely conducted and is a topic of great novelty. The frequent interactions of sound and ice will lead dramatic reflection, attenuation and modal dispersion which will pose additional difficulties and uncertainties to localization due to the intense environmental sensitivity of the MFP method. This paper constructs the basic MFP source localization scheme in under-ice environment. Firstly, a hybrid acoustic propagation model called OASR-KRAKENC that considers the effects of ice reflection and attenuation are constructed to generate the replicas in under-ice waveguide. Then the incoherent broadband MFP approach of conventional linear correlation (Bartlett) and minimum variance (MV) algorithm are verified by the real data collected from the northern Yellow Sea in winter. The localization performance of real data validates the feasibility of MFP in under-ice environment.