【作者】 郭维华；

【导师】 康绍忠；

【作者基本信息】 中国农业大学 ， 农业水土工程， 2014， 博士

【摘要】 在全球气候变化日益加剧的环境下,陆地生态系统碳循环的研究已成为热点。绿色植物是碳循环过程当中最重要的组成成分,尤其是自然系统较为脆弱的区域,植物的功能更加明显。中国西北早区生态系统脆弱,葡萄作为西北旱区广泛种植的经济果树,对当地的大气碳库起着重要作用,其碳通量有着独特的变化规律和机理,其研究对于认识整个陆地生态系统的碳循环过程有着重要意义。本研究利用涡度相关系统,于2008年至2013年在我国西北旱区甘肃武威地区葡萄园进行了6年的CO2通量观测试验,分析了不同时间尺度下葡萄园CO2通量的变化规律,确定其碳源/汇属性及强度；区分出不同时间尺度下影响碳通量变化的主要因子；通过对比实测土壤碳通量数据,得到了适宜模拟西北旱区土壤碳通量的模型；在以上研究基础上,通过对比分析,建立了适宜模拟葡萄园碳通量的理论模型。具体研究结果如下：(1)葡萄园生态系统2008年至2011年的净CO2交换量分(NEE)别为-820、-824、-961以及-992gC.m-2·y-1,总体呈明显的碳汇。葡萄园生态系统NEE呈明显的季节变化规律,在萌芽期葡萄园整体表现为弱的碳源,而到新梢生长期以后转变为碳汇,并且固碳量持续增大,至生育末期开始减小。其NEE日变化规律呈双峰型曲线,午间会出现“光合午休”现象,灌溉会减轻“午休”现象的发生,使其峰值持续的时间增长。(2)在不同时间尺度上,影响葡萄园生态系统净CO2交换量的主要因子各不相同：小时尺度,冠层导度和净辐射是主要影响因子。在日尺度上,冠层导度、温度以及饱和水汽压差为主要影响因子。月尺度上气温、冠层导度以及大气CO2含量对NEE影响最明显。(3)温度模型和Q1o模型由于仅考虑温度对土壤碳通量的影响,因此在模拟时效果并不理想,R2分别为0.10和0.32。本研究通过考虑温度和水分两方面的因素,建立了改进的水分温度综合模型(W-T模型),其模拟效果得到明显提高,R2达到0.78。(4)常用的碳通量模型SMPT-SB和Crop-C模型在模拟西北旱区葡萄园碳通量变化时,效果均不理想,模拟值分别高估60%和30%。在考虑葡萄园的特殊性：植被和土壤对碳通量贡献的差异性后,结合SMPT-SB模型与土壤的综合模型,本研究建立了适宜模拟葡萄园碳通量的SMPT-SW模型,其模拟效果有显著提升,模拟值相较实测值高估约8%,R2达到0.85。(5)根据IPCC第五次工作报告中提出的两种碳排放情景,利用SMPT-SW模型对西北旱区葡萄园2015至2065年的植被初级生产力(NPP)进行了预测。预测结果表明在低碳排放情景下,2065年葡萄园NPP可达1299gC-m-2·y-1,是2015年的1.25倍。其中2015至2020年间为快速增长阶段。在高碳排放情景下,2015至2030年间为快速增长阶段,该15年间葡萄园NPP增长了24%,2030至2065年间葡萄园NPP却开始下降,到2065年时下降至1213gC-m-2·Y-1。更多还原

【Abstract】 On the premise of global change, carbon flux cycle of terrestrial ecosystem has become a research focus. Green vegetation is the most important part of carbon cycle; it has important effect on atmospheric carbon pool. Moreover, it is a fragile ecosystem in arid regions of Northwest China, and wind grapes are widely grown there, so the vineyard plays an important role for the local atmospheric, research on it has significance means for understading the whole carbon cycle of terrestrial ecosystem.Based on above reasons, carbon flux was measured using eddy covariance system in a wine vineyard in arid northwest China during2008to2013, and analysis the dynamic changes of vineyard’s carbon flux under different time scales, and it’s carbon source/sink properties; then distinguish the main factor of the carbon flux under different time scales; by comparing the different types of model, getting the proper model for simulation local soil carbon flux; and based on the above, by comparing different types of carbon flux model, establishes a sutable model for simulation the frame vegetation ecosystem carbon variation. The main results are as follows:(1) The net ecosystem CO2exchange were-820,-824,-961, and-992gC-m-2-y-1form2008to2011respectively, and showed a significant carbon sink. And vineyard NEE was positive value at the early growth stage, higher negative value at the mid-growth stage, and low negative value at the later growth stage. Diurnal variation of NEE was "W" shaped curve in sunny day, but "U" shaped curve in cloudy day.(2) The main factors affecting hourly NEE were canopy conductance (gc) and net radiation (Rn). The main factors affecting daily NEE were gc, air temperature (Ta) and vapour pressure deficit (VPD). The main factors affecting month NEE were gc, air temperature (Ta) and atmospheric CO2density.(3) The effect of temperature and Q10soil carbon models in the simulation are not good, with R2of0.10and0.32respectively, because they only consider the influence of temperature on soil carbon flux. In this study, we established an improved water-temperature integration model (W-T model) by considering both temperature and moisture factors, and the simulation results have been improved obviously, R2reached to0.78.(4) When using the mainly ecosystem scale carbon flux model to simulate the vineyard carborn flux variation, such as SMPT-SB and Crop-C model, the results are not as expcted, the simulation are overvalued by60%and30%. This is due to not considering the plant and soil have different carbon flux varation in the frame vegetation ecosystem. After considering this situation,, we establishes a new ecosystem carbon model called SMPT-SW combined with the SMPT-SB model and W-T soil carbon model, to simulate the vineyard’s NEE. The simulation result has a significant increase compared with above two models, the simulation is overvalued about8%, and R2is0.85.(5) According to the two kinds of carbon emission scenarios proposed by IPCC5th working report, using the SPMT-SW modle to predicte the net primary productivity of vineyard in arid regions of Northwest China from2015to2065. The results showed that in low carbon emission scenario, the vineyard’s NPP will be1299gC-m-2-y-1at the year of2065,1.25times that of2015. From 2015to2020among them, the NPP growth is rapid. Meantime, under the high carbon emission scenario, it will be a rapid growth stage for vineyard’s NPP from2015to2030, the NPP increased by24%over15years, but after2030, the NPP will begin to decline, untill the year of2065, the NPP will fell to1213gC-m-2·y-1.更多还原