草地植被盖度遥感抽样调查方法研究  

Research on Remote Sensing Sampling Survey Method of Grassland Vegetation Coverage

作　　者：丁成翔[1] Ding Chengxiang(Academy of Animal Husbandry and Veterinary Science,Qinghai University,Xining 810016,China)

机构地区：[1]青海大学畜牧兽医科学院,西宁810016

出　　处：《青海科技》2022年第5期154-160,共7页Qinghai Science and Technology

基　　金：青海省“昆仑英才·高端创新创业人才”项目;青海省畜牧兽医科学院基本科研业务费自主课题项目(MKY-2019-03)。

摘　　要：为快速掌握区域草地资源变化情况,及时采集草地植被盖度数据,本文以青海省海南州为例,采用Sentinel-2遥感数据结合地面调查,反演出该区域草地植被盖度近似真值,以5000 m×5000 m网格抽样框提取出MOD13Q1植被盖度作为抽样底图,选取5项评价指标,评价了基于自然间断点分级算法的分层抽样方法。结果表明:基于自然间断点分级法的分层抽样满足方法要求,其样本容量、抽样误差、回归模型拟合的决定系数、均方根误差、平均相对误差分别为33、0.8%、0.966、3.32%、4.06%,经验证可备选作为草地植被盖度调查快速监测方法。In order to quickly grasp the changes of regional grassland resources,collect grassland vegetation coverage data in time,and try to establish a rapid grassland monitoring method combining remote sensing sampling and ground investigation in the target area,this paper took Hainan prefecture in Qinghai Province as an example,used sentinel-2 remote sensing datas and ground investigation to reverse the approximate true value of grassland vegetation coverage in this area,and extracted mod13q1 vegetation coverage as the sampling base map with a 5000 m×5000 m grid sampling frames.The sample size,sampling error,determination coefficient of regression model fitting,root mean square error and average relative error were selected as evaluation indicators to evaluate the sample size and survey accuracy of the stratified sampling method based on the natural discontinuity classification algorithm.The results showed that the stratified sampling based on the natural discontinuity classification method had the best overall effect,and its sample size,sampling error,determination coefficient of regression model fitting,root mean square error and average relative error were 33,0.8%,0.966,3.32%and 4.06%respectively.It had been verified that it would be used as a rapid monitoring method for grassland vegetation coverage survey.

关 键 词：草地植被 盖度 遥感 反演 分层抽样 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]