PROCEEDINGS ARTICLE | November 8, 2014
Proc. SPIE. 9259, Remote Sensing of the Atmosphere, Clouds, and Precipitation V
KEYWORDS: Statistical analysis, Data modeling, Satellites, Remote sensing, Vegetation, CCD cameras, Fractal analysis, Analytical research, Spherical lenses, Statistical modeling
Spatial heterogeneity of the animal-landscape system has three major components: heterogeneity of
resource distributions in the physical environment, heterogeneity of plant tissue chemistry,
heterogeneity of movement modes by the animal. Furthermore, all three different types of
heterogeneity interact each other and can either reinforce or offset one another, thereby affecting
system stability and dynamics. In previous studies, the study areas are investigated by field sampling,
which costs a large amount of manpower. In addition, uncertain in sampling affects the quality of field
data, which leads to unsatisfactory results during the entire study. In this study, remote sensing data is
used to guide the sampling for research on heterogeneity of vegetation coverage to avoid errors caused
by randomness of field sampling. Semi-variance and fractal dimension analysis are used to analyze the
spatial heterogeneity of vegetation coverage at Heihe River Basin. The spherical model with nugget is
used to fit the semivariogram of vegetation coverage. Based on the experiment above, it is found,
(1)there is a strong correlation between vegetation coverage and distance of vegetation populations
within the range of 0～28051.3188m at Heihe River Basin, but the correlation loses suddenly when the
distance greater than 28051.3188m. (2)The degree of spatial heterogeneity of vegetation coverage at
Heihe River Basin is medium. (3)Spatial distribution variability of vegetation occurs mainly on small
scales. (4)The degree of spatial autocorrelation is 72.29% between 25% and 75%, which means that
spatial correlation of vegetation coverage at Heihe River Basin is medium high.
