An analysis of some diverse approaches to modelling terrestrial net primary productivity

摘要: We inter-compare the mathematical formulation of ten models terrestrial net primary productivity (NPP) and their functional responses to temperature (T), carbon dioxide (CO2), soil water (W) photosynthetically active radiation (PAR). The span a broad spectrum complexity approaches from original, empirical Miami model, through β-factor, global-average box models, dynamic global vegetation (DGVMs) TRIFFID BIOME3. Five separate photosynthesis respiration, although only three directly consider biochemistry. Equations for all are given in complete, compact, standardized format. NPP differ markedly: β-factor show tiny increase with temperature, predict modest S-shaped response saturating above 35 ◦ C (a surrogate drought effects), biochemical quasi-biochemical peaked an optimum range 15–25 except DEMETER (<10 C). Qualitative differences whether respiration exhibits or exponential have effect on NPP. CO2 is qualitatively similar but one model that misrepresents it. Where stress represented, it same where declines under high moisture. PAR saturating, saturation point differs considerably. When two environmental variables combined, second variable typically just scales first. In increases CO 2, also noticeably derivative different plant types (PFTs) 4 grasses greatest maximum at higher than any C3 PFT. Boreal, needle-leaf trees generally lowest Additional uncertainties modelling stomatal conductance, light-use efficiency, scaling canopy level discussed. suggest concerted effort standardize definitions, notation, units PFTs would add degree transparency currently lacking modelling. © 2003 Elsevier B.V. All rights reserved.