Field Observation And Model Optimization Of Carbon Cycle In Typical Terrestrial Ecosystems

Carbon cycle of terrestrial ecosystem plays an important role in global carbon cycle.The study of terrestrial ecosystem carbon cycle and its response to global change is significant in predicting atmospheric CO₂ concentration and addressing climate change.In this research,field observation data and remote sensing data of Xilinhaote National Climatic Observatory and Jinzhou Agroecosystem Observatory during growing seasons were used to analyze the key environmental factors and biotic factors which drive the ecosystem respiration(R_eco).Then the performances of different ecosystem respiration models were compared to pick out the opportune model for the optimization of Vegetation Photosynthesis and Respiration Model?VPRM?.Furthermore,the optimized ecosystem model was verified in the two ecosystems and its limitations and probable improvement directions were analyzed considering the model mechanisms.Finally,on the basis of the above work,the potential of sun-induced chlorophyll fluorescence?SIF?in estimating gross primary productive in semi-arid grassland ecosystem and maize cropland ecosystem was explored and evaluated.Several major findings are obtained from paper:?1?R_eco is affected by temperature,moisture and biotic factors.The multiplicative respiration model which considering both the environmental factors and the biological factors,can model R_eco of the semi-arid grassland ecosystem and the maize cropland ecosystem better than other models,with a coefficient of determination?R²?of 0.61 and 0.55,respectively.Moreover,this model also shows better effects in simulating R_eco of these two ecosystems,R²between the modeled and the observed R_eco are 0.572 in semi-arid grassland and 0.613 in maize cropland ecosystem.?2?Optimization of the respiration part of VPRM can improve the simulation effect of net carbon exchange?NEE?during night of the growing seasons in the two ecosystems,R² between the modeled NEE and the observed NEE increased from 0.299 to 0.574 in the semi-arid grassland ecosystem and increased from 0.208 to 0.582 in the maize cropland ecosystem.However,in the whole time of the growing seasons,optimized VPRM don't show a significant improvement in modelling NEE,probably for the reason that daytime NEE is mainly dominated by vegetation photosynthesis.?3?The WRF-VPRM model simulate the meteorological elements well and R² between the simulated and observed values of air temperature at 2 m level and downward shortwave radiation were 0.865 and 0.884.Though diurnal peaks of NEE simulated by VPRM and the optimized VPRM were underestimated,the daily variation characteristics of NEE were illustrated well.WRF-VPRM also can reflect the trend of atmospheric CO₂ concentration at Xilinhaote,R² between the simulated and the observed CO₂ concentration is 0.548,with a mean deviation of 13.8 ppm.?4?During the growing seasons of the semi-arid grassland ecosystem and the maize cropland ecosystem,SIF and GPP have significant correlations,their R² of linear relationship are 0.564 and 0.511.SIF can simulate the trend of GPP during the growing season.However,SIF is sensitive to soil moisture in the growing season.When the soil water content is low,the drought stress caused photoinhibition of plant photosynthesis,resulting in some differences in the trends of SIF and GPP,which makes the applicability of SIF in modelling and monitoring GPP greatly affected by the extent of soil drought stress.