We analyze sea level variation data acquired by a differential GPS and gravity data acquired by a superconducting gravimeter (SG) at Syowa Station, Antarctica, in an eight month period of 1998. At frequencies between 0.2 and 2.5 mHz in the seismic normal mode band we observe similar spectral peaks in both of the data sets. We also observe high coherence and zero phase between the two data sets at the frequencies of these peaks. The results of response analysis and simple mode calculation suggest that the observed peaks in the SG data are due to the effects of ocean water attraction and loading associated with sea level variation, a possible cause of which is the seiche in L¿tzow-Holm Bay around the station. Applying a transfer function method to both of the data sets, we can reduce the background noise due to the oceanic effects in the SG data.