The same is true for remote sensing data in which the use of one single scale (grain) might perturbe analysis, leading to misleading results or simply to hidden patterns. From this point of view, multiscale approaches are strongly encouraged. An inappropriate match of satellite spatial resolution and grain size of field data may hide actual spatial heterogeneity with sub-pixel variability remaining undetected (Small, 2004 and Rocchini, 2007). Multiple scales of analysis would therefore increase the probability of detecting a possible correlation between species AP 24534 and spectral variability (Foody, 2004).
From a mathematical point of view, the distribution of species diversity over space is a peculiar case of the so called switched systems, i.e. hybrid systems resulting from both continuous and discrete dynamics with a high amount of different potential variables acting as main drivers of species diversity response (Liu et al., 2014). As it has been demonstrated by all case studies mentioned throughout this paper, spectral information can be a good proxy of diversity; however, care must be taken in using only remotely sensed variables without considering additional multiscale drivers like climate, topographic variables, and biotic interactions.
From a mathematical point of view, the distribution of species diversity over space is a peculiar case of the so called switched systems, i.e. hybrid systems resulting from both continuous and discrete dynamics with a high amount of different potential variables acting as main drivers of species diversity response (Liu et al., 2014). As it has been demonstrated by all case studies mentioned throughout this paper, spectral information can be a good proxy of diversity; however, care must be taken in using only remotely sensed variables without considering additional multiscale drivers like climate, topographic variables, and biotic interactions.