Evaluation of five models for constructing forest NPP-age relationships in China based on 3121 field survey samples

Abstract. Forest net primary productivity (NPP), representing the biomass carbon gain from the atmosphere, varies significantly with forest age. Reliable forest NPP-age relationships are essential for forest carbon cycle modelling and prediction. These relationships can be derived from forest inventory or field survey data, but it is unclear which model is the most effective for simulating forest NPP variation with age. Here, we aim to establish NPP-age relationships for China’s forests based on 3121 field survey samples. Five models, including the Semi-Empirical Mathematical (SEM) function, the Second-Degree Polynomial (SDP) function, the Logarithmic (L) function, the Michaelis-Menten (M) function, and the Γ function, were compared against field data. Results of the comparison showed that the SEM and the Γ function performed much better than the other three models. SEM also outperformed the Γ function in tracking forest NPP-age curves at old ages and therefore is regarded as the best NPP-age model. The finalized forest NPP-age curves for five forest types in six regions of China can facilitate forest carbon modelling and future carbon projections in China and may also be useful for other regions.