Analisis Perilaku Mean Dataset Perubahan Garis Pantai pada Hasil Spasialtemporal Metode Empirical Orthogonal Function (EOF)

The purpose of this study is to apply EOF method on shoreline change resulting spatialtemporal analysis mode 1 and also to to prove the mean behavior on the spatial or temporal of its EOF outcomes. The data used was obtained from shoreline coordinates of shp file of the Bali island map and the result of breaking waves study, by using a one-dimensional modeling yielded dataset or prediction shoreline changes during 91 months. Calculation EOF 1, the input matrix or dataset was initially not reduced by the mean but EOF 2 vice versa. Each matrix receive the same treatment were calculated covariance, eigen value, eigen vector and principal component. EOF calculations obtained the last five eigen value, the last five eigen vector, trace, principal component and variant data. Based on the results obtained were compared parameters of two matrices mentioned before. Spatially results both of EOF 1 and EOF 2 shows the same eigen vector represented by the first mode of eigen vector. Similarly, the eigen value, trace and variance of data, produce the same information. Significant difference occurs in the principal component (temporal). EOF 1 shows that the value of the first month produces a positive value, second month until month 91th output are minus. EOF 2 shows the value of the principal component the first month until the 37th month are in a positive position, then from month 38th to month 91th yielded negative results. Nevertheless EOF 1 and EOF 2 showed shoreline changes tend to be erosion.