A Light Use Efficiency Model Based On Photochemically Reflective Vegetation Index And Its Application

As an important part of the terrestrial carbon cycle,the gross primary prodution(GPP)is an important indicator for quantifying the carbon budget of the biosphere.Accurately estimating GPP is critical to understanding the response of ecosystems to elevated atmospheric CO2 concentrations and the development of relevant policies.This paper estimates the regional GPP and analyses its temporal and spatial patterns in China based on MODIS data,site flux data and meteorological data,combined with improved light use efficiency model based on photochemically reflective index(PRI-LUE model).Based on the analysis of the correlation between PRI and LUE at the canopy scale,this paper introduces the vegetation index and meteorological factors that characterize the canopy structure and hydrothermal conditions to improve the PRI-LUE model.Based on the improved PRI-LUE model,combined with remote sensing data and meteorological data,this paper simulates terrestrial ecosystem GPP from 2006 to 2008 in China,and analyzes its temporal and spatial patterns and the consistency between simulated GPP and global GPP products(GPP_MPI)and sun-induced chlorophyll fluorescence(SIF).The main research contents and conclusions are as follows:(1)Correlation of PRI and LUE at canopy scaleThe selection of PRI reference band will affects the correlation of PRI and LUE:selecting MODIS land surface reflectance band 1 as reference band,PRI is better correlated with LUE,and the R2 of PRI and LUE is 15.3%higher than that of band 10 and 15.5%higher than that of band 12.The correlation between LUE and PRI grouped by observation angle was not improved,and that between LUE and non-grouped PRI was better.There were differences in the correlation of PRI and LUE among different vegetation types,and correlation was better in CRO,GRA and DBF,and poorly correlated in SHB,MF,ENF and EBF.Compared with the choice of reference band and observation angle,the difference of vegetation type has a greater impact on the relationship of PRI and LUE.(2)The construction of PRI-LUE modelRegardless of the vegetation type,the simulated LUE by the linear least squares method(LUE_siml)and the observed LUE have an R2 of 0.35 and an RMSE of 0.40 g C MJ-1.Considering the influence of vegetation type on PRI and LUE.The simulated LUE(LUE_sim2)and the observed LUE have an R2 of 0.48 and an RMSE of 0.36 g C MJ-1,compared to LUE_sim1,the R2 of LUE_sim2 and observed LUE increased by 36.0%,and RMSE decreased by 10.5%.Vegetation index(LAI,LSWI)and environmental factor(Ta)were introduced to analyze the effects of LAI,LSWI and Ta on the relationship between PRI and LUE,and obtain the simulated LUE(LUE_sim3).The R2 of LUE_sim3 and the observed LUE is 0.51,and the RMSE is 0.35 g C MJ-1.Compared with LUE_sim2,the R2 of LUE_sim3 and the observed LUE is increased by 4.8%and the RMSE is decreased by 2.1%.Among the 6 plant types of CRO,DBF,GRA,MF,WET and WSA,LUE_sim3 is significantly higher than LUE_sim2,R2 is increased(ΔR2 is 0.02-0.09),and RMSE is decreased(ΔRMSE is-0.03～0.00 g C MJ-1).After introducing the vegetation index and environmental factors,The ability of simulating LUE with PRI has been improved in various vegetation types and in the whole.Based on LUE_sim3,the PRI-LUE model is constructed,which lays the foundation for the simulation of regional GPP.(3)Temporal-spatial distribution pattern of GPP in ChinaThe average annual total of GPP in China from 2006 to 2008 simulated by PRI-LUE model is 5.79 Pg C yr-1.The distribution trend of GPP is generally the highest in the southeast coast,which in turn decreases toward the northwest inland.The annual average GPP maximum value is mainly concentrated in Hainan Island,Taiwan Island,parts of Guangdong,parts of Guangxi,western and southern Yunnan,and southeastern Tibet.The GPP of these areas exceeds 2000 g C m-2 yr-1.In Jiangsu,Anhui,parts of Henan,parts of southern China and parts of northeast China,GPP values range from 1500 to 2000 g C m-2 yr-1.The average annual GPP minimum value is mainly concentrated in the drought and cold regions of Xinjiang,Qinghai,Tibet,Ningxia,Gansu and Inner Mongolia and the GPP values are less than 200 g C m-2 yr-1.China’s regional monthly GPP total reached the highest in July(1.44 Pg C month-1),the lowest in January(0.05 Pg C month-1),and in the summer(June,July,August)GPP accounted for 63.8%of the whole year,while in winter(January,February,December)GPP accounted for only 3.1%of the whole year.Terrestrial ecosystem GPP in China has obvious characteristics of high in summer and low in winter.(4)Consistency of GPP with GPP_MPI and SIF in ChinaThe annual total of GPP(GPP_sim)simulated by PRI-LUE model and that of GPP_MPI pixel by pixel in China,with R2 of 0.66(p<0.001,n=2265)and RMSE of 461.18 g C m-2 yr-1.In the 99.1%vegetation area of China,the monthly scale GPP_sim was significantly positively correlated with GPP_MPI(R2>0.108,P<0.05,n=36);the monthly average R2 of GPP_sim and GPP_MPI was 0.74.In 91.8%of China’s vegetation areas,the monthly scale GPP_sim was significantly positively correlated with SIF(R2>0.11,p<0.05,n=36);the mean R2 of monthly scale GPP_sim and SIF was 0.52.There was a significant positive correlation between the two GPP totals on the monthly scale(p<0.001,n=36),R2 was 0.93,and the RMSE was 0.14 Pg C month-1.GPP_sim was significantly positively correlated with SIF(R2=0.95,p<0.001,n=36).The correlation of monthly GPP_sim and SIF has similar seasonal differences with the correlation of monthly GPP sim and GPP_MPI:the correlation ranges from strong to weak in autumn,spring,summer and winter.