摘要
为改进营造林成效人工调查方法和现有遥感监测方法的局限,研究不同分辨率、不同地域条件、不同造林年限的营造林成效遥感监测效果。选取南、北方各1个具代表性的行政县为研究区,以生长1~5 a的营造林小班为对象,采用涵盖中、高分辨率的5种遥感影像,从不同分辨率的适用性、不同造林年限的保存状况监测、NDVI成林状况监测、造林前地类监督分类4个方面对营造林成效进行监测,并通过地面调查验证监测精度。结果表明:1)宜采用与营造林小班面积相适应的分辨率进行遥感监测,建议营造林小班平均面积<2 hm 2的地区选取2 m或更高分辨率。2)相较于北方,南方营造林保存状况随造林年限增加上升较快,遥感监测保存率易偏高。3)NDVI成林状况遥感监测总精度达87.5%,北方精度高于南方3.2%,原因是南方易受到杂草杂灌的干扰。4)造林前地类监督分类总精度达72.7%,北方高于南方14.3%,原因是北方地类完整单一,易于识别。5)营造林成效采用遥感监测,成本仅为人工调查的40.2%。营造林成效遥感监测在监测效果和精度上能够满足调查需要,监测成本也较人工调查有很大优势,可作为营造林成效监测的有效手段。
In order to improve the limitations of artificial investigation method,to solve the existing problems in remote sensing monitoring method to investigate the afforestation effectiveness,it was proposed to study the effectiveness of afforestation with remote sensing monitoring method from the view of different resolutions,different geographical conditions,and different afforestation years.Two representative administrative counties that were selected both from southern and northern China were taken as the study areas.One to 5 years old subcompartments of afforested plantations were taken as the research objects.Remote sensing images with medium to high resolutions were extracted.The effectiveness of the afforestation was examined from aspects:the applicability of different resolutions,monitoring results to preservation conditions for different afforestation years,monitoring of forest establishment with normalized difference vegetative index(NDVI),and supervision of the land classification before afforestation.Field investigation was conducted to verify the monitoring accuracy.1)Remote sensing monitoring should be carried out according to the resolution of afforestation subcompartment.It was suggested that the subcompartment with an average area of less than 2 hectares should select 2(or higher)meters resolution.2)Compared with the north,the preservation status of afforestation in the south increased rapidly with the increase of afforestation year,and the preservation rate of remote sensing monitoring tended to be high.3)The total precision of remote sensing monitoring of NDVI forest condition was 87.5%,the precision in the north was 3.2 percentage points higher the south due to the disturbance of weeds and shrubs in the south.4)The total accuracy of supervision classification of forest land reached 72.7%,14.3 percentage points higher in north than in south.The reason was that the land class in the north was complete and single,and to be identified easily.5)The cost of remote sensing monitoring of afforestation effec
作者
刘晓双
李才文
龚直文
LIU Xiao-shuang;LI Cai-wen;GONG Zhi-wen(Northwest Inventory,Planning and Designing Institute of National Forestry and Grassland Administration,Xi'an 710048,Shaanxi,China;College of Economics and Management,Northwest A&F University,Yangling 712100,Shaanxi,China)
出处
《西北林学院学报》
CSCD
北大核心
2021年第3期203-209,共7页
Journal of Northwest Forestry University
基金
国家自然科学基金(31300538,31400540)
林业公益性行业科研专项(201504424)。
关键词
营造林成效
NDVI
监督分类
afforestation effectiveness
NDVI
monitoring classification
