Phenology-explicit remote sensing estimation of V25cmax spatiotemporal dynamics and uncertainty in Moso bamboo forests

The maximum rate of carboxylation at 25 °C (V25cmax) is a fundamental photosynthetic parameter critical for estimating carbon fluxes in terrestrial ecosystems. However, its phenological variability is often oversimplified in current terrestrial biosphere models, resulting in substantial uncertainty in carbon cycle assessments. As a dominant forest type in subtropical China, Moso bamboo (Phyllostachys edulis) exhibits a distinctive pattern of “explosive growth” and pronounced phenological rhythms, which collectively lead to complex spatiotemporal variations in V25cmax. Despite its ecological importance, remote sensing–based studies on V25cmax in Moso bamboo forests remain limited particularly regarding high-accuracy estimation methods capable of explicitly capturing phenological dynamics across different growth stages. Furthermore, there is a notable gap in the systematic quantification and interpretation of the sources of estimation uncertainty and their dependence on phenological phases. Using Moso bamboo forests in Anji County as a case study, this research integrates Sentinel-2 imagery with in situ measurements to develop phenology-specific models coupling between chlorophyll content (Cab) and V25cmax. High-resolution Cab and leaf area index (LAI) were retrieved using the PROSAIL model combined with a look-up table (LUT) approach, enabling the estimation of V25cmax spatiotemporal dynamics from 2021 to 2023. Based on a covariance propagation framework the study further quantified the contributions of multiple uncertainty sources—including reflectance measurements, Cab/LAI inversion, and Cab-V25cmax model parameters—to the overall uncertainty in V25cmax estimation, and assessed their variations across phenological stages. The results show that:(1) A significant, phenology-dependent coupling exists between Cab and V25cmax (R2 > 0.6), with both the functional form and sensitivity of the relationship varying considerably among phenological stages-underscoring the necessity of stage-specific modeling.(2) The estimated V25cmax exhibits clear seasonal dynamics, peaking during the main growth period (MGP) and declining during the bamboo shoot emergence period (BSP) and dormancy periods (DP), consistent with the characteristic growth rhythm of Moso bamboo. Moreover, V25cmax displays marked spatial heterogeneity—manifested as a “low-center, high-edge” pattern—and notable interannual variability. (3) Uncertainty analysis reveals that the dominant error sources arise from Cab and LAI retrieval uncertainties and model parameter uncertainties, together with their correlated terms. At the leaf scale, model parameters and their correlations dominate during BSP and RGP, exceeding 70 %, whereas Cab-related uncertainty becomes most influential in DP. At the canopy scale, Cab and LAI uncertainties and their correlated terms prevail from MGP onward, while reflectance uncertainty remains negligible (<0.01 %). This study is the first to explicitly identify the primary sources of uncertainty in V25cmax estimation and their phenological dependencies. It establishes a high-resolution, phenology-aware framework for mapping V25cmax in Moso bamboo forests, offering valuable insights for improving carbon cycle modeling, enhancing our understanding of bamboo forest responses to climate change, and supporting sustainable regional forest management.