Simulating Forest Carbon Budget In China With The Integration Of Remote Sensing And Inventory Data

As the largest carbon pool of terrestrial ecosystems,forests play a dominant role in the global carbon cycle.Up to 2008,China's forest cover is 20.36%,making up 5%of the global forest area.Recent studies indicated that forests in China were acting as a carbon sink during the past decades.However,the magnitude of that sink and its spatiotemporal distribution and influenced factors are still inconsistent.Carbon strength(source or sink)of forests is affected by past disturbance(fire,harvest,and age effect)and nondisturbance(climate change,CO2 concentration,and nitrogen deposition)factors.Therefore,the effects of all these factors must be taken into account when simulating spatial and temporal patterns of forest carbon budget.This study is to estimate forest carbon budget in China and evaluate their responses to different factors by integrating forest inventory data(FID),remote sensing data,and an Integrated Terrestrial Ecosystem Carbon(InTEC)model,hoping to reduce current uncertainties in the terrestrial cycle processes.For these purposes,China's forest biomass carbon stocks(BCS)was updated using seven national forest inventories from 1973 to 2008 in combination with refined continuous biomass expansion factor(CBEF)model.The effects of forest area and biomass carbon density(BCD)changes on BCS dynamics in different inventory periods and regions were identified.Then,a viable approach for combining field observations and remotely sensed forest height was developed to map forest stand age in China,which is a key input in InTEC model.The age map was evaluated using national,provincial,and county FID.The spatial pattern of forest age in China was examined.Finally,this study updated InTEC model using the relationships between net primary productivity(NPP)and stand age established from field observations for major forest types in different regions.Forest stand age,referenced NPP,climate,CO2 concentration,nitrogen deposition,and other spatial data were used to drive the updated InTEC model.Eddy covariance measurements and FID-based estimates were used to compare and validate model results.Annual spatial distributions of carbon budget in China's forests were simulated for the period from 1901 to 2012.The contributions of different factors to forest carbon budget were analyzed.Following conculusions can be drawn:(1)During the period from 1973—1976 to 2004-2008,total BCS of all types of forests in China increased from 4.93 Pg C to 8.12 Pg C with increases of 2.13 Pg C by forest stands and 1.06 Pg C by other types of forests,respectively.Total BCS of forest stands increased by 51.8%,indicating a carbon sink of 0.07 Pg C yr-1 from 1973 to 2008,mainly driven by area expansion.BCS of forest stands increased in all regions but the northeast region during the study period.The increase of BCS in the densely populated south and east regions was mainly driven by the increase in BCD while it was mainly due to the area expansion in the north,northwest,and southwest regions with relatively abundant land resources.BCS increased in all regions from 1999 to 2008.Forest stands in China acted as an average biomass carbon sink of 0.17 Pg C yr-1 from 1999 to 2008.BCD of young and middle-aged forests is largely smaller than the value of mature forests.Currently,about 67%of forest stands in China are at young and middle ages and will have a large carbon sequestration potential in the future.(2)The forest age map in 2005 was produced using remotely sensed forest height and age-height relationships indirectly obtained from field observations for major forest types in different regions of China.The estimated results match the FID estimates very well,with a determination coefficient R2 of 0.87,a RMSE(Root Mean Square Error)of 4 years,and 2-8 years biases for mean forest ages in different regions and a R2 of 0.53 and a RMSE of 12 years for provincial mean forest ages.In provinces with 86.7%of national total forests,the difference between retrieved and FID-based provincial mean ages are less than 10 years.The validation using provincial and county FID also show the significant agreement between retrieved forest age and ground truth,with R2 and RMSE values ranging from 0.16 to 0.32 and 11 to 21 years,respectively(P<0.01).The spatial pattern of forest age in China showed significant heterogeneity.In southern and eastern China,most forest stand age was less than 40 years.Forests more than 120 years old were mainly in the northeast mountains and the southwest subalpine mountains of China,accounting for 4.7%of the total forest area.In southeastern Tibet,Xinjiang,northern Heilongjiang and Yunnan,northeastern Inner Mongolia,and southern Guizhou,forest ages were mostly between 60 and 120 years.The forest age and its standard deviation averaged 43 and 16 years in China.The averaged forest age is 51,46,32,26,59,and 52 years in northeastern,northern,eastern,southern,southwestern,and northwestern China,respectively.(3)Site-by-site comparisons of simulated net ecosystem productivity(NEP)using the updated InTEC model with measured values at 14 ChinaFlux forest sites show that the model generally captures the magnitude of NEP(R2 = 0.75,RMSE = 50 g C m-2 yr-1,n = 37,P<0.01),although in some cases the model underestimates or overestimates the measured values.The simulated estimates of net biome productivity(NBP)are consistent with the FID estimates of biomass carbon sink at the provincial level(R2 = 0.43,RMSE = 2.98 Tg C yr-1,n= 168,P<0.01).These validated results indicate that InTEC model can act as a practically useful tool to simulate forest carbon budget in China.The overall pattern of forest NBP showed an increase trend from 1901 to 2012.The temporal changes of living biomass carbon were similar to the total NBP of forests while the changes of soil carbon lagged behind that of living biomass carbon.On average,China's forests were a relatively stable carbon sink before the 1920s,with the value of 46 Tg C yr-1.Forest NBP had a great fluctuation from the 1920s to the 1970s,showing a weak carbon sink.Since 1980s,the carbon sink of China's forests gradually increased.The average carbon sink was 186.9 Tg C yr-1 during the period from 1982 to 2012,with 68%of the sink in living biomass and 32%in soil.The whole China's forests could sequester 212.7 Tg C yr-1 from 1999 through 2008,or equal 12.5%of the contemporary fossil-fuel CO2 emissions.The spatial pattern of forest NBP showed a good agreement with forest stand age.The forest carbon sink extended from northern and northeastern China to eastern,southwestern,and southern China since 1970s.Forests in northwestern China had a decreasing carbon sink.Up to the period of 2000-2012,there was a general enhancement in China's forest carbon sink,with a high value of 300 g C m-2 yr-1 in north Daxing' anling Mountains,Changbai Mountains,and southwestern China and over 50 g C m-2 yr-1 in eastern and southern China.Total forest carbon sinks of China is 239.3 Tg C yr-1 in this period.(4)Six numerical simulations are conducted to examine the individual and integrated effects of disturbance and nondisturbance factors on China's forest carbon budget.If disturbance factors were omitted for the period of 1901-2012,the simulated carbon sink would be reduced by 25.0 Tg C yr-1.If nondisturbance factors were omitted,the carbon sink would be reduced by 41.8 Tg C yr-1.The contributions of disturbance and nondisturbance factors to NBP varied spatially and temporally across China's forests.In most decades,these factors increased forest carbon sinks,with the largest of 101.3 Tg C yr-1 and 90.2 Tg C yr-1 from 2000 to 2012 contributed by disturbance and nondisturbance factors,respectively.Disturbance effects on NBP were larger than nondisturbance factors before 1950s and after 1980s in northern,eastern,and southwestern China.In northeastern and southern China,nondisturbance effects were generally greater than disturbance factors during the study period.However,if disturbance effects were omitted in this period in northeastern China,the carbon sink would be significantly overestimated(36.9%).Disturbance factors played a leading role in determining the magnitude of NBP in northwestern China relative to disturbance factors.As for the contributions of the individual nondisturbance factors,CO2 concentration and nitrogen deposition increase the forest carbon sink.NBP showed a large interannual fluctuation under climate changes,acting as an average carbon source of-13.3 Tg C yr-1 during the study period.CO2 concentration,nitrogen deposition,climate change,and disturbance factors contributed 21.4,15.4,-13.3,and 25.0 Tg C yr-1 to the averaged NBP from 1901 to 2012 in China,respectively.Disturbance effects were larger than the individual nondisturbance factors in most regions,but the effects of CO2 concentration and nitrogen deposition were dominant in southern China.