Development of remote sensing and increased availability of satellite imagery of different spatial, spectral, temporal and radiometric characteristics makes information obtained from such sources of vital importance for studying and mapping vegetation. Vegetation indices have a significant role in vegetation change detection and tracking, whether in quantity or quality terms. Each index has specific significance and performance characteristics. A multiple regression statistical analysis of average vegetation index values (NDVI, NDWI, GNDVI, EVI and SAVI) was performed for 2012, 2013 and 2014 periods of a wider Cesma forest area near Vrbovec, Croatia. Further, rasters using three-year average values, sums, variances and standard deviations for all five indices were created. Differencing of average NDVI index values for years 2005 and 2014 was also performed. Imagery chosen was from the active vegetation period and used as a basis for cluster analysis detection of significant change areas. Cesma forest area was selected due to previous field monitoring and point analysis conducted (2012, 2013 and 2014) that serve as validation for this research. Finally, a raster analysis of select areas, surrounding accumulation dam and encompassing Cesma forest area exclusively, was conducted. The intention was determining vegetation index change dynamics. Spaciotemporal analysis around accumulation dams determined vegetation changes in dam areas. The advantage of the applied method is that, by using the Principal Components Analysis - PCA, it allows change detection, tracking and monitoring on wide areas more promptly than with other methods.The analysis itself was made using 92 LANDSAT images acquired over a 10-year period.