Principal component analysis (PCA) and complex PCA (CPCA) are applied to examine medium-term morphological behavior of a sandy beach before Gamo lagoon located at the mouth of Nanakita River along Sendai Coast, Japan, using a 12-year series of bottom survey data. The basic features and applicability of PCA and CPCA to beach morphology data sets are preliminary investigated with model topography data; CPCA is found to have the strong advantage of capability of representing propagating-wave-like migration of topography by using only single mode, whilst PCA necessitates two modes to reproduce such migrating processes. However, the results of CPCA for spatially uniform erosion model show that the both of real and imaginary part of the eigenfunction cn, (t) are appreciably distorted in their time series plots owing to the Fourier series expansion in the Hilbert transform procedure. The results of PCA and CPCA for the morphology data set show that erosions and accretions of the submerged terrace formed in front of the river mouth appear in the first mode of CPCA and have the most significant influence on the medium-term morphological processes. The second mode of CPCA demonstrates that the sediments previously discharged from the river return again into the nearshore region, and furthermore, the topography changes due to cross-shore sediment transport emerge in the third mode of CPCA.