Study On Characteristics And Influencing Factors Of Carbon Exchange Of Desert Area In XINJIANG

The Taklimakan desert (the largest mobile desert in china) and the Gurbantunggut desert (the largest semi-fixed desert in China) have great impact on local land-atmosphere interaction and regional climate. Research on variations in CO2 flux and its environmental influence factors over different desert area in XINJIANG not only provide prerequisite dataset for carbon budget assessment in desert ecosystem, but also bring advantage to quantify regional and global carbon cycle. In this study, long-term (5-year) CO2 flux and environmental variables were observed at four sites across Taklimakan desert and Gurbantunggut desert in XINJIANG The GOSAT L4 product data were downloaded for CO2 flux comparison and cross-validation. The above data will be used to address three-fold issues:(1) to find the most appropriate method for data post-processing and assessment in XINJIANG desert fluxes observation, (2) to investigate CO2 flux variations and elucidate the underlying environmental control mechanisms (3) to analyze the carbon exchange and quantify carbon sink/source capacity of the two deserts. The main results are as follows:(1) Annual mean of atmospheric CO2 concentration of TAKLIMAKAN desert was 387.3 ppm in 2013. Atmospheric CO2 concentration showed obvious diurnal variation with peak value observed in daytime and nadir value observed at nighttime. The profile of atmospheric CO2 in near surface layer showed "inversion" pattern with high concentration observed at upper level and low concentration observed at low level The CO2 concentration of the desert has significant correlation with air te mperature and wind speed.The annual mean of subsurface CO2 concentration at depth of 40 cm at Tazhong site was 662.73μmol · mol-1, which was 20% higher than the value observed in Xiao tang (562.55 μmol·mo-1).Taklimakan Desert soil CO2 concentration increases with soil water content and soil temperature, but shows negative relationship with atmospheric pressure. With desert ecosystem carbon uptake rate up, Soil CO2 concentration grow up and fallow down from carbon uptake rate decline.(2) Double coordinate rotation can be used for tilt correction for CO2 flux observation over such undulating terrain in the mobile desert. The WPL correction accounting for air density fluctuation contributes to 80% of the CO2 flux 30 min averaged CO2 fluxes can be used to investigate carbon exchange of desert while fluxes at 15min scale are more suitable for studying fine turbulence structure of turbulence heat flux. Spectrum and co-spectrum analysis showed that the open-path eddy covariance system can respond quickly and accurately to high frequency turbulence at two deserts. In addition, current energy balance ratio and footprint area can meet the requirement of eddy flux measurement in desert.(3) At daily and monthly scale, the moving desert area where hinterland of Taklimakan desert behaved as carbon sink. The average daily NEE (net ecosystem exchange) was -0.4 gC m-2 d-1. The monthly NEE reached the maximum in June with-19.0 gCm-2month-1. The average value of monthly NEE was-8 gCm-2 month-1. Hinterland of the Taklimakan desert sequestrated carbon in daytime and reach its maximum sink at noon. To a contrary, the desert released carbon at nighttime but with very smaller magnitude.Similarly, the transition zone between north part of the Taklimakan desert and Populus stands of also absorbed CO2 in daytime and showed weak CO2 efflux at night. At daily and monthly scale, this area acted as carbon sink and average daily NEE of-0.6 gCm-2d-1. The average monthly NEE was -22 gCm-2month-1 and peaked in July.This area sequestrated carbon in daytime and reach its maximum sink at noon. To a contrary, released carbon at nighttime but with very smaller magnitude.The transition zone between north part of the Taklimakan desert and ancient riverbed area of Tarim River behaved like carbon source in April, carbon sink in March and from May to October.In monthly scale, this area also behaved as carbon sink with average of -6.0 g Cm-2mon-1 and occurred peak value in July. The transition zone sequestrated carbon in daytime and released carbon at nighttime but with very smaller magnitude.The maximum sink occurred at 9-10 am. in from March to May, and noon in from Jun to October.Compared with three different land type area of the Taklimakan desert, the moving desert area of hinterland and transition zone of north border to Taklimakan desert behaved as carbon sink in daytime and carbon source in nighttime.In daily and monthly scale, the carbon sink rate of transition zone of sand and Populus greater than mobile desert area and ancient riverbed area.(4)The carbon sink capacity of Taklimakan desert is also confirmed by GOSAT observations. The daily NEE was -0.23 gCm-2d-1 and the maximum was -0.9 gCm-2d-1.(5)Artificial oasis have changed carbon exchange pattern in Taklimakan desert. In growing season, the artificial oasis notably absorbed CO2 from atmosphere with average carbon sink intensity of-20 gCm-2 month-1. On the contrary, the artificial oasis area acted as carbon source in non-growing season with average intensity of 5 gCm-2 month-1.(6) At daily scale, the Gurbantunggut desert behaved like carbon sink in March, April, May, July and September. The average sink intensity was -0.3gCm-2d-1. The desert switched to carbon source in June and August. The average source intensity was 0.2 gCm-2d-1.At monthly time scale, this desert acted as carbon sink in March to May and July with maximum sink intensity of -18 gCm-2mon-1. While this area became carbon source in June, August and September with maximum source intensity of 4.5 gCm-2mon-1.The Gurbantunggut desert absorbed carbon in daytime and the sequestration was more obvious in the morning. Conversely, the desert released carbon in nighttime.(7) Compared with Taklimakan and Gurbantunggut desert, their carbon exchange is different each other in different time scale.In monthly scale,two desert behaved as carbon sink in May and July.In Jun and August, Gurbantunggut desert acts carbon source while Taklimakan desert acts carbon sink.Nocturnal CO2 flux seems strongly carbon sink of Gurbantunggut desert while Taklimakan desert as weakly carbon source.(8) At hourly scale, the main impact factors of carbon exchange are global radiation, net radiation and soil moisture at depth of 5 cm.0.14 m3·m-3 of 5cm soil moisture was a threshold for carbon exchange of this area.At daily scale, air temperature, global radiation, net radiation and soil skin temperature are the most important influencing factors. At monthly scale, air temperature, global radiation, net radiation, soil skin temperature, soil moisture of top layer and VPD (Vapor Pressure Deficit) become main control factors of carbon exchange.0.04 m3·m-3 of 5cm soil moisture was a threshold for carbon exchange of this area.In Gurbantunggiit desert, the influence of environmental factors to carbon exchange is more significant at daily scale. The daily air temperature,5cm soil moisture and temperature, net radiation, VPD are main determinant of daily carbon exchange.0.1 m3·m-3 of 5cm soil moisture was a threshold for carbon exchange of this area.(9) The abiotic process and air-soil CO2 concentraion gradientes in surface boundary layer were considered to be the major causes of such desert carbon uptake.