Application of Fractal Dimension to Assess the Effect of Scale on the Sensitivity of Landscape Metrics

The sensitivity of landscape metrics to the scale effect is one of the most challenging issues in landscape ecology and quantification of land use spatial patterns. In this study, fractal dimension was employed to assess the effect of scale on the sensitivity of landscape metric in the north of Iran (around Sari) as the case study. Land use/ cover maps were derived from Landsat-8 (OLI sensor) image processing and its spatial scale was downgraded to 30, 60, 120, 150, 200, 250, and 300 by cell-center method. After that, landscape-level metrics were quantified. Finally, linear regressions were formed for the every metric based on the logarithmic transformation and the Coefficient of Determination and Fractal Dimension were computed as well. The coefficient of determination for all measures of diversity was zero and for other measures in two general categories: high sensitivity (R- redundant) and without sensitivity (N- no effect). Results acquired from the two indicators were perfectly delineated the sensitivity of landscape metrics to the scale effect (Coefficient of Determination) as well as the direction and magnitude of the landscape metrics (Fractal dimension).