基于空间插值数据支持下新疆伊犁地区草地类型判别与分类研究

Automatic classification of grassland type in Xinjiang Ili based on spatial interpolation of remote sensing and other data

在遥感和地理信息系统技术支持下,我国草地类型学研究有新的进展。本研究在伊犁地区草地群落高度、盖度、地上生物量、地下生物量、草地表层土壤容重、土壤全碳、土壤有机碳、土壤全氮、土壤全磷等空间分布数据的基础上,通过对不同草地类型不同指标的特征值分析,研究确定各草地类型不同指标的阀值范围,采用决策树分类法,实现新疆伊犁地区草地类型自动判别。研究结果表明,利用伊犁地区不同草地类型在群落高度、盖度、地上生物量、地下生物量、草地表层土壤容重、土壤全碳、土壤有机碳、土壤全氮、土壤全磷9个指标的特征值作为草地类型划分的依据,可以简单直观地反映各类草地的空间分布面积和分布范围,与20世纪80年代草地调查数据比较,分布趋势一致,结果可靠。同时,本研究为伊犁地区草地资源利用管理提供基础,为伊犁地区草地资源监测和信息管理平台建设提供依据,对于指导新疆伊犁地区草地畜牧业生产具有现实意义。

With support of remote sensing and geographic information system technology,the inventory of grassland types in China has advanced rapidly.This study used a range of input data including among others grassland community height,ground cover,aboveground biomass,underground biomass,the surface soil bulk density,soil total carbon content,soil organic carbon content,soil total nitrogen content,and soil total phosphorus content.With the collected data,a spatial interpolation algorithm was applied using a Decision Tree Classifier approach to produce an automated classification and index of grassland type in Xinjiang Ili.The grassland types identified based on these nine input variables,consistently matched grassland survey data collected in the 1980 s,and showed the methodology to be reliable This study provides a tool for decision makers involved in managing the utilization of grassland resources and livestock production in the Xinjiang Ili region.