摘要
在棉蚜危害棉花的主要生育期测试蚜害后不同严重度叶片的理化参数和光谱,分析并比较了不同时期、不同品种蚜害棉叶理化参数和光谱反射率特征,确定了蚜害棉叶光谱敏感波段,建立了蚜害棉叶严重度估测模型。结果表明:蚜害棉叶的理化参数和光谱特征明显,蚜害棉叶平均厚度,含水量,温度和叶片CHL.b均增加,叶片而平均CHL.a,CHL.a+b和Cars含量减少。不同时期、不同品种蚜害棉叶光谱反射率在可见光均表现先升后降的特征,但不同品种蚜害棉叶在近红外波段反射率不同。434～727和648nm可分别作为蚜害棉叶的光谱敏感波段和最佳波段。基于敏感波段建立的蚜害棉叶遥感估测模型均达到显著相关,其中以(R1589-R648)/(R1589+R648)为自变量建立的估算模型精度最高,相对误差最小(0.128),可作为蚜害棉叶严重度最佳识别模型。该研究为遥感大面积监测棉花蚜虫提供了理论依据及参考。
The spectrum and physical-chemical parameters were measured on cotton leaves infected by aphid with different severity levels（SL）at main cotton growth periods.Meanwhile,the reflectance and physical-chemical parameters of cotton leaves infected by aphid were analyzed and compared in different cotton growth periods and varieties.The sensitivity wave bands of cotton leaves infected by aphid were confirmed,and the estimating models of leaves infected by aphid were established.The results showed that there was a significant difference in the spectrum and physical-chemical parameters of cotton leaves infected by aphid.The thickness,water and Chl.b increased,while Chla,Chl.a＋b and Cars content decreased in leaves infected by aphid.Besides,in visible region,the reflectance of cotton leaves infected by aphid has shown going up first and then down in different cotton growth periods and varieties with increasing of SL.However,in NIR region,it has shown discrepancy in varieties.The 434-727 and 648nm can be used as sensitive and optimal aphid-band for cotton leaves.Estimation models for leaves infected by aphid were all in significant correlation.Among all the models,the model of（R1 589-R648）/（R1 589＋R648） had the best estimation precision,RE was the smallest（0.128）,and it was commended as best models to estimate SL of leaves infected by aphid.The study provides an experimental reference for monitoring spectrum of cotton infected by aphid with remote sensing in large areas.
出处
《光谱学与光谱分析》
SCIE
EI
CAS
CSCD
北大核心
2010年第11期3093-3097,共5页
Spectroscopy and Spectral Analysis
基金
国家(863计划)项目(2006AA10Z207,2006AA10A302)
国家科技支撑项目(2007BAH12B02)
兵团绿洲生态农业重点实验室开放课题项目(200804)资助
关键词
棉花
蚜虫
光谱特征
识别模型
Cotton
Aphid
Spectrum characteristics
Recognition models
