Carbon Dynamics Simulation In The Eastern Part Of Northeast Forest Ecosystems In China Using Ibis

Integrated Biosphere Simulator (IBIS) is designed to integrate a variety of terrestrial ecosystem phenomena within a single, physically consistent model that can be directly incorporated within AGCMs. It belongs to a new generation of dynamic global vegetation models (DGVMs), which can simulate complexity and long time biosphere processes. Also IBIS represents the future trends of carbon model development.Based on original IBIS framework, we simulated net primary production (NPP), soil respiration (Rs) and climate conditions (temperature and precipitation) in Maoer Mountain Natural Reservoir Station and compared these simulations with observed data set, thus parameters and biogeochemical processes in IBIS were modified. Scaling-up from point scale to regional scale, forest ecosystem carbon dynamics and climate conditions were simulated in Northeast China by the modified IBIS, also temporal and spatial distribution variations were evaluated. Under model sensitivity analysis, climate scenarios were set here to evaluate effect of climate change and vegetation cover (LAI) on carbon balance according to the third authority report of IPCC. Main results are listed as below:1. Model modification. Eight plant functional types were derived form the original twelve, as well as seven vegetation types from fifteen according to the vegetation characteristics of our study area. Soil spin-up procedure in belowground carbon&nitrogen cycling module is not considered in this modified IBIS. Litter carbon pools were evaluated directly to make a model simplification and totally accorded with the real conditions of soil balance.2. Model validation. IBIS simulated well with the precipitation and temperature in growing season, but little agreement with climate variety in late spring and winter. Modeled errors of NPP, GPP and soil respiration in seasonal dynamics were smaller than these reported literatures. Overall, modified IBIS can represent the real dynamics of NPP, GPP and soil respiration in Northeast China and applied carbon simulation in the study area.3. Rs. Two years of Rs variation in2004and2005were simulated, results showed that:Rs had a great seasonal pattern since the highest Rs in summer, persisting52-57%of the total yearly Rs value, the lowest in winter. Contribution ratio (RC) of autotrophic respiration (Ra) showed a clear seasonal patterns also, which can be expressed by a compound function. RCmin occured in April and May, while RCmax in October. RC was much higher than30%reported by Raich. Compared to heterotrophic respiration (Rh), Ra persisted a higher value. Simulated Ra falls in0.122-0.663kg Cm-2a-1conducted by a lot of temperate forests, Rh, was much lower compared to0.31-0.692kg Cm-2a-1, which maybe the reason for low soil respiration simulation. Rs in different ecosystems showed a great variation, broadleaf-conifer mixed forest persisted the highest value of Rs, while larch forest the lowest. Rs spatial pattern was dependent on the spatial variations in T and P. When P was less than465mm, and T was less than-3℃, ecosystem respiration was restricted.4.NPP. Two years of NPP variation in2004and2005were simulated, results showed that: NPP had a great seasonal pattern since the highest NPP in summer, persisting53-70%of the total yearly NPP, the lowest in winter and the most obvious change in spring. NPP values in2005were much higher than that in2004, with different changing expend in various vegetation types. Broadleaf-conifer mixed forest had the highest NPP,0.89kg C m-2a-1, mixed forest the lowest,0.68kg C m-a-1. NPP to GPP ratio for evergreen coniferous forest fell in0.36-0.56, broadleaf in0.25-0.56, which accorded well with Waring et al.1998. NPP spatial pattern was dependent on the spatial variations in T and P. When P was less than460mm, and T was less than-3℃, vegetation growth was restricted. Broadneadle leaf mixed forest maintained the biggest NEP, with the avegrage ranging from0.72to0.74kg C m-2a-1, while hardleaf forest the least, ranging from0.47-0.53kg C m-2a-15. Sensitivity analysis.1) LAI. Responses of NPP, Rs, Ra, Rh, to the variations of LAI input showed that NPP decreased with LAI increasing or decreasing, as Rs, Ra, and Rh, on the contrary.2) Temperature. Vegetation growth can be accelerated by temperature increasing, but exhibit a threshold (+3℃observed in our research). NPP was not sensitive to temperature various, with sensitivity index (SI) less than0.1. Responses of Rs, and Ra to temperature were mediate with SI between0.1and0.2. Rh was very sensitive to temperature variations with SI increasing first and then decreasing, SI were between2.07-2.61.3) Precipitation. NPP was more sensitivity to precipitation decreasing than to increasing, with SI more than0.2. Rs in different species was more sensitive to precipitation various with+10%and-5%, with higher SI in coniferous forest than in broadleaf forest. Response of Ra to precipitation variations was less than Rh, but SI degree reached media or maximum. The different responses of Ra and Rh to precipitation, explained a diverse requirement of Ra and Rh to precipitation, which gives a big challenge in regional soil respiration simulation.6. Carbon dynamic simulating under the future climate scenarios.1) Temperature increased with precipitation decreased. NPP decreased under this climate scenario.2) Both temperature and precipitation increased, NPP for most vegetation types were increased, with evergreen conifer, broadleaf-conifer and larch forest except. Rs increased under both of the above climate scenarios. Climate change had more effect on Rs than on NPP, this agreed well with the result of Griffis et al.