通过陆面过程模型模拟青藏高原多年冻土关键水热特征

Permafrost is a result of long-term energy-water exchange of the ground and atmosphere system on the Qinghai-Tibet plateau (QTP). Its distribution and hydrological and thermal characteristics are concerned by plateau climate study, cold region engineering and water resources study of China. While the empirical modeling methods are usually absent of mechanism description and unable to simulate hydrological and thermal processes, however they have obtained extensive applications to permafrost modeling on QTP. On contrast, the physically explicit land surface models (LSMs) who are capable of depicting the energy and water budgets between atmosphere, vegetation and soil, are constrained by insufficient data and parameters available on QTP so that no application yet are made to areal simulations with LSMs. LSMs were originally developed for simulating near-surface energy and water exchange. Therefore, although some frozen soil parameterization schemes have been implemented in LSMs, they cannot meet the requirements of permafrost research that needs accurate soil temperature simulations at deep depths. In this study, the Noah LSM is improved in terms of thermal parameterization scheme and soil heterogeneity. The data stress is greatly alleviated when permafrost surveys of QTP are available, supported by a programme of the Chinese Minister of Science and Technology, with which five years investigation on five typical areas and several profiles across QTP were carried out. By extending site soil profiles to a surveyed area and then to the whole QTP, the improved Noah LSM is able to simulate the whole QTP and to produce not only permafrost spatial distribution, but some key hydrological and thermal characteristics such as mean annual ground temperature and soil ice content which are important spatial information for cold region engineering and research. The effects of environment to the development of permafrost are also discussed based on simulations. This work lays solid foundation to future study of evaluating responses of permafrost on QTP under different climatic change scenarios.

多年冻土是青藏高原地气系统水热交换的产物，其分布和水热特征对于青藏高原气候研究、寒区工程和中国水资源研究有重要意义。目前青藏高原多年冻土模拟以经验和统计模型为主，但由于缺乏对机理的描述一般不能刻画冻土水热过程。陆面过程模型有明确物理过程，能够系统描述大气-植被-土壤的水热交换，但由于受青藏高原可用数据的限制，并没有得到很好的应用。现有陆面过程模型尽管有考虑冻土参数化方案，但模拟表明不能满足多年冻土研究对深层土壤地温精确模拟的需求。本研究从热力学方案和土壤异质性着手改进Noah模型。结合科技部青藏高原冻土调查专项取得的典型区详尽调查数据，建立从点到典型区最终扩展到整个青藏高原的土壤参数面域化方案，使得改进的Noah模型可以用于模拟整个青藏高原，并最终得到青藏高原多年冻土分布以及寒区工程和研究十分关注的年平均地温、地下冰含量等关键水热特征。本研究也从水热耦合方面探讨了环境对冻土发育发展的影响。

青藏高原发育着大面积的多年冻土，在历史气候变化背景下已经发生了严重的退化。气候变化导致多年冻土水热性质也发生了变化，以往的研究只能在站点尺度研究冻土的水热过程，对于整个青藏高原冻土水热性质的变化并不清楚。本项目旨在通过改进陆面过程模型，实现对青藏高原面上的水热特征模拟，从水热耦合角度理解青藏高原环境—多年冻土相互作用。主要研究内容包括在唐古拉站点的模型适用性评价及改进、典型区（西昆仑和改则）模拟和青藏高原模拟。本项目建立了较高质量的模型驱动数据、土壤质地数据和边界条件，研究了模型参数迁移策略及多尺度验证方案，并结合青藏高原冻土环境特征改进了Noah LSM模型，较好克服了技术难点。重要结果包括：(1) 研究表明被广泛应用在青藏高原的经验/半物理模型均存在高估情况；(2) 提出新的遥感地表温度数据重建算法，较好解决模型上边界温度问题；(3) 形成了适合青藏高原冻土水热状况模拟的数值模拟方法；(4)1980-2010间青藏高原多年冻土从面积和热状况上均发生显著退化。项目产出了适合青藏高原冻土水热状况模拟的Noah-Tibet模型；产出了分18层、总深度超15m的青藏高原土壤质地类型数据集、1980-2010共31年青藏高原多年冻土分布及关键热属性分布图。项目的研究结果提高了多年冻土研究的建模技术手段，加深了我们对气候变化下青藏高原冻土响应规律的认识，为研究未来气候变化情景下多年冻土响应奠定基础。