Fig. 2: Comparison of power law estimates to a common solution of Taylor’s Law.
From: Taylor’s law predicts unprecedented pulses of forest disturbance under global change
We compared our power law exponent estimates to a common statistical solution of Taylor’s law (Fig. 2), which assumes the power law to result purely from identically and independently distributed sampling from a skewed distribution. Specifically, we compared our estimates to an analytical solution of this statistical artifact developed by Cohen and Xu26, as well as to independently and identically distributed draws from a log-normal distribution calibrated with empirical means and variances. We found substantially lower power law estimates in our analysis compared to both the analytical and simulation-based statistical solution, which suggests that the observed pattern cannot be explained solely by identically and independently distributed sampling from a skewed distribution.
