【作者】 程琨；

【导师】 潘根兴； Stephen M. Ogle；

【作者基本信息】 南京农业大学 ， 土壤学， 2013， 博士

【摘要】 工业革命以来人类活动引起的温室气体排放量的急剧增加是引起全球气候变化的主要原因,对各领域的温室气体减排是全球减缓气候变化的主要途径。据估计,全球农业源温室气体排放量占人为温室气体排放总量的10%-12%,而农业温室气体技术减排潜力占全球减排潜力的20%。2005年中国农业共排放温室气体约8.2亿吨二氧化碳当量,占全国温室气体总排放的11%,同时约占全球农业源温室气体排放的13%-16%。所以,农业在中国乃至全球的应对气候变化行动中应当发挥重要的作用。因此,中国农田温室气体排放与减排潜力的计量与评价成为中国农业减缓气候变化的重要课题。本论文从构建中国农田生产及温室气体数据库出发,采用数学统计与模型相结合的研究方法,研究中国农田温室气体排放及其减排潜力的统计计量与模型模拟,进而探讨农田温室气体减排的总体潜力及技术途径,为建立农田温室气体排放计量方法及定量表征,并为国家农田温室气体减排政策及技术选择提供科学依据。主要结果如下：1.本文是基于中国农作物生产数据库、国家耕地监测点有机碳数据库和农田长期试验数据库进行研究的。中国农作物生产数据库是通过对中国农村统计年鉴、中国农业生产成本收益资料汇编、中国水利统计年鉴中数据进行搜集而建立的,主要包括各种农作物播种面积、产量、化肥投入量、农药投入量、农膜使用量、机械燃油用量和灌溉量,该数据库用来计算农作物生产碳足迹。国家耕地监测点有机碳数据库包含299个国家级耕地监测点近20余年的有机碳数据,这些耕地监测点分布在全国各个区域,有着不同的种植制度,该数据库用来分析中国农田表土有机碳近20余年的变化情况,还用来估算农田表土固碳潜力。农田长期试验数据库是通过搜集已发表文献中长期试验数据搜集而建立的,该数据库包括试验点地理信息、种植制度、管理模式、农田投入等农田基础信息,还包括有机碳含量(试验开始前和结束时)、氧化亚氮和甲烷排放量、作物产量等数据。该数据库主要用来估算农田表土固碳潜力、开发DAYCENT甲烷子模型以及验证DAYCENT模型。2.基于299个国家级耕地监测点20余年有机碳数据库,分析了中国农田表土有机碳近20年的变化情况,从而评价了农业发展中土壤固碳趋势。结果表明,全国约80%的监测点有机碳年均相对变化幅度在-1.5%-7.5%之间。整体上,中国农田呈固碳趋势；其中,华北、华东和西南地区农田表土固碳明显。对旱地和稻田两种管理模式下农田有机碳数据分析结果显示,稻田和旱地有机碳含量均呈现显著的增加,而稻田有机碳含量增加的监测点数目占监测点总数的比例高于旱地。证实了我们对中国农田近20多年来的土壤固碳趋势以及稻田固碳明显强于旱地的认识。同时,基于中国农田表土有机碳数据库,本研究计算和比较了中国不同区域、不同类型的农田土壤有机碳储量。中国农田表土有机碳储量为36.44tha-1,其中,西南地区农田有机碳储量最高,为42.96tha-1,而西北地区最低,为25.18tha-1。旱地有机碳平均密度为29.14tha-1,远低于稻田(43.73tha-1)。西南地区和东北地区旱地有机碳储量最高,分别为38.45tha-1和36.43tha-1,而华南地区稻田的有机碳密度最高,为55.97tha-1。3.农作物生产的碳足迹是指在某个作物生长过程中由人为投入的生产资料或者器械使用所带来的总的温室气体排放量,并以碳当量(carbon equivalent, CE)来表示。本研究采用生命周期评价-碳足迹分析研究方法,基于中国农作物生产数据库,分析了农作物生产的碳足迹及其构成的变化趋势。中国农作物生产单位面积的平均碳足迹为0.78±0.08t CE ha-1yr-1,单位产量农作物的平均碳足迹为0.11±0.01t CE t-1yr-1。由肥料施用引起的温室气体排放占总碳足迹的60%,而氮肥施用量的变化可以解释15年来碳足迹变化的97%。1993-2007年15年间,单位面积的碳足迹增加了49%,单位产量的碳足迹下降了21%。尽管作物产量与碳足迹呈极显著的正相关关系,但2003-2007年单位碳投入的作物产量(碳利用效率)呈下降趋势。该结果揭示了中国农作物生产碳成本较高,而有着巨大的减排空间。中国水稻、小麦、玉米和大豆生产单位面积的平均碳足迹分别为2472、794、781和222kg CE ha-1,单位产量的碳足迹分别为0.37、0.14、0.12和0.1kg CE kg-1。旱作作物如小麦、玉米和大豆生产78%的碳足迹来自氮肥施用,氮肥施用引起的温室气体排放包括氮肥生产的间接温室气体排放和氧化亚氮的农田直接排放：水稻生产的碳足迹主要来自甲烷排放的贡献(69%)。不同区域间旱作作物碳足迹的差异主要是由于氮肥施用量的差异,而甲烷排放可以解释85%的区域间水稻碳足迹变异。减少这些作物化学氮肥施用量30%,可以产生每年60Mt CO2-eq.(二氧化碳当量,CO2-eq.,下同)的减排量,证实提高氮肥利用率是减少粮食生产碳足迹的重要途径。4.固碳减排计量是进行减缓气候变化的碳交易机制的基础工作,而适合项目计量的计量方法学是实现交易的基础工具。本研究以测土配方施肥项目为例,从边界和基准线的设定、碳库和关键排放源的选取、固碳减排的计量方法等方面探讨了基于碳交易的固碳减排计量方法学的编制问题,提出了以常规施肥下温室气体排放量为基准线,施肥下作物生长田块为边界,以有效氮施用带来的氧化亚氮排放、稻田甲烷排放和施肥器械的排放为边界内关键排放源,确定肥料配方过程中的温室气体排放为泄漏,选择农田土壤有机碳库作为计量碳库的一整套方法学理论框架,并提出了三种计量方法作为参考。5.采用有机碳饱和极限法和生态恢复法估算了农田生物物理(自然)固碳潜力,并采用情景法预测了国家规划或工程建设计划下农田土壤固碳技术可达潜力。通过有机碳储量现状与有机碳饱和容量或未开垦土壤有机碳储量进行对比得到农田生物物理固碳潜力,而技术可达潜力则通过计算不同良好农田管理模式下有机碳积累速率来实现。饱和极限法估算的中国农田表土生物物理固碳潜力为2.21Pg,恢复法估算的生物物理潜力为2.95Pg。预测实施保护性耕作20年中国农田固碳可达0.62Pg,实施配方施肥项目20年的固碳潜力为0.98Pg,两项措施的综合固碳潜力相当于中国2007年温室气体总排放的40%到60%。因此,农业发展计划下农田固碳在中国减缓气候变化中扮演着非常重要的角色。6.稻田甲烷排放在农业温室气体排放中占较大比例,因此,对稻田甲烷排放的准确预测是稻田温室气体减排的重要依据。本文探讨了生态系统模型DAYCENT用于预测稻田的甲烷排放,根据土壤有机质、温室气体排放数据库的实例数据,本研究开发验证了稻田甲烷排放子模型。该模型通过模拟土壤水热状况、水稻植株生长、土壤有机质分解和甲烷由土壤向大气的排放过程来模拟稻田甲烷化过程。使用97个中国稻田试验点对模型进行了开发和验证,其中25个试验点(91个观测值)被用来进行模型参数化,72个试验点(204个观测值)被用来对模型进行验证。通过对比甲烷排放的模拟值与实测值,表明开发的DAYCENT甲烷子模型可以很好地模拟中国稻田甲烷排放(线性回归的决定系数(R2)高达0.83)。模型灵敏度分析结果表明,该模型对与产甲烷基质数量相关的参数最为敏感。7.运用DAYCENT模型及开发的DAYCENT甲烷子模型,模拟预测了中国农田系统不同管理模式下温室气体减排潜力。首先采用350个农田点位的产量、有机碳和温室气体排放数据对模型进行了验证。模拟值与实测值有很好的线性相关关系(R2值在0.71到0.85之间)。模拟效率最高的是作物产量(0.83),有机碳变化的模拟效率最低(0.65)。继之,用DAYCENT模型对不同管理模式下旱作系统和稻作系统的产量和温室气体排放进行了模拟,并采用单位产量的温室气体排放量对不同管理模式的温室气体减排潜力进行了评价。旱作系统在减少化学氮肥施用、施用有机肥和少耕配合秸秆还田的管理模式比常规管理模式温室气体排放显著的减少,在该模式下的不同轮作系统减排潜力为0.31-0.83Mg CO2-eq.Mg-1。而稻作系统减排潜力最高的管理模式为减少化学氮肥施用配合间歇淹水管理,潜力为0.08-0.36Mg CO2-eq.Mg-1.本论文从数据库构建,到计量和模拟方法的开发和应用,完善了用于评价农田温室气体排放及预测减排潜力的数据库-计量方法-模型预测方法,发展了农业碳计量方法学,构建了农业生产碳足迹计量评价分析框架,并提出了中国农业生产高碳投入与低碳效率特征,揭示了农田生产的温室气体减排潜力及技术途径,为中国农业生产固碳减排提供了科学依据和技术支撑。更多还原

【Abstract】 Anthropogenic greenhouse gas (GHG) emissions have increased rapidly since pre-industrial times, which is a major factor contributing to climate change. Reducing emissions of carbon dioxide (CO2) and other greenhouse gases (GHGs) to earth’s atmosphere and sequestrating soil organic carbon (SOC) in terrestrial ecosystems are identified as two of the most pressing modern-day environmental issues. Agriculture accounted for10-12%of total global anthropogenic emissions of greenhouse gases, and20%of global GHG mitigation potential could be achived from agriculture, which indicates the importance of agriculture in global climate change mitigation. China’s GHG emissions in the agriculture sector in2005were estimated to be820Mt CO2-eq., contributing11%to the nation’s total emissions, and also contributing13%-16%to global GHG emissions in agriculture. Therefore, China’s agriculture plays an important role in global climate change mitigation. Assessment of the SOC dynamics and GHGs emissions accurately in the cropland is important for police makers. The objective of this study was to account and simulate GHG emissions and the mitigation potentials for China’s cropland using various approaches, and then discussed the realization of the GHG mitigation potential. The main results obtained were as follows:(1) Four datasets, included Chinese crop production dataset, national monitoring network dataset and long-term cropland experiments dataset, were built to do the following researches. Chinese crop production dataset includes planting area, crop yield and various agricultural inputs. National monitoring network dataset includes the SOC contents from299sites in the last20years. Long-term cropland experiments dataset were collected from the scientific field studies that reported SOC changes and GHG emissions in Chinese cropping systems under various management practices.(2) Data of topsoil SOC contents from the national monitoring network was used to analyze the SOC dynamics and the sequestration status in China’s croplands. The data set comprises299observations across China’s mainland. The relative annual SOC changes of80%of monitoring sites were distributed between-1.5%and7.5%. Topsoil SOC of China’s cropland was in a general trend of accumulation with a frequency of79.1%, which were mainly distributed in the North, East and Southwest China’s croplands. The difference of land use effect on soil carbon sequestration was significantly, the SOC content of rice paddy was significantly higher than that of dry croplands, and the frequency of increasing SOC content was also greater than that of dry cropland. The topsoil SOC datasets of China’s cropland were used to calculate the SOC densities in different regions and cropland types. The current average SOC density of China’s cropland was estimated as36.44t ha-1. The greatest SOC density of42.96t ha-1was found in southwest China, and the least was in northwest of China (25.18t ha-1). There was a significant difference in present SOC density between dry cropland and rice paddies, with the latter being greater by almost10t ha-1than the former. The greatest SOC density for dry cropland was in south west (38.45t ha"1) and North-East of China (36.43t ha-1), and for rice paddies in south China (55.97t ha-1).(3) Life cycle analysis-carbon footprint (CF) calculation approach was employed to analyze the changes of CFs in China’s crop production and identify the contributions of various agricultural inputs to total CFs. The mean overall CF of China’s crop production was estimated to be0.78±0.08t CE ha-1yr-1and0.11±0.01t CE t-1yr-1,for land use and bulk production respectively. For the duration the data covered, the carbon intensity under cultivation land use was seen to increase since1993. Among the total, fertilizer induced emissions exerted the largest contribution of-60%. Compared to the UK, the estimated overall CF of China’s crop production was higher in terms of cultivation land use. While there was a significant positive correlation of carbon intensity with total production, carbon efficiency was shown in a decreasing trend during2003-2007. Therefore, low carbon agriculture should be pursued, and the priority should be given to improving fertilizer use efficiency in agriculture of China. Then data of cultivation area, grain yield, application rates of fertilizer, pesticide, diesel, plastic film, irrigated water, etc., for the major grain crops in China were collected from the national statistical archive and CF of direct and indirect carbon emissions associated or caused for these agricultural input was assessed with published emission factors. In general, paddy rice, wheat, corn and soybean in China had the mean CFs with2472,794,781and222kg carbon equivalent (CE) ha-1 in area, and0.37,0.14,0.12and0.1kg CE kg-1yield in grain yield, respectively. For dry crops,78%of the total CFs was contributed by N fertilizer use, with which direct soil N2O emission and indirect emissions from N fertilizer manufacture. For flooded rice paddy, direct CH4emission contributed69%to the total CFs. Moreover, the variations in CFs of dry crops among different provinces could be mostly explained by the difference in N fertilizer application rates while CH4emissions could explain85%of the variation in the CF across provinces for paddy rice. When a reduction in N fertilization by30%is considered, a potential of GHGs reduction of60Mt CO2-eq from production of these crops could be projected. This work highlighted opportunities to gain GHG mitigations in grain crops production associated with good management practices in China.(4) Carbon trading has been developed rapidly in recent years under the context of climate change mitigation, and quantifying carbon sequestration and GHGs emission reduction in the projects is the basis for carbon trading. Therefore, it is vital to develop the methodologies for quantifying the projects. Recommended fertilization use fertilizers rationally to increase production, and also reduce greenhouse gas emissions and improve soil carbon storage at the same time. The methodology on quantifying carbon sequestration and GHGs emission reduction was discussed from the aspects of the setting of the boundary and baseline, the selections of carbon pool and key GHGs emission sources and measurement methods, with the purpose of preparing for the recommended fertilization methodology in future. A methodology framework was identified to quantify the GHG emission reduction in recommended fertilization. The GHG emissions in conventional fertilization was set as a baseline and the boundary was scaled in field plot. In this framework, the key GHG sources include N fertilizer induced N2O emissions, CH4emission from rice paddy and the GHG emissions involved in fertilization equipment, and the soil carbon pool was selected to be quantified. GHG emissions during the process of fertilizer formula determination was regared as a leakage.(5) To assess the topsoil carbon sequestration potential (CSP) of China’s cropland, two different estimates were made:a) a biophysical potential (BP) using a saturation limit approach based on SOC accumulation dynamics, and a storage restoration approach from the cultivation-induced SOC loss, and b) a technically attainable potential (TAP) with a scenario estimation approach using SOC increases under best management practices (BMPs) in agriculture. Thus, the BP is projected to be the gap in recent SOC storage to either the saturation capacity or to the SOC storage of uncultivated soil, while the TAP is the overall increase over the current SOC storage that could be achieved with the extension of BMPs. The recent mean SOC density of China’s cropland was estimated to be36.44t ha-1, with a BP estimate of2.21Pg C by a saturation approach, and2.95Pg C by the storage restoration method. An overall TAP of0.62Pg C and0.98Pg C was predicted for conservation tillage plus straw return, and recommended fertilizer applications, respectively. This TAP is comparable to40%-60%of total CO2emissions from Chinese energy production in2007. Therefore, carbon sequestration in China’s cropland is recommended for enhancing China’s mitigation capacity for climate change. However, priority should be given to the vast dry cropland areas of China, as the CSP of China is based predominantly on the dry cropland.(6) The prediction of CH4emissions from rice paddies could play a key role in GHG mitigation efforts associated with agriculture. We describe a methanogenesis sub-model that has been developed in the DAYCENT ecosystem model for estimating CH4emissions and assessing mitigation potentials for rice paddies. Methanogenesis is modeled based on the simulation of soil hydrology and thermal regimes, rice plant growth, SOM decomposition, and CH4transport from the soil to atmosphere. A total of97sites from China’s rice paddies were used to develop and evaluate the model, in which25sites (91observations) were used for parameterization and72sites (204observations) were used for model evaluation. Comparison of modeled results with measurements demonstrated that CH4emissions in rice paddies of China can be successfully simulated by the model with an overall R2of0.83, and included an evaluation of CH4emissions for a range of climates and agricultural management practices. The model was most sensitive to parameters influencing the amount of labile C available for methanogenesis(7) We uses the DAYCENT ecosystem model to predict GHG mitigation potentials associated with soil management in Chinese cropland systems. DAYCENT was evaluated with data from350experiments in China’s cropland, including measurements of N2O, CH4emissions and SOC stock changes. In general, the model was reasonably accurate with R2values for model predictions versus observations ranging from0.71to0.85. Modeling efficiency varied from0.65for SOC stock changes to0.83for crop yields. Mitigation potentials were estimated on a yield basis (Mg CO2-eq. Mg-1Yield). The results demonstrate that the largest decrease in GHG emissions in rainfed systems are associated with combined effect of reducing mineral N fertilization, organic matter amendments and reduced-till coupled with straw return, estimated at0.31to0.83Mg CO2-eq. Mg-1yield. A mitigation potential of0.08to0.36Mg CO1-eq. Mg-1yield is possible by reducing N chemical fertilizer rates, along with intermittent flooding in paddy rice cropping systems.In this thesis, a database was built to develop and apply the accounting and simulating methods, and both of soil carbon sequestration and greenhouse gas emissions were considered. This thesis improved the approach for evaluating and predicting of GHG emissions reduction potential in the cropland, developed an agricultural carbon accounting methodology, and constructed an analysis framework for assessing carbon footprint of crop production. The characteristic of high carbon input and low carbon efficiency in Chinese agricultural production was proposed, and the GHG emission reduction potential and its technical approach were revealed. This thesis provide a solid scientific basis and technical support for carbon sequestration and GHG mitigation in Chinese agricultural production.更多还原