现有的生态系统模型在模拟干旱胁迫下森林生态系统总初级生产力(Gross Primary Productivity,GPP)时存在较大的不确定性,需要探索提高模型解释干旱对GPP影响机理能力的方法。以2011年、2013年和2015年安吉毛竹林通量塔观测数据、MODIS归一化植被指数(MOD13Q1)为数据源,采用结合涡度相关技术的光能利用率模型EC-LUE(Eddy Covariance-Light Use Efficiency)开展毛竹林GPP模拟,并提出提高模型模拟干旱条件下GPP精度的方法。研究结果表明:与非干旱条件相比,干旱条件下GPP与日均温度(Temperature,T)和水蒸气压差(Vapor Pressure Deficit,VPD)的相关关系发生显著改变,由正相关转变为负相关。GPP随T增加而降低的临界温度为15.0 ℃。GPP随VPD增加呈指数递减关系。将VPD添加到EC-LUE模型(EC-LUEVPD)显著提高了干旱条件下GPP预测精度,2011年和2013年干旱期间GPP模型模拟误差分别降低了7.70%和13.74%。ECLUEVPD模型的预测精度得到明显改善,提高了模拟干旱条件下毛竹林GPP的能力,为模拟气候变化背景下竹林 GPP提供可行的方法。
英文摘要:
Existing ecosystem models are relatively uncertain in estimating forest ecosystem gross primary productivity(GPP)under drought stress.It is necessary to increase the ability of models to explain effects of drought stress on GPP.In this study,the model of Eddy Covariance-Light Use Efficiency(EC-LUE)was used to estimate GPP of moso bamboo forest based on EC flux data and Normalized Difference Vegetation Index (MOD13Q1).A way was proposed to increase the accuracy of GPP estimates under drought stress.The results showed that the relationships between GPP and temperature(T)and vapor pressure deficit(VPD)changed from positive to negative.The critical T value was 15.0 ℃ when GPP decreased with the increases of T.There was an exponential decrease of GPP together with an increase in VPD.By including the VPD limitation of GPP in the EC-LUE model(EC-LUEVPD),it was found that the accuracy of GPP estimates significantly increased. The relative root mean square error(RMSEr)for GPP estimates in 2011 and 2013 decreased by 7.70% and 13.74%,respectively. The EC-LUEVPD increased the ability to explain GPP under drought stress. All these provides an effective way to estimate GPP of bamboo under the background of global climate change.