Nonlinear parsimonious forest modeling assuming normal distribution of residuals

Abstract To avoid the transformation of dependent variable, which introduces bias when back-transformed, complex nonlinear forest models have parameters estimated with heuristic techniques, can supply erroneous values. The solution for accurate provided by Strimbu et al. (Ecosphere 8:e01945, 2017) 11 functions (i.e., power, trigonometric, and hyperbolic) is not based on heuristics but could contain a Taylor series expansion. Therefore, objectives present study are to unbiased estimates variance following predicted variable identify an expansion that does induce numerical mean variance. We proved in expectation depends illustrated new modeling approach two problems, one at ecosystem level, namely site productivity, individual tree stem taper. unbiased, more parsimonious, precise than existing less parsimonious models. This focuses research methods, be applied similar studies other species, ecosystem, as well behavioral sciences econometrics.