2018 年 74 巻 2 号 p. I_979-I_984
Shoreline positions of Kashima Coast facing the Pacific Ocean, which is approximately 16 km long with Hasaki Fishery Port at the south end and Kashima Port at the north end, have been observed with four land-based X-band radars and Synthetic Aperture Radar (SAR) satellite. X-band radars observe shoreline positions continuously in time but do not cover the whole coast. On the other hand, SAR covers the whole spatial domain, but data is available only a few times in a year. The purpose of the present work is to propose a data fusion method which combines different shoreline data observed by X-band radars and SAR satellite with the help of Garcia's method, a Penalized Least Square (PLS) regression based on Discrete Cosine Transform (DCT). Garcia's method is initially applied to shoreline positions dataset derived from X-band radars, and its performance has been checked for this dataset with artificial gaps. Then Garcia's method is executed to combine Radar and SAR shoreline positions dataset together. The data fusion result is verified by survey data, and we confirm that our fusion method performs reasonably well to process shoreline data set.
