期刊文献+

基于神经网络的西藏拉萨地区生态环境分类 被引量:3

Ecoregion Using Artificial Neural Networks in Lhasa Area, Tibet
下载PDF
导出
摘要 根据生态环境分类指标的科学性、完备性、简洁性和数据的可获取性,选取了影响拉萨地区生态环境的主要地形和气候因子高程、坡向、≥0℃积温、年平均温度、年平均降水量、潜在蒸散量和湿润度等7个有代表性的指标,利用GIS的空间内插方法将所有这些指标转成100 m×100 m的空间珊格数据,再根据每个指标特定的地理和环境意义进行指标的分带,对7个指标进行主成分分析后提取主要信息。通过选择4个典型样区作为训练区,对拉萨地区的生态环境进行了分类。结果表明,拉萨地区的主要生态环境类型包括河谷农业类型、山地草原类型、高山草甸类型及高山裸岩及冰雪类型。其中,高山草甸和山地草原生态环境类型占主导,分别为10 768.52 km2和10 646.6 km2,各占总面积的36.61%和36.20%,而河谷农业类型占总面积的10.75%。此外,拉萨地区分布有较大面积的高山裸岩及冰雪区生态环境类型,面积为总面积的14.16%。作为特殊类型的生态环境类型,拉萨地区境内的纳木错的湖泊面积是668.76 km2,占该湖面积的近一半和拉萨地区总面积的2.27%。 Based on the topographical and climatic conditions in Lhasa area located at the central Tibetan Plateau, elevation, aspect, ≥0℃ accumulated temperature, mean yearly temperature, yearly precipitation amount, potential evapotranspiration and humidity is selected as the main ecoregion indicators. All of indicators are interpolated into spatial data with 100 × 100 m resolution. Considering 7 indicators, the PCA (Principal Component Analysis) is used for deriving the main information, which is input data of ANN (Artificial Neural Networks). Then, the ecoregion for Lhasa area studied using ANN. The result shows that there are five main ecoregion types in Lhasa area, namely valley agriculture zone, mountain steppe, alpine meadow, snow and ice and rock, and lake. Among them, alpine meadow and mountain steppe are main ecoregion types, which are 10768.52 km^2 and 10646.60 km^2, respectively. The area of these two types is 70% of total Lhasa area, and valley agricultural zone is 10.75 % of total area. There is a large area of snow and ice and rock type, which is 14.16 % of total area. The Namsuo Lake is a special ecoregion type in Lhasa area. Nearly half area of Nansuo Lake is located in the Lhasa area and its area is 2.27 % of total study area. The main environmental questions in different ecoregional type are as follows: 1. In valley agriculture zone Although there is certain improvement in the environment of this zone through implementing the integrated agriculture development, particularly through the ways of afforestation and building shelter belt for crop land, there is few organic matters returning to crop land, so soil organic matter content can not increase and soil structure can not get improving because of energy shortage in this area. Therefore, there are difficulties to improve environment further. Soil and water losses in some mountain-valley area is relatively strong also because of human activity such as over-cutting shrub. 2. In mountain steppe zone As a result of relatively big slope in this ecoregional type the water storage capacity is lower and the environment is comparatively dry and soil lay is shallower. In addition to human activity the water and soil losses in this area is stronger. This type of ecoregion is the most vulnerable in this study area. 3. In alpine meadow zone Because of large temperature difference in this area the freeze-thaw action is comparatively obvious. AIthough vegetation such as the meadow grass grows good, the grass degeneration caused by overgrazing exists in some part of this area.4. In snow, ice and rock type Because of low temperature and shallow soil lay in this area, the soil erosion caused by ice and snow is strong. However, at present this area basically belongs to natural ecosystem and is difficult to use for human,but once destroyed it is very hard to restore.
作者 除多
出处 《山地学报》 CSCD 北大核心 2005年第4期391-398,共8页 Mountain Research
基金 国家自然科学基金资助项目(40361001) 中国气象局配套资金资助。~~
关键词 生态环境分类 神经网络 西藏拉萨地区 ecoregion, ANN, Lhasa area, Tibetan Plateau
  • 相关文献

参考文献6

  • 1刘南威 郭有立.综合自然地理[M].北京:科学出版社,1994.2.102~103.
  • 2中国科学院青藏高原综合科学考察队.西藏自然地理[M].北京:科学出版社,1982.17-159.
  • 3西藏自治区拉萨市农牧局.西藏拉萨土地资源[M].北京:中国农业科技出版社,1991..
  • 4张增祥 除多.西藏自治区"一江两河"地区环境综合评价与动态监测研究[M].北京:宇航出版社,1998..
  • 5西藏自治区一江两河开发建设委员会办公室.西藏自治区"一江两河"地区综合开发生态环境规划[R].,1997.4..
  • 6J. A. Richards, et al. Remote Sensing Digital Image Analysis[M]. Berlin: Springer-Verlag, 1994. 340~341

共引文献22

同被引文献52

引证文献3

二级引证文献27

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部