A Study Of Nighttime Ecosystem Respiration And Ecosystem Respiration Model In A Semiarid Stipa Grassland Of Xilin Gol

Under the background of global change,it is particularly important to estimate carbon reserves of grassland ecosystem in China and analyze the contribution of vegetation to global climate change and ecological value.In this paper,based on the observed data of near-stratum turbulent flux of semi-arid stipa in Inner Mongolia from 2011 to 2013 from xilinhaote national climatic observatory,the data of soil temperature,soil moisture and conventional meteorological observation data and satellite remote sensing data,etc.,the characteristics of ecosystem nocturnal respiration in semi-arid stipa grassland in Inner Mongolia were analyzed,and the environmental factors that driving ecosystem respiration were discussed.The ecosystem respiration model was established,improved and applied by using the key factors affecting ecosystem（the data of growth season in 2010 were also used to establish the model）.The main conclusions are as follows:（1）Respiration of this ecosystem R_eco showed a single peak change from 2011 to 2013,with significant interannual differences.The nocturnal annual average values were 0.026,0.045 and 0.042mgCO₂·m^-2·s^-1,respectively.There was no significant difference in the size of Reco in non-growing season,and the interannual differences mainly came from the differences in growing season.（2）Soil temperature and temperature of ecosystem showed single-peak change,and there was little interannual difference;seasonal dynamics of soil moisture content had no obvious regularity,and there were obvious interannual differences,and the growth season was significantly larger than the non-growth season;precipitation had obvious interannual differences,mostly occurred in the growth season,the proportion of precipitation in the growth season of 2011-2013 accounted for 0.79,0.85 and 0.85 of the total annual precipitation,respectively.Enhanced vegetation index（EVI）showed a single peak change,land surface water index（LSWI）could not express the water condition of ecosystem in the non-growing season with snow cover,and there were interannual differences in the growing season,similar to EVI;Land surface temperature（LST）showed a single peak change,and there was no obvious interannual difference.（3）In the growing season,R_eco is exponentially correlated with soil temperature at 10cm underground(T₁₀),and is linearly correlated with soil moisture content at 10cm underground(VWC₁₀).Respiration is affected by water stress.When VWC₁₀>20%,the correlation between R_ecoco and T₁₀ is better;On the monthly scale,the monthly average R_eco is linearly correlated with the monthly accumulated precipitation;The monthly average R_eco was exponentially correlated with the monthly mean temperature T_air;LST had a better response to soil temperature T₁₀,LSWI could not well reflect the size of ecosystem VWC₁₀;R_ecoco had a good correlation with LST,LSWI and EVI in the growing season,and had an exponential correlation with LST,a linear correlation with LSWI,and a linear correlation with EVI,that is to say,using MODIS remote sensing data to retrieve R_eco in the growing season is completely reasonable and feasible.（4）T-P respiration model,improved Q₁₀ respiration model and remote sensing respiration model based on remote sensing data can well inverte the respiration of semi-arid stipa grassland in Inner Mongolia.The simulation effect of T-P respiratory model will be affected by drought stress;Q₁₀ respiratory model improved by EVI has obvious optimization,and the determination coefficient R² has increased by about 10%;The remote sensing respiratory model based on pure remote sensing data performs well in the study area,and the correlation coefficient between model simulation value and observation value is higher,and the correlation between R_eco observation value and EVI_s·W_s is higher.It is reasonable and feasible to use EVI_s·W_s as an important driving factor of respiratory model.