Estimation Of Vegetation Gross Primary Productivity Using Sun-induced Chlorophyll Fluorescence And Its Response To Drought In China

In recent years,global warming has led to frequent drought disasters,which has had a great impact on the carbon cycle of the global ecosystem.The total primary productivity of vegetation is the total organic carbon fixed by plants in the ecosystem through photosynthesis per unit time.Compared with the green vegetation index,solar-induced chlorophyll fluorescence,as a byproduct directly involved in photosynthesis,is more sensitive to vegetation photosynthesis.Using solar-induced chlorophyll fluorescence to estimate global and regional vegetation GPP has become a new model to estimate vegetation GPP.Therefore,based on remote sensing SIF,this paper uses machine learning method to estimate vegetation GPP in China,and takes standardized precipitation evapotranspiration index(SPEI)as drought index to explore the temporal and spatial characteristics of vegetation GPP and drought in China and analyze the influencing factors of drought.In addition,the response of vegetation GPP to drought and its driving factors were further evaluated.The main conclusions are as follows:(1)Different SIF-GPP models are constructed by three kinds of remote sensing SIF,selected site GPP,other auxiliary data and three methods.The results show that the SIF-GPP model based on random forest and gradient boosting regression tree is better than the SIF-GPP model based on linear relationship.Finally,the SIF-GPP model jointly constructed by GOSIF and the random forest method has the best performance,and based on this model,the vegetation GPP in China is estimated.The annual and seasonal GPP of vegetation in China showed a significant upward trend from 2001 to 2021.The spatial distribution of annual and seasonal GPP in China increased successively from northwest to southeast.From 2001 to 2021,the annual GPP of vegetation in most areas of China showed an increasing trend,and the downward trend was mainly in the western region.(2)Based on the SPEI,the annual and seasonal SPEI showed a downward trend during 1950-2021 in China.The northwest and some southeastern regions of China showed a trend of wetting,while other regions showed a trend of drought,especially in North China.From 1950 to 2021,China’s annual precipitation showed an overall downward trend,and annual potential evapotranspiration showed an overall upward trend.71.27% of China’s drought changes are mainly affected by changes in precipitation,and 28.73% are mainly affected by changes in PET.The regions dominated by PET on the change trend of SPEI are mainly in Tibet,Xinjiang,North China and parts of Northeast China,and the changes in SPEI in the rest of the regions are mainly dominated by precipitation.(3)The correlation between China’s vegetation GPP and SPEI at the annual and seasonal scales is strong,which indicates that drought has a significant impact on vegetation GPP,and the time scale of vegetation GPP’s response to drought has a large heterogeneity in spatial distribution.The SPEI and GPP of different vegetation types have strong correlation coefficients,and the correlation coefficient between SPEI and GPP of grassland is the largest.Compared with grassland and shrubs,the time scale of GPP response to drought in forest and farmland is longer.The correlation coefficient of vegetation GPP and SPEI in different dry and wet climate regions is in the order of semiarid region > arid region > sub humid region > humid region.The response time of vegetation in arid and semiarid regions is smaller than that in sub humid and humid regions.Actual evapotranspiration is the most important environmental factor affecting vegetation GPP response to drought.