Simulation Of Atmospheric CO₂ Concentration By WRF-STILT Model

CO₂,as the main greenhouse gas,not only contribute to global warming,but also deeply affect earth's climate system.Its concentration can reflect the composition change of atmophere and indicate the process of CO₂ exchange between land and atmosphere,Making it valuable for police making and enviromental accessments to quantify CO₂ flux at regional scale?102-106 km2?and lager scales.And this method to predict future C02 concentration changes is based on high-precision and continuous atmospheric CO₂,along with atmospheric transport models.Cities and agricultural ecosytems play an vital role in global carbon cycle as carbon source and sinks,seperately.The agricultural ecosytems had more homogeneous underlying surface and its atmospheric CO₂ was controlled by NEE,while cities had more inhomogeneous surface,and was mostly affected by anthropogenic emissions.In our study,2 typical underlying regions,which are located in agricultural area?U.S.Corn Belt?and city?Nanjing?seperately,were chosen to observe and model atmospheric CO₂ concentration.For the agricultural observation site,it's in the world largest corn growing area,and provided tall tower multiple CO₂ observations?2007-2016?as eddy covariance for flux,CO₂ concentration and isotope informations,Nanjing city was chosen as the urban city site,which was located in Yangtze River Delta region.In this study,WRF-STILT transport model and CO₂ flux at high spatialtemporal resolution was used to simulate atmospheric CO₂ concentration for these 2 typical underlying surface,and EC observed CO₂ flux along with bayesian inversion method was used to evaluate prior CO₂ flux.What's more,the composition of CO₂ concentration enhancement,tall tower concnetration footprint was also investigated,the major results are as follows:?1?WRF-STILT model can capture strong seasonal and diurnal variations of observed CO₂ concentration at a 100m height in the U.S.Corn Belt.Inter-annual variation of tall tower concentration footprint did not show obvious variations,and about 80%of CO₂ enhancement was contributed by past 12 hours.The model modeled ?¹³C was close to keeling plot derived results,and also found the concentration components can only represent flux components in some certain area,which was strongly advised to use atmospheric tranport model to simulate the main soure area.The RMSE is 19.61ppm,and in the growing season?June-September?,annual average diurnal variation of observation is 20.16ppm,similar with modeld 24.30ppm.the results also shows that even it's in this agriculture domained area,CO₂ enhancement from fossil buring can reach to 6.61ppm,and to some extent it reflects that fossil emissions may be overestimated or the NEE was underestimated.What's more,we also find it's necessary to ued high spatial resolution fossil emission data,which indicats that modeld CO₂ enhancement from low spatial resolution is only 0.4 times that of high resolution.The uncertainty of underlying surface or the difference in modeld NEE can lead to large bias in CO₂ enhancemnt?-12ppm?,this is the potential reason of overestimation in recent study.It is also found that modeled CO₂ enhancement from fossil and NEE CO₂ flux has strong negative relationship with ?¹³C and ?¹⁸O in heating seaon and growing season,respectively.The results relied on eddy covariance method and bayesian method indicated overestimation of CO₂ flux.And isotope information was suggested in future carbon flux studies.?2?At the Nanjing observation site,where the fossil had high spatial heterogeneity,special CO₂ seasonal average diurnal variation and high CO₂ concentration can still be modeled.And when comparing with diurnal variation od CO₂ emission,the boundary layer height was the main controlling factor for the diurnal variation of atmospheric CO₂ concentration.The area that had effect on 34m observation site covered the whole Yangtze River Delta region,while the east of Anhui province and west of Jiangsu province had much larger effect on it.On the basis of CO₂ background,85%of CO₂ enhancement was contributed in the past one day.And it was also found that the conponents of CO₂ concentration can not represent the conponents of regional CO₂ flux at certain extent,especially at the area that has high spatial heterogeneity.Footprint should be considered in these related analysis.?3?Modeled anthropogenic CO₂ concentration enhancement in urban was about 20ppm in spring and summer,while it could reach 300ppm in autumn and winter.When compared with agricultural area,the anthropogenic CO₂ concentration enhancement was less than 7ppm.It's also found that even annual net ecosystem exchange in Corn Belt is negative?carbon sink?,while its concentration enhancment was positive,it was caused by difference of boundary layer height in daytime and nightime.That in the nighttime,which had much lower boundary layer height,the enhancement was larger even it had the same flux?opposite sign?.The extent of observed CO₂ concentration was between 350 and 420ppm,and it will drop below background in the growing season.And when it's in Nanjing,the concentration in the whole year of 2014 was between 400 and 550ppm,and did not had the seaonal variation like that in agricultural area,and the WRF-STILT model can well simulate the two different concentration change character.