【摘要】 土壤碱化对干旱区农牧业发展危害严重,当前对碱化程度的评估主要为实验室理化性质测定及实地土壤光谱测定。土壤pH值是土壤碱化程度预测的重要参数,为讨论高分辨率卫星数据快速、准确获取土壤碱化程度的可行性,以准噶尔盆地东南缘新疆奇台碱化土壤为研究对象,通过对采样点Quickbird数据的反射率及其土壤pH数据的相关分析,探究其对土壤碱化程度预测的"敏感"波段,并建立影像反射率与土壤实测pH值之间的多种回归预测模型,选择模型判定系数、均方根误差等进行模型精度评价,得到土壤碱化程度预测的最优模型。结果表明:Quickbird影像各波段反射率均能较好反映研究区土壤的碱化程度。630~690 nm范围的band3是预测碱化土壤pH预测的最"敏感"波段。band1、band2和band3所建立的碱化土壤pH值定量模型精度最高,其预测精度达到82.41%,明显高于LANDSAT数据所建立的模型。更多还原

【Abstract】 The study aimed at searching for the most sensitive bands of Quickbird to estimate soil pH values. Soil salinization or alkalization is an important problem of land degradation and environmental degeneration in arid regions. Laboratory determination and research on spectral data are the main methods to estimate the degree of soil salinization or alkalization. In this study,Qitai County,a typical region with soil salinization or alkalization along the southeastern marginal zone of the Junggar Basin in Xinjiang Uygur Autonomous Region,was selected as the study area to explore the feasibility of using the Quickbird image in studying soil salinization or alkalization.Through the correlation analysis between the values of reflectance from Quickbird and the soil pH values at all the sample sites,it was found that the "sensitive"bands could be used to predict soil salinization or alkalization and to build the different linear regression forecasting models. The results showed that all the bands of Quickbird could well reflect the degree of soil salinization or alkalization,and band3 of the Quickbird( 630- 690 nm) is the most"sensitive"band used to predict the pH values of saline or alkaline soil. The precision of the inversion model developed based on band1,band2 and band3 was the highest( 82. 41%). The quantitative model of Quickbird to predict soil pH values provides a new approach for monitoring soil salinization or alkalization.更多还原