中国东北半干旱区能量水分循环过程的资料同化与数值模拟研究

The Northeastern China is one of the areas sensitive to global change. Under the circumstances of climate change, and the water-energy cycle and local climate effects has significantly changed. It is a big challenge to simulate the feathers of water-energy cycle using regional meteorological model over this area. The purpose of this work is how to fully and effectively use the various kinds of conventional and non-conventional observations to generate a more accurate model initial field, and reduce the error in initial and boundary fields, and using the data assimilation method to improve simulation results. Therefore, a cycling data assimilation method will be developed based on 3D-var method and the near-surface meteorological elements will be assimilated into the WRF model, and so are the multi-source remote sensing observations data will be integrated in numerical model. The purpose of this work is to improve the simulated atmospheric conditions and land surface energy budget components with better initial land surface characteristics, in contrast to the ones in default model. The high spatial and temporal data assimilation dataset will be generated and the credibility of this datasets will also be verified. It is very important for improving a good performance in numerical simulation over semi-arid area northeastern China. The assimilation datasets generated by this work can also be applied to research on climate change and environmental monitoring.

东北地区是全球变化最为敏感的区域之一。由于气候变化的影响，该地区的能量水分循环过程和局地气候效应发生了显著变化。如何利用区域大气模式准确地模拟出该地区水热循环特征，是一个很大的挑战。本研究的目的是充分、有效地利用当前不断涌现的各种常规、非常规观测资料来形成较为准确的模式初值场，减少模式初值和边值误差，利用资料同化的方法改善模拟状况。基于此，本项目拟发展基于WRF-3Dvar的循环同化法，同化该区域地面气象观测资料，并将多源陆表遥感观测数据与数值模式相融合。给出一个全局最优的大气状态和较为准确的下垫面信息，改善对该区域能量水分循环过程的模拟，最终得到高时空分辨率资料同化数据集，并对此数据集的可信度进行检验。本研究对提高中尺度大气模式在该地区的预测能力至关重要，所得数据集也可用于气候变化、干旱监测等方面研究。

使用地球观测系统的中分辨率成像光谱仪(EOS-MODIS)提供的归一化植被指数(NDVI)产品估算植被覆盖度和航天飞机雷达地形测绘任务(SRTM)制成的数字高程模型(DEM)数据遥感产品替换WRF 模式默认的植被覆盖度和地形高度, 并且利用WRF 模式及其先进的三维变分同化系统(WRF-3DVar)循环同化东北半干旱区自动气象站近地面气象要素,对东北半干旱区的温度场、湿度场、风场和能量场的结构及其日变化特征进行了较为细致的模拟研究. 通过4 组数值模拟试验分别探讨了同化气象要素与改变模式地表参数引起的不同下垫面潜热、感热的分配关系和降水、土壤湿度变化引起的地表能量通量模拟效果. 结果表明, WRF 模式能够较好地模拟出东北半干旱区夏季的近地面温度、风向、净辐射、感热和潜热等要素的变化特征及日变化规律. 同化试验(Case 2)模拟的近地面气温、相对湿度、风速相比控制性试验有所改善; 陆面参数试验和集合试验改善了感热和地表热通量的模拟. WRF 模式能较好地模拟出下垫面土壤湿度随时间变化的规律, 集合试验土壤湿度模拟结果与4 个通量站观测值相比无太大差别, 但降水的模拟有待改善. 本研究利用卫星遥感资料改善模式下垫面陆面参数, 利用气象资料同化改善近地面大气要素模拟精度, 这是将各种不同空间和时间尺度的多源数据与数值模拟融合的有益尝试. 此研究生成的东北地区资料同化数据集可用于气候变化、干旱监测等方面, 对深入了解半干旱区气候的形成和维持机理具有重要的意义.