| Gross Primary Productivity(GPP)describes the photosynthesis amount in terrestrial ecosystem,which is one of the most important parameters in global carbon cycle evaluation.Remote sensing method provides continuously spatial and temporal data that helps for estimating GPP.Although there are serval series of global GPP remote sensing products,most of them explained around 70% of GPP variance.The main reasons for the high uncertainties of GPP models are:(1)remote sensing GPP models haven't fully considered the environmental effects for photosynthesis;(2)models haven't considered the impact of forest vertical stratification and stand age;(3)remote sensing GPP models haven't fully considered the plant traits of photosynthesis process.This study aimed to solve three main reasons mentioned above and focus on four research topics including: environmental factors,forest canopy stratification,forest stand age,spectra characteristic.The specific themes are:(1)Improved global estimations of gross primary productivity of natural vegetation types by incorporating plant functional type.In this section,we built an improved framework for estimating GPP by considering the growing stress variance of light,temperature and water and the maximal potential GPP in different plant functional types.Then we applied the estimating results from this model and did an inter comparison with three other remote sensing GPP products.The results showed that our model reduced at least 13% uncertainties than MODIS GPP product.(2)Assessing the forest canopy vertical stratification to photosynthesis.In this section,we evaluated effects of vertical stratification on meteorology and plant trait difference for remote sensing GPP estimation.We used a 3-D radiative transfer model and three GPP models for parameterization,and then we built a multilayer GPP model.Results showed that the meteorological factor had less difference during growing season but had high variance at leaf off period.The new model incorporated vertical stratification reduced the underestimation of understory photosynthesis.(3)Investigating the pattern between canopy scale BRDF reflectance and light use efficiency(LUE)in different stand age.On one hand,there is no research for investigating the patterns of LUE in different stand age in a similar environment.On the other hand,canopy reflectance has the information of sun-view effect and photosynthesis variance.In this section,we utilized the canopy scale spectro radiometer for investigating the relationship between photochemical radiation index and LUE and then get the spatial and temporal LUE data in different stands.Results showed that the young stand had higher LUE than middle age and mature forest under hot and dry conditions.This indicated that the large scale GPP estimation at forest area need to consider the photosynthetic capacity variance of stand age.(4)Utilized the satellite based high spatial resolution remote sensing red edge data for GPP estimation.Vegetation red edge is the comprehensive spectra characteristic of vegetation from remote sensing data.However,there is not much study focus on estimating GPP by satellite based red edge data.In this study,we used the Sentinel-2 high spatial resolution remote sensing data for GPP estimation.Results showed that red edge vegetation index(CIr)had higher estimation accuracy than the non red edge indices and it showed more detail information in the spatial heterogeneity area.Thus,the next generation of global GPP estimation needed to consider the high spatial resolution red edge data.The remote sensing GPP products are developing.This study tried to improve the reduced the uncertainties which limited the remote sensing GPP products.We attempted to combine the environmental stress,forest vertical stratification,stand age,spectra characteristics to improve the GPP models that helps to get the next generation of remote sensing based GPP products and for global carbon cycle evaluation. |
PDF Full Text Request