Application of Fractal and Gray-Level Co-Occurrence Matrix Indices to Assess the Forest Dynamics in the Curvature Carpathians—Romania

• Published in: Sustainability 2019-12, Vol.11 (24), p.6927
• Main Authors: Ciobotaru, Ana-Maria, Andronache, Ion, Ahammer, Helmut, Radulovic, Marko, Peptenatu, Daniel, Pintilii, Radu-Daniel, Drăghici, Cristian-Constantin, Marin, Marian, Carboni, Donatella, Mariotti, Gavino, Fensholt, Rasmus
• Format: Article
• Language: English
• Subjects: Agricultural land; Algorithms; Biodiversity; Climate change; Curvature; Deforestation; Evolution; Forest ecosystems; Forests; Fractal analysis; Fractal geometry; Fractals; Image contrast; Landsat satellites; Local communities; Matrix methods; Mountains; Pixels; Regional development; Regional planning; Satellite imagery; Satellites; Spatial distribution; Sustainable development
• ISSN: 2071-1050; 2071-1050
• DOI: 10.3390/su11246927

The mountain ecosystems face significant damage from deforestation and environmental forest changes. We investigated the evolution of tree types of cover areas, deforested areas and total deforested areas from Curvature Carpathians using Gray-Level Co-occurrence Matrix and fractal analysis. The forest dynamics mapping was one of the main objectives of this study and it was carried out using multiple fractal and GLCM indices. We approached the analysis of satellite forest images by calculation of four fractal indices such as Pyramid dimension, Cube Counting Dimension, Fractal Fragmentation-Compaction Index and Tug-of-War lacunarity. We also calculated fractal dimension because it is an index of complexity comparing how the detail in a pattern changes with the scale at which it is measured. Fractal dimension is useful for estimation of irregularity or roughness of fractal and natural objects that do not conform to Euclidian geometry. While the fractal dimension quantifies how much space is occupied, the Tug-of-War lacunarity complements fractal dimension with its ability to quantify how space is occupied. Analysis was further supplemented by the Gray-Level Co-occurrence Matrix analysis because it quantifies spatial probability distributions of gray level values between pixel pairs within an image. The calculated Gray-Level Co-occurrence Matrix features included Angular Second Moment, Contrast, Correlation, Inverse Difference Moment and Entropy. Such comprehensive analysis has the advantage of combining fractal analysis that extracts quantitative information about the morphological complexity of the image with the spatial distribution of the gray pixel intensities as calculated by the co-occurrence features provided by Gray-Level Co-occurrence Matrix. Evolution of deforested areas, expansion of agricultural land and the increased demand for quality timber have affected the forests ecosystems and, the regional sustainable development of local communities.