基于Landsat8遥感影像青海省草地产草量监测研究  

Monitoring grassland yield in Qinghai Province based on Landsat8 remote sensing image

作　　者：李娜[1] 李婷[2] 王苑[2] 乔占明[2] LI Na LI Ting WANG Yuan QIAO Zhanming(Station of Quality Supervision and Inspection of Surveying and Mapping Products of Qinghai,Xining 810001 ,China Basic Geographic Information Center of Qinghai, Key Laboratory ofGeo - spatial Information Technology and Application of Qinghai, Xining 810001, China)

机构地区：[1]青海省测绘产品质量监督检验站,青海西宁810001 [2]青海省基础地理信息中心,青海省地理空间信息技术与应用重点实验室,青海西宁810001

出　　处：《青海大学学报（自然科学版）》2016年第5期63-68,共6页Journal of Qinghai University(Natural Science)

基　　金：青海省第一次全国地理国情普查(2014);地理空间信息工程国家测绘地理信息局重点实验室经费资助项目

摘　　要：为了更好地开展青海省复杂草地类型特点下的草地产草量遥感监测,选择2015年8月草地主生长季Landsat8遥感影像,以地理加权回归模型为基础,结合地面样点,按统一建模和分草地类型的分类建模两种方法,对青海省天然草地产草量进行估算。结果表明:(1)草地类型空间分布特征的分类建模评估精度达75%,优于统一建模的61%;(2)两种模型总产草量估算值差异较小,但不同区域两种模型估产值差异较大。利用遥感监测技术方法,将草地类型空间分布区域引入模型,能更好地反映实地草地产量,满足青海省草原生产力各类管理的需求。In order to better monitor the yield of complex grassland types of Qinghai, the remotesensing images Landsat 8 in August, 2015 were selected and combined with field sampling. Accordingto geographical weighted regression model, unified model and grass type classification modelwere constructed to estimate the yield natural grassland in Qinghai. Results show that： （ 1 ） the accuracyof the spatial distribution of grassland type classification modeling is 75% , and 61% for the unifiedmodeling; （ 2 ） The difference of estimated total forage yield of the two models is small, butwith bigger difference in regional yield estimation. By using the technology of remote sensing monitoringand introducing the spatial distribution of grassland type into the model, the actual grasslandyield can be reflected, which meets the needs of the various management of grassland productivity inQinghai province.

关 键 词：草地生产力 植被覆盖度 地理加权回归模型 青海省 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] X144[自动化与计算机技术—控制科学与工程]