摘要
利用1988—2013年Landsat TM/ETM+/OLI时间序列分类数据,根据采伐迹地的时序变化特征,提取长汀县1988—2013年森林采伐信息,并对监测结果进行精度评价。结果表明:长汀县各时相森林采伐监测的总体精度达90%以上,除1994-11-02、2001-12-31外,其余时相的kappa系数均接近或者超过0.8,利用采伐迹地特有的地表覆盖变化特征,能将光谱特征极为相近的采伐迹地、裸地、休耕期的耕地做进一步区分;长汀县采伐迹地分布最密集的区域集中在西南部,其次是东南部、东北部;长汀县1988—2001年森林采伐的面积为500~1 000 hm2,2003年以后随着森林覆盖面积持续增加,森林采伐面积增加,其中2010年森林采伐的面积超过2 800hm2,中心区域采伐迹地面积增加与该区域森林覆盖条件显著改善有直接关系。
We employed time-series Landsat TM/ETM ＋/OLI classification data during 1988-2013 in Changting County, and the time-series variation features of the forest harvest land to extract forest harvest information and evaluate the accuracy of monitoring results. The results showed that the overall accuracy was above 90% for each temporal classification image, and the kappa coefficients were close to or above 0.8 for most of the time-series images, except for the datas on November 2, 1994 and December 31, 2001. Although the harvested land, bare land and fallowed farm land have similar spectral features, we can further distinguish them according to the special surface coverage variation characteristics of forest harvested land. The most densely harvested land distributed in the southwest of Changting County, next were the southeast and the northeast. The forest harvest acreage reached about 500-1 000 hectares during 1988-2001, it presented an increasing trend after 2003 along with the increasing of forest acreage, and the forest harvesting area was more than 2800 hectares in 2010. Obviously, there is a direct relationship between the increasing acreage of the harvested area and the significant improvement of forest coverage in the central area.
作者
谢旭
陈芸芝
汪小钦
陈雪娇
Xie Xu Chen Yunzhi Wang Xiaoqin Chen Xuejiao(National Engineering Research Center of Geospatial Information Technology, Spatial Infolnaation Research Center of Fujian Province, Fuzhou University, Fuzhou Fujiang 350002)
出处
《西南林业大学学报（自然科学）》
CAS
北大核心
2017年第4期167-174,共8页
Journal of Southwest Forestry University：Natural Sciences
基金
国家科技支撑计划课题(2013BAC08B01)资助
国家自然科学基金项目(41401488
41571330)资助
关键词
LANDSAT
时序
分类
森林
采伐
Landsat, time-series, classification, forest, harvest
