Phenology has a significant effect on forest growth and directly affects the forest ecosystem carbon cycle. Bamboo forests possess strong carbon sequestration capacities. However, it is not clear whether variations in phenology increase or decrease carbon uptake and storage in subtropical bamboo forests. We first extracted the length of the growing season (LOS) by coupling a data assimilation algorithm and MODIS leaf area index (LAI) data, and then the LOS was used to drive the integrated terrestrial ecosystem carbon-budget (InTEC) model to simulate gross primary productivity (GPP) and net ecosystem productivity (NEP) in Zhejiang Province from 2001 to 2017. Our results showed that the LOS estimation using the assimilated LAI time series was more reliable than that of the MODIS LAI and enhanced vegetation index (EVI). The annual average LOS increased on average by 0.76 day yr⁻¹ from 2001 to 2017. The GPP and NEP simulations based on the LAI assimilation-based phenology indicated that bamboo forest ecosystems possess strong carbon sequestration capacities and act as carbon sinks, with mean annual GPP and NEP values of 434.74 ± 257.93 g C m⁻² yr⁻¹ and 141.42 ± 82.54 g C m⁻² yr⁻¹, respectively, during 2001–2017. An increase of one day in the regional annual LOS increases the annual average GPP and NEP by 1.34 g C m⁻² yr⁻¹ and 0.75 g C m⁻² yr⁻¹, respectively. Moreover, the interannual variation of NEP was significantly correlated with precipitation and temperature, whereas GPP was not. Our results demonstrated that phenology extraction based on LAI data assimilation should play an important role in the simulation of bamboo forest productivity with ecological process models. The variation in phenology induced by climate change can strengthen the bamboo forest carbon sink, which is of great significance for subtropical forests coping with climate change in the future.

物候对森林生长具有重要影响，并直接影响森林生态系统的碳循环。竹林具有很强的固碳能力。但是，尚不清楚物候变化是否增加或减少了亚热带竹林的碳吸收和储存。我们首先通过结合数据同化算法和MODIS叶面积指数（LAI）数据提取生长季节的长度（LOS），然后使用LOS来驱动陆地生态系统综合碳预算（InTEC）模型来模拟2001年至2017年浙江省的初级生产力（GPP）和净生态系统生产力（NEP）。我们的结果表明，使用同化LAI时间序列进行的LOS估计比MODIS LAI和增强的植被指数（EVI）的估计更可靠。^-1从2001年到2017年。基于基于LAI同化的物候学的GPP和NEP模拟表明，竹林生态系统具有强大的固碳能力并充当碳汇，年均GPP和NEP值为434.74±257.93 g C m在2001– 2017年期间分别为⁻² yr ^-1和141.42±82.54 g C m ⁻² yr ^-1。区域年度LOS增加一天，年均GPP和NEP增加1.34 g C m ^-2 yr ^-1和0.75 g C m ^-2 yr ^-1， 分别。而且，NEP的年际变化与降水和温度显着相关，而GPP则没有。我们的结果表明，基于LAI数据同化的物候提取在利用生态过程模型模拟竹林生产力中应发挥重要作用。气候变化引起的物候变化可以加强竹林碳汇，这对于今后应对气候变化的亚热带森林具有重要意义。