Coastal modification of a scene employing multispectral images and vector operators

Abstract

Changes in sea level, wind patterns, sea current patterns, and tide patterns have produced morphologic transformations in the coastline area of Tamaulipas Sate in North East Mexico. Such changes generated a modification of the coastline and variations of the texture-relief and texture of the continental area of Tamaulipas. Two high-resolution multispectral satellite Satellites Pour l’Observation de la Terre images were employed to quantify the morphologic change of such continental area. The images cover a time span close to 10 years. A variant of the principal component analysis was used to delineate the modification of the land-water line. To quantify changes in texture-relief and texture, principal component analysis was applied to the multispectral images. The first principal components of each image were modeled as a discrete bidimensional vector field. The divergence and Laplacian vector operators were applied to the discrete vector field. The divergence provided the change of texture, while the Laplacian produced the change of texture-relief in the area of study.