干旱区生态环境知识资源中心

The division of arid areas is important in water and land resources management, planning and for a long-term agricultural, economic and social planning. Northwest China (NW) dominates the main arid areas in China. There is thus a need to adopt adequate concepts relative to the scope of arid areas of NW China and identify its climate types and characteristics. In this study, we analyzed climatic data over the last 30 years (1981-2010) from 191 stations in three provinces and three autonomous regions of NW China. The factor-cluster analysis technique (FC), an objective and automated method was employed to classify the dry/wet climate zones. The traditional methods with predefined thresholds were adopted for providing a comparison with FC. The results showed that the wet/dry climate zones by FC were mainly distributed along mountains, rivers and desert borders. Climate-division boundaries relied heavily on the major terrain features surrounding the grouped stations. It also showed that the climate was dry in the plain sandy areas but relatively wet in the high mountain areas. FC method can reflect the climate characteristics more fully in NW China with varied and complicated topography, and outperform the traditional climate classifications. Arid areas of NW China were defined as four climate types, including five resultant classes in FC classifications. The Qinling and Da Hinggan Mountains were two important boundaries, besides main administrative boundaries. The results also indicated that there are some differences between two traditional classifications. The precipitation moved and fluctuated to an extent, which confirmed that climate change played an important role in the dry/wet climate zoning, and the boundaries of dry/wet climate zones might change and migrate with time. This paper is expected to provide a more in-depth understanding on the climate characteristics in arid areas of NW China, and then contribute to formulate reasonable water and land management planning and agricultural production programs.