img

官方微信

高级检索

中国沙漠 ›› 2014, Vol. 34 ›› Issue (6): 1562-1567.DOI: 10.7522/j.issn.1000-694X.2013.00427

• 生物与土壤 • 上一篇    下一篇

克里雅河流域土壤盐分光谱定量分析

陶兰花, 塔西甫拉提·特依拜, 姜红涛, 买买提·沙吾提, 吴雪梅   

  1. 新疆大学 资源与环境科学学院/绿洲生态教育部重点实验室, 新疆 乌鲁木齐 830046
  • 收稿日期:2013-09-03 修回日期:2013-10-28 出版日期:2014-11-20 发布日期:2014-11-20
  • 作者简介:陶兰花(1988- ),女(蒙古族),新疆博乐人,硕士研究生,主要从事干旱区资源遥感定量研究.Email:tlh1003@163.com
  • 基金资助:
    国家自然科学基金重点基金联合项目(U1138303);教育部长江学者和创新团队(IRT1180)资助

Quantitative Retrieval of Soil Salt Content Using Hyperspectral Data in the Keriya River Basin

Tao Lanhua, Tashpolat·Tiyip, Jiang Hongtao, Mamat·Sawut, Wu Xuemei   

  1. College of Resources and Environment Sciences/Key Laboratory of Oasis Ecology under Ministry of Education, Xinjiang University, Urumqi 830046, China
  • Received:2013-09-03 Revised:2013-10-28 Online:2014-11-20 Published:2014-11-20
  • Contact: 塔西甫拉提·特依拜(Email:tash@xju.edu.cn)

摘要: 本文对克里雅河流域进行野外调查、采集土壤样品及其光谱反射特性的测量,通过比较不同光谱预处理的方法建立偏最小二乘回归(PLSR)模型,并利用决定系数(R2)、均方根误差(RMSEP)、残留预测偏差(RPD)对模型的稳定性和预测能力进行检验.结果表明:反射率一阶微分是预测土壤样本盐分含量的最佳光谱指标.PLSR模型在建立土壤光谱与盐分含量关系时较为适用,R2RMSERPD分别为0.77、0.25和1.88.利用反射光谱估算土壤中盐分含量,通过各种光谱预处理方法可以提高估算精度,可以为该区土壤盐渍化评价和生态环境调查提供依据.

关键词: 高光谱, 土壤, 盐分含量, 偏最小二乘回归法

Abstract: Spectral reflectance properties of soil samples, collected in the area along the Keriya River, were measured. A partial least squares regression (PLSR) model was establish by comparing different preprocessing methods spectral reflectance and a theoretical foundation for the quantitative inversion of soil salt content was provided, then the root mean squared error (RMSE) was introduced to test the predictability and precision of the model, coefficient of the determination (R2) was used to evaluate stability of the model. The results demonstrated that: the first derivate reflectance was optimal index for predicting salt content; the PLSR was the optimal model to establish the relationship between the soil spectrum and salt content, in which the R2, RMSE and rate of prediction to deviation was 0.77, 0.25 and 1.88, respectively.

Key words: hyperspectral, soil, salt content, PLSR

中图分类号: