Simulating And Predicting Plantation Carbon Dynamics In Consideration Of Forest Disturbance And Recovery Histories

The carbon cycle of ecosystems has a decisive impact on climate change.Forests are the largest carbon pool in terrestrial ecosystems and play an essential role in achieving the goal of carbon neutrality.The carbon sink effect of plantation has been considered to be a possible mechanism for mitigating global warming,and accordingly has become one of the vital directions of global change studies.The research topics on the impacts of forest disturbance and recovery pattern changes on carbon sequestration capability,and carbon cycle mode variations on climate change have been a focus of different-level governments and academia.The overall objective of this work was to develop a comprehensive framework that integrates remote sensing and process models to improve carbon accounting accuracy,and to realize the accurate identification of carbon source and sink characteristics in Southern China's plantations at an annual time step and a 90 m spatial resolution,with an emphasis on three typical plantations:Chinese fir(Cunninghamia lanceolate),Oak(Quercus L.)and Pine(Pinus massonia)in Lishui city,Zhejiang province.The major research contents and results were summarized as follows:(1)Using long-term Landsat time series observations and Vegetation Change Tracker model(VCT),a historical data set on forest disturbance and recovery events in the study area from 1989 to2019,was first created or mapped.Based on it,defining three forest age groups including VCT-IPR(Within Period's disturbance and then Recovered),VCT-PPR(Pre-Period's disturbance and then Recovered)and VCT-PSF(Persisting Forest),followed by proposing a three-tiered conceptual framework for estimating the ages of the three types of forest.For VCT-IPR meaning that forest disturbances occurred with the time window of Landsat time series,its age was calculated by subtracting the year of being disturbed from the year of spectrally recovery,both directly read from the Landsat time series stack,plus two years of the seedling age used for afforestation.For VCT-PPR meaning that forest disturbances occurred before the start year of the Landsat time series stack,1989,its age was assigned as the average spectral recovery years of the three dominant tree species in the study area.For VCT-PSF meaning that no disturbances occurred within the time window of the Landsat time series stack,its age was estimated by modelling the sample's ages from the national forest inventory with diverse spectral and textural indices derived from remote sensing observations using random forest in conjunction with recursive feature elimination method.Then,a spatial integration(overlay)process of the three estimations was done to gain an annual map of forest age,followed by a recurrence process to generate an inter-annual map set of forest age in the entire study area.By analyzing the changes in forest disturbance and recovery areas in various counties and cities in Lishui City and their causes,it was found that the main cause responsible for the sudden forest changes and forest quality was forest logging and extreme weather(such as snow and high temperature events).Post-disturbance artificial regeneration and the implementation of forest protection measures were the main reasons contributing to the increase of forest area.The age analysis results showed that between 1989 and 2019,due to the implementation of forest protection policy,the average forest age of the three plantations of Chinese fir,pine and oak increased by 18.98yr,20.81 yr and 21.15 yr,respectively.These established forest age maps can guide forest managers to determine the amount and location of forest harvesting according to the management objectives of different tree species,and can help assess the sustainable use status of forest resources,improve the accuracy of carbon accounting,and accurately predict future development trend of forest stands.(2)Using forest age,biomass carbon density of live standing trees and the relationship calibration data among different carbon pools collected from on-site inventories and literature investigations to drive IBIS(Integrated Biosphere Simulator)model,a 100-year duration carbon density and carbon flux curves of Chinese fir,pine and oak in the study area under the average climate conditions was simulated first,and the carbon flux change trends under three climate scenarios(RCP 2.6,RCP 4.5 and RCP 8.5)from 2020 to 2050 were predicted.The results indicated that forest age was an important factor in controlling forest carbon density,and the average carbon density usually increased with an increased age before overmature.Carbon flux had a certain correlation with annual precipitation and annual accumulated temperature,but the correlation between carbon flux and annual accumulated temperature was higher.The net carbon sequestration capacity of young forests and overmature forests was relatively small,while the net carbon sequestration capacity of middle-age forests was relatively large.Specifically,the carbon density curves of aboveground biomass of Chinese fir,pine and oak all showed that the growth rate was high before reaching the mature forest status,and it gradually slowed down after maturity.The net primary productivity(NPP)ranked as Chinese fir>pine>oak;The change trend of Net Ecosystem Productivity(NEP)was basically same as NPP.They all increased in the early stage of growth,and then gradually decreased with aging.The change trend of Net Biome Productivity(NBP)was that oak tree always behaved as carbon sinks within 100 years(0.15?2.5 Mg·ha^-1·yr^-1),and Chinese fir and pine transferred as carbon sources after 63 years and 73 years respectively.Under the three climate scenarios,NPP,litter decomposition and soil decomposition all increased with the increase of CO₂ concentration,whereas the degenerated process decreased with the increase of CO₂concentration.(3)A LUCAS(Land Use and Carbon Scenario Simulator)model was built for the study area by integrating forest disturbance data set,forest age data,carbon density and carbon flux data.With this modelling,the changes in carbon source and sink trends of the three plantation forest ecosystems in the study area from 1989 to 2050 and the reasons responsible for these changes were examined.Also,the changes in carbon sources and sinks under future climate scenarios and management patterns were predicted.According to different disturbance strategies,different experiments were designed,and the improved model was used to quantitatively analyze the influence of disturbance and climate change on the forest carbon source and sink changes.The results showed that forest age structure reflected the spatial and temporal changes of forest ecosystem disturbances and determined the amount of carbon accumulation.Disturbance not only changed the regional carbon cycle in a short time,but also affected forest regrowth after disturbance.From 1989 to 2019,the total carbon pools of the three plantation ecosystems showed an upward trend(an increase of 0.16?0.22 Tg C),and the highest increase proportion was observed in the above-ground biomass carbon pool.The simulation study of the total ecosystem carbon storage of the three tree species under different climatic scenarios in the next 30 years showed that the total ecosystem carbon storage of oak plantations under the three RCP concentrations was higher than the historical average climatic conditions,and varied with CO₂ concentrations.However,the total ecosystem carbon storage of Chinese fir and pine plantations under the three climate models was less than the carbon storage under the historical average climate conditions.Through a two-factor analysis of variance on the influencing factors of carbon flux,it suggested that there was no interaction between disturbance and climatic factors on carbon flux(P>0.05),and both factors significantly(P<0.05)affected the carbon fluxes of the three main plantation ecosystems in the region in the next 30 years.In this study,by implementing VCT algorithm,global vegetation dynamic model IBIS and the system model LUCAS,the spatial and temporal dynamics of carbon flux and carbon cycle process of the three typical forest ecosystems in Lishui City,Zhejiang Province were clearly characterized.The research methods and results are useful for evaluating the carbon cycle process of forest ecosystems in Lishui City,and they also provide a model basis and theoretical reference for the final establishment of a carbon cycle coupling model suitable for Southern China's forest ecosystems,which can simulate the structure and function of forest ecosystems at different scales,thereby providing a theoretical model for more accurately predicting the contribution of forest carbon sinks to carbon neutrality in the future.