Logistic vegetation trajectories and their implications for land surface phenology

Abstract

Vegetation trajectories describe canopy development and underpin phenological metrics, including the start and end of the growing season. We derive a generalized logistic function (GLF) to represent canopy growth under explicit assumptions: exponential development combined with a Michaelis-Menten-type saturating response yields an ordinary logistic function, and a power transformation introduces asymmetry. Using the GLF as a common mathematical basis, we relate trajectory shape to phenological metrics and date uncertainty. We derive shape-dependent relative thresholds for derivative-defined transition points and recover the 9.18% onset threshold for an isolated ordinary logistic growth curve under the small-slope approximation. For a symmetric double-logistic seasonal trajectory, we derive the realized amplitude, full width at half maximum and a critical dimensionless temporal separation between growth and senescence components at which interior third-derivative extrema merge. Exact curvature-change transition points also depend on the metric assigned to the time and response axes. Local noise propagation and 4,800 nonlinear fits show how trajectory geometry affects date uncertainty: additional shape flexibility can reduce bias while increasing uncertainty in the seasonal tails. Together, these results explain how trajectory shape determines the location and existence of phenological transition points and uncertainty in their estimated dates. The framework supports phenology retrieval that preserves the underlying trajectory and interprets each metric through its mathematical definition and uncertainty.