基于遥感信息与DNDC模型同化的稻田温室气体排放量估算

Human-induced greenhouse gas (GHG) emissions are closely related to global climate changes. Methane emissions from paddy rice fields could account for approximately 10% of anthropogenic methane emissions. Since China is one of the major rice-producing countries, accurate estimates of methane emissions from paddy rice fields can provide scientific basis for making national policies. By describing the detailed role of soil microbes in the ecosystem, DNDC biogeochemical model has the ability to calculate agricultural greenhouse gas emissions accurately. However, due to difficulties in obtaining land surface parameters, it is difficult to derive the high-resolution spatial distributions of agricultural methane eimissions. This research proposes to study the rice fields in Jiangsu Province by assimilating remote sensing information and DNDC model. We would build an assimilation model RS-DNDC by applying cost functions, and perform model validation with multi-souce data. By developing models from site-scale to spatially-distributed simulations, the proposed assimilation model is able to estimate total methane eimssions from paddy rice fields, while providing their high-resolution spatial information.

人为温室气体排放与全球气候变化有紧密联系，而稻田排放的甲烷占人为甲烷排放总量的约10％。我国是水稻大国，准确估算稻田甲烷温室气体排放量，能为国家制定粮食政策和国家控制农业温室气体排放提供科学依据。生物地球化学模型DNDC，深入刻画土壤微生物在生态系统中的作用，合理计算农田温室气体排放量，却由于陆面参数获取等困难，难以获得高分辨率的空间分布信息。本研究选择江苏省稻田为具体研究对象，结合空间遥感RS信息与DNDC模型，利用代价函数构建同化模型RS-DNDC，结合多源数据进行模型验证和不确定性分析，实现模型从点位模拟到分布式计算的提升，在估算稻田甲烷温室气体排放总量的同时，获取其高分辨率的空间分布信息。

项目将遥感信息与生态系统过程模型进行同化，深入刻画了水分、光照、温度等环境因子协同作用下的植物生长机理，发展建立了植被生理生态机理模型GPD、植被冠层模型、植被物候模型、融合遥感信息的DNDC模型等。以理论模型为基础，进行了水氮管理对稻田生态系统服务与功能的影响、农田光能利用率多尺度评估、植被物候模拟、太阳直射与散射辐射对遥感反演作物总初级生产力影响、植被覆盖连续监测方法等方面的研究工作。研究提高了生态系统过程模拟的准确性，为解决陆面过程模型参数率定等困难提供了参考方案。..项目直接成果体现在利用同化遥感信息的DNDC模型估算了全国稻田温室气体排放量，并获取了其高分辨率空间分布图。在项目的支持下，研究共发表SCI期刊检索论文10篇，项目负责人晋升副教授职称，参加国际学术会议3次、国内学术会议1次，获各类奖项3项，圆满取得了项目的预期成果要求，为后续研究打下了良好基础。