【作者】 卢燕宇；

【导师】 黄耀；

【作者基本信息】 南京农业大学 ， 土壤学， 2007， 博士

【摘要】 温室气体引起的全球气候变暖和平流层臭氧的损耗已成为全世界面临的两大环境问题．氧化亚氮（N₂O）是大气中仅次于CO₂和CH₄的第三大痕量温室气体，农田是大气申N₂O的重要来源，然而对于农田N₂O排放的精确估计一直存在着较大困难，其原因主要在于气候、土壤以及农业管理方式等影响农田N₂O排放的因素在时间和空间上都存在巨大的变异性。中国作为农业大国，其农田对全球N₂O排放的影响已经成为全球变化研究的焦点之一，并且中国作为《联合国气候变化框架公约》的缔约国之一，编制N₂O排放的国家清单已成为刻不容缓的工作。准确估计我国的农田N₂O排放量，研究其时空变化规律，不仅对于我国的科学研究和国家利益具有一定的重要性，同时也是全球变化研究的内容之一，有助于揭示全球气候变化的规律，为减缓和应对气候变化提供有效对策．本文的研究目的为：结合我国农业生态系统的特点，在IPCC推荐的排放系数法的基础上，建立适用于我国农田的N₂O排放估算模型；在GIS技术支持下，估算中国农田N₂O直接排放量，并研究中国农田N₂O排放的时空变异规律；结合气候模式，分析未来气候变化情景下中国农田N₂O直接排放的变化趋势；探讨在区域尺度上减缓中国农田N₂O排放的对策。本研究通过充分调研现有文献资料，结合本课题组观测结果的基础上，采取一定的规則对所获取的数据进行取舍和整理后，建立了包括N₂O年排放量、施氮量、环境要素、地理位置、作物类型等要素的数据库。统计分析结果表明，在大尺度样本空间內农田N₂O排放与降水及氮肥施用量呈显著正相关，与气温、土壤pH值、有机碳及N含量无显著相关．在IPCC推荐的排放系数法的基础上，通过考虑农田N₂O排放与年降水量间的关系，本文建立了针对旱地农田N20排放的估算方程A＼O-N=（1．49±0．73）P+（0.0186±0.0027）P·F（r²=0.61，P＜0.0001）。拟合度分析表明，相较于IPCC缺省方法，引入了降水因子的估算方程能够提高对农田N₂O排放的解释性，对N₂O排放具有更好的模拟效果，也更适用于中国农田的N₂O排放估算。根据中国农业生产和种植制度的特点，本文引入水田N₂O排放系数，结合降水修正的排放系数，在GIS技术的支持下，估算了中国农田N₂O直接排放量．结果表明，在1980—2000年间中国农田化学氮源N₂O平均年排放量为167．22 GgN₂O-N yr^-1，约占氮肥用量的1．02％，其中旱作农田是主要的N₂O释放源，约占中国农田N₂O直接排放量的86％。中国农田N₂O排放具有较强的时空变异性．在分布格局上，N₂O排放通量和排放系数均表现出明显的地区间差异，东部较高，西北偏低．在时间序列上，中国农田N₂O排放量在上世纪后二十年升高了约149 Gg N₂O-N，增加了1．55倍，同时受氮肥用量和气候条件等因素的影响，不同地区N₂O排放量增加的幅度也具有较大差异。由于考虑了N₂O排放的主控因子，本文采用的方法能够更为合理的估计出农田N₂O排放的年际问变异。本研究估算的不确定性主要来自于两个方面，输入数据空间化过程中的偏差和估算方程的不确定性。结合本研究的特点，采用了IPCC推荐的误差传播公式来量化和合并估算过程中的不确定性。分析表明，本研究对中国农田N₂O年排放量估算的不确定性约为21．3％。利用气候模式驱动农田N₂O估算模型，本文分析了未来气候变化情景下中国农田N₂O排放的变化趋势。结果表明，农田N₂O排放对气候变化具有较强的敏感性。虽然在较长的时间尺度上，中国农田N₂O排放在趋势上并未发生明显的改变，然而，N₂O排放的年际变异进一步变大，波动性增强，同时其空间分布格局也发生了一定的变化，N₂O的排放中心有向北移动的趋势。基于中国农田N₂O排放和氮肥用量的地区间差异，本研究设置了中国氮肥区域分配情景。情景分析结果表明，在化学氮肥施用总量接近的条件下，通过调整中国氮肥的地区间分配，可以削减约11．7％的农田N₂O排放量。与此同时，经过氮肥用量的区域性调整后，由于提高了氮肥的利用率，中国的粮食产量具有一定的增加潜力。本研究通过大样本空间数据的分析，探讨了农田N₂O排放与各影响因素间的关系，基于这种关系所建立的N₂O排放估算模型，在保持排放系数法操作简单的基础上提高了估算的准确性。在GIS技术的架构下，本文将栅格化数据集和估算模型相结合，研究了较长时间序列内中国农田化学氮源N₂O直接排放的时空变异规律，系统的回答了中国农田N₂O排放的过去、现在和将来，这将有助于我们评价中国农田对大气温室气体浓度升高的贡献，为制定有关温室气体的的决策提供科学支持，对合理预判和应对未来全球变化的趋向具有重要意义。更多还原

【Abstract】 Nitrous oxide （N₂O）, an important atmospheric trace gas, has been paid muchattention due to its substantial contribution to global warming and ozone depletion, thetwo severest impacts on the global environment. Although the N₂O budget remainspoorly understood at present, fertilized agricultural soils have been believed to be amajor source of annual global N₂O emission. Owing to the large spatial and temporalvariabilities in the agricultural practices and environmental factors, it is difficult toprecisely estimate agricultural N₂O emission at a large regional scale.China is one of the largest agricultural countries in the world. Huge amount ofsynthetic fertilizer have been applied to maintain high crop yield to feed the increasingpopulation. The impact of Chinese agricultural activities on atmosphere N₂O incursincreasing attentions. Moreover, according to the UNFCCC, China is obliged tocompile greenhouse gas inventories in order to make proper policy to reduce theiremissions. The estimation of N₂O emission from agricultural fields in China is alreadymore than a scientific topic.The objectives of this study are, 1) to establish empirical models in which theemission factor and background emission for N₂O in the IPCC methodology wererectified by key environmental parameters; 2) to simulate long term spatio-temporalcharacteristics of direct N₂O emission in China’s agriculture by linking the statisticalmodels to spatial datasets via GIS; 3) to predict the response of agricultural N₂Oemission to climate change through adopting the output of climate model to drive N₂Oestimate models; 4) to examine potential mitigation option of agricultural N₂Oemission on regional scale.In this study, we collected direct N₂O measurements that were published in peer-reviewed English and Chinese journals between 1982 and 2003. After dataextraction following strict criteria, the employed data set remained of 206measurements from 42 sites. For each data point, we recorded the annual total N₂Oemission, the synthetic nitrogen application rate, the precipitation and meantemperature, the location of the measurement and the soil physiochemical parametersincluding soil total N content, soil organic carbon content, and pH. Both nitrogen input（F） and precipitation （P） were found to be largely responsible for temporal and spatialvariabilities in annual N₂O fluxes, while no significant correlations were observedbetween N₂O emission and other parameters of temperature, pH, organic carbon andnitrogen.Based on the definition of N₂O emission factor and its quantification by theIntergovernmental Panel on Climate Change, we established an empirical model forupland （N₂O-N=1.49 P+0.0186 P·F）, in which both emission factor and backgroundemission for N₂O were rectified by precipitation. Fitness analysis suggests that theinclusion of precipitation into regression eqution could improve the predictability, andis more suitable to quantify N₂O emissions from agricultural fertilized fields in ChinaConsidering the large paddy rice planting area in China, the cropping-specificemission factors were introduced to estimate fertilizer-induced N₂O emissions fromrice paddies, while the precipitation-rectified emission factor to the uplands. Annualfertilizer-induced N₂O emissions from China’s agricultural fields were simulated bylinking these models to spatial datasets via GIS. During the period 1980 to 2000, themean annual emission was estimated to be 167.22 Gg N₂O-N yr^-1, equivalent to 1.02%of the national total of synthetic fertilizer N input. Upland was the major source ofannual N₂O emission from China’s agriculture, contributing about 86% to the total offertilizer-induced N₂O emissions from agricultural fields in China.Direct N₂O emission from China’s agricultural fields occurs essentially with greatspatial and temporal variability. This study showed a significant spatial distributionpattern of direct N₂O emission in agricultural fields. The maximum direct N₂Oemission occurred in the major district of crop production, where is characterized ashumid climate in China. By contrast, N₂O emission rarely exceeded 0.10 kg N₂O-Nha^-1.yr^-1 in the West region, where is generally dominated by dry climate. During thelast two decades in the 20th century, the amount of direct N₂O emission from Chineseagricultural fields has increased by about 149 Gg N₂O-N yr^-1, equivalent to 155% of the initial. The increasing range of N₂O emission varied widely with different regions,which is mainly attributed to the quantity of synthetic nitrogen application and climateconditions. In contrast with the IPCC default methodology, this study may providemore insights into the inter-annul variations in N₂O emissions.Uncertainty estimate is an essential element of a complete inventory of N₂Oemission. The uncertainties of this study were principally originated from the modelsand spatializing process. Similar to the uncertainty estimate in the IPCC methodology,the error propagation equation was adopted to quantify the uncertainties of this study.Results indicated that the combined uncertainty in this study was about 23%.In this paper, we simulated direct N₂O emission from Chinese agriculture from2005 to 2050 by linking N₂O estimate models with FGOALS’s climate changescenario. Results suggest that N₂O emission from agricultural fields is sensitive toclimate change. Although the trend of direct N₂O emission from Chinese agriculturemay be constant essentially in the future, the inter-annual variations of N₂O emissionwill be enhanced. Moreover, the changing spatial distribution pattern of N₂O emissionwas predicted, and the high emission centers will move northward.According to the regional variations of direct N₂O emission and synthetic nitrogenapplication, we simulated a fertilizer application scenario. Scenario analysis indicatedthat the adjustment to the spatial distribution of fertilizer application in China might bea potential mitigation option of agricultural N₂O emission with the premise of insuringcrop production.更多还原