土壤湿润条件下基于光谱对称度的盐渍土盐分含量预测  

Predicting Soil Salinity Based on Spectral Symmetry under Wet Soil Condition

作　　者：刘娅[1,2] 潘贤章[1] 王昌昆[1,2] 李燕丽[1,2] 石荣杰[1,2] 周睿[3] 解宪丽[1] 

机构地区：[1]中国科学院土壤环境与污染修复重点实验室(南京土壤研究所),江苏南京210008 [2]中国科学院大学,北京100049 [3]中国科学院南京分院,江苏南京210008

出　　处：《光谱学与光谱分析》2013年第10期2771-2776,共6页Spectroscopy and Spectral Analysis

基　　金：国家自然科学基金项目(41071140);中国科学院战略先导性项目(XDA0505050902)资助

摘　　要：近年来光谱技术以其经济、高效的优势在土壤盐渍化监测研究中得到重视,但是由于土壤水分对反射光谱影响很大,土壤湿润条件下监测精度难以满足农业生产需求。通过对盐土土柱室内模拟蒸发过程中的反射光谱和水分、盐分变化的连续监测,利用多元逐步回归方法,建立了1 370～1 610nm光谱对称度与土壤表层含盐量、含水量之间的线性关系模型,r为0.863;用该模型反演表层土壤含盐量,实测值与预测值之间线性关系的r为0.656(n=54),RMSE为2.059g·kg-1。利用光谱对称度可以实现土壤湿润条件下土壤盐分含量预测。There has been a growing interest in using spectral reflectance as a rapid and inexpensive tool for soil salinity monito- ring in recent years. However, since soil moisture often exerts a tremendous influence on soil reflectance, the monitoring accura- cy under various moisture conditions cannot fully satisfy the requirements of agricultural practice. In the present paper, a linear model was built to relate the spectral symmetry in the band of 1 370-1 610 nm with the salt content and moisture content of the saline soil based on regularly measured data of reflectance, soil moisture and salt content of the surface of 5 soil columns during the simulated evaporation process in laboratory. The results showed that the model was good with r greater than 0. 8. By inver- sing the model, soil salt content then was predicted after moisture content was determined. The results showed that the predic- tion accuracy was acceptable with a root mean square error （RMSE） of 2.059 g . kg-1 and an r of 0. 656. The results demon- strated the feasibility of using spectral symmetry to predict soil salt content under various moisture conditions.

关 键 词：土壤盐渍化 土壤湿润条件 光谱对称度 土壤含盐量 预测 

分 类 号：S127[农业科学—农业基础科学] TP79[自动化与计算机技术—检测技术与自动化装置]