近年来黄河下游土地次生盐渍化呈现反复和逐渐加剧的态势，对农业生产和生态安全造成危害。本文以黄河三角洲小开河引黄灌区为研究区，基于野外实地调查的土壤盐分含量以及Landsat8 OLI多光谱影像，分析土壤样品的光谱曲线特征，利用诊断指数法选取诊断指数较大的波段反射率数据作为自变量，土样盐分数据为因变量，分别采用多元线性回归模型和BP神经网络模型构建土壤含盐量反演模型。结果表明：土壤盐渍化程度越高，影像光谱反射率越低，且在近红外波段反射率最高；BP神经网络模型的反演精度优于传统的多元线性回归模型，其R2为0.9808，RMSE为1.0595，平均相对误差为15.4%，拟合精度较高，能够为灌区盐渍化治理提供基础依据。

In recent years，secondary salinization of the lower reaches of the Yellow Ｒiver has been repeated and gradually intensified， causing damage to agricultural production and ecological security． This paper takes the Xiaokaihe Irrigation District of the Yellow Ｒiver Delta as the study area． Based on the soil salt content and Landsat8 OLI multi－spectral imagery，the spectral curve characteristics of soil samples are analyzed． According to the diagnostic index method，the band reflectance data with larger diagnostic index is an independent variable and the salt salinity is the dependent variable，and the soil salt inversion model is constructed by multiple linear regression model and BP neural network model． The results show that the higher the degree of soil salinization，the lower the spectral reflectance of the image and the highest reflectance is in the near－infrared． The inversion accuracy of BP neural network model is better than that of the traditional multiple linear regression model． It＇s Ｒ2 is 0．980 8 and ＲMSE is 1．059 5，the average relative error is 15．4%，and the fitting accuracy is high，which can provide a basis for the salinization treatment in the irrigation district．