Effects Of Phenological Changes On Carbon Budget In Boreal Forests

Vegetation phenology,as an indicator of global climate change,can affect vegetation photosynthesis and ecosystem respiration by affecting the length of the vegetation's growing season,which in turn affects the carbon budget of terrestrial ecosystems.As the main body of terrestrial ecosystems,forests play a very important role in regulating the carbon cycle of terrestrial ecosystem.The boreal forest is located in a high latitude area and has a vast area,accounting for 14.5% of the total land area and 30% of the total forest area.It is an ideal place to monitor global climate change.At present,the research results of different scholars present different views on the impact of phenological changes on carbon sources / sinks in northern forests.In this paper,the Net Ecosystem Productivity(NEP)of the northern forest vegetation was estimated based on remote sensing data such as MODIS and RSCFLUX model from 2001 to 2017.The observation results of FLUXNET site were used to verify the estimation results,and the temporal and spatial changes of GPP and NEP were analyzed;then the widely used Savitzky-Golay filtering method is used to reconstruct the NDVI time series curve,the Double Logistic curve is used to construct the vegetation growth curve,and the maximum slope method is used to calculate the main parameters of the vegetation phenology: Start of Season(SOS),End of Season(EOS)and Length of Season(LOS),and analyzed the spatial and temporal distribution of the three parameters and the differences between different vegetation types;Finally,the linear correlation statistical method was used to analyze the key parameters of carbon budget NEP and growth Seasonal length LOS response relationship.The results showed that the spatial distribution of GPP and NEP in the boreal forest showed a gradual increasing trend from high latitude to low latitude,and there was a high consistency with the distribution of vegetation types.The average range of GPP is 0-1265 g C·m-2·a-1,and the average range of NEP is-706-415 g C·m-2·a-1.In the study area,on the whole,both GPP and NEP showed a significant upward trend.Among them,the increasing trend of GPP was 2.65 g C·m-2·a-1 per year,and the increasing trend of NEP was increase by 1.1 g C·m-2·a-1.From the pixel scale,most of the pixels have no significant changes in GPP and NEP.Only about 17-18% of the pixels have a significant change trend.Among them,almost all GPP pixels with significant trends have an upward trend,with an average increase of 0-32 g C·m-2·a-1 per year;while in areas where NEP has a significant change trend,64% of vegetation NEP shows an upward trend,with an average annual increase of 0-44 g C·m-2·a-1,about 36% of the vegetation NEP showed a decreasing trend,with an average annual decrease of 0-61 g C·m-2·a-1.From the perspective of different vegetation types,the GPP averages of all vegetation types have an upward trend.In terms of NEP,in addition to permanent wetlands,the NEP averages of other vegetations show different degrees of growth,especially the growth of evergreen needle forests and mixed forests.From 2001 to 2017,the start date of the growing season of boreal forests vegetation was mainly concentrated between the 120 th and 180 th days,from the beginning of May to the end of June.The end date of the average growing season is mainly concentrated on the 280 th to 330 th days,from mid-October to the end of November.The average growing season length(LOS)is mainly concentrated on 110-195 days.The average length of growing season(LOS)of forest vegetation in boreal forests is relatively large,and the average value of LOS generally shows a trend of gradually increasing from high latitudes to low latitudes.In most areas,the start date of the growing season SOS is in advance,with an average annual advance of 0-6 days;in most areas,the end date of the growing season is in a delayed trend,with an average of 0-2 days delayed each year;Season length LOS generally shows an extension trend,with an average extension of 0.3 to 3 days per year.Among them,the average extension of the average LOS of evergreen needle forest and mixed forest is larger,reaching an average annual extension of 0.88 days and 0.78 days.The correlation analysis results of NEP and LOS show that within the area,12% of the pixels show significant correlation and have more obvious spatial distribution characteristics.The areas where the two are significantly positively correlated are located exactly in the interlaced areas of different vegetation types(such as Scandinavian and Russian taiga).In vegetation types such as evergreen needle forest,mixed forest,and deciduous needle forest,NEP and LOS showed different degrees of positive correlation,and R2 reached 0.42,0.44,and 0.32 respectively;while in permanent wetlands,they showed significant negative Relatedly,R2 reached 0.64.