南方红壤典型水土流失区植被分类及植被类型变化的遥感评估——以福建省长汀县河田地区为例

Vegetation Classification and Variation Assessment in a Typical Red Soil Erosion Area in Southern China:Hetian,Changting County of Fujian Province

以福建省长汀县河田地区的植被为研究对象,选取2010年Landsat 5和2014年Landsat 8遥感影像,基于影像的植被光谱特征曲线和纹理波段,对该区进行植被分类和植被类型变化的研究。精度验证表明,两年份的植被分类的总精度分别为85.62%和86.44%,Kappa系数分别为0.804和0.818。研究结果表明:2010~2014年间河田地区共增加植被面积590.40hm^2,并且两年份的植被类型均以马尾松为主,其面积占各自年份植被总面积的69%以上,而其它植被类型的面积比例各异。分析发现,政策的支持是该区植被在这4年间发生变化的主要驱动力。研究表明该方法能够有效地对植被进行分类,并且可以准确地掌握河田地区各植被的分布及面积,较好地了解各植被的变化及变化原因,为水土流失治理及植被优化提供科学决策依据。

This paper focused on the vegetation classification and the evaluation of vegetation change on the basis of the spectral characteristic and texture of various vegetation types in remote sensing imagery.Two Landsat images of Hetian district in Changting County,dated on 29 October 2010and 7October 2014,were employed to do the classification and evaluation.The validation showed that the total accuracy of the vegetation classification was 85.62%and 86.44%,with corresponding Kappa coefficients of 0.804 and 0.818,respectively,for the two study years.The results showed that the area of vegetation in Hetian district had increased by 590.40hm^2 from 2010 to 2014.The vegetation in the two study years was dominated by Masson Pine,which accounted for more than 69% of the total vegetation area.It is suggested that the government policy is the main driving forces to the vegetation change during the four year period.The result indicated that the remote sensing method used in this study could effectively classify the vegetation,and accurately reveal the variation of the vegetation in Hetian area in the study years.This helps a better understanding of the vegetation change and causes for the change in the area.The results will provide a scientific support for the decision makers for the further control of soil and water loss with optimization of vegetation types.