基于开裂土壤纹理特征及光谱响应的苏打盐碱土含盐量在线测量方法研究

It is urgent and important to determine the area of saline-alkali soils in Songnen Plain as well as rapidly and accurately measure the salt contents of these soils, which is very significant for determining the salinity levels, ameliorating the saline-alkali soils and increasing the crop yields. Salt mineral contents can change the mechanical properties of saline-alkali soils and affect the desiccation-cracking process of these soils to a large content. In this project, the surface states of soil samples will be monitored during water evaporates in order to study the effects of water content and salt content on desiccation-cracking process of the saline-alkali soils in Songnen Plain. After that, quantitative relationship will be performed among the water contents, the salt contents and the texture features in order to establish prediction models for salt contens of these saline-alkali soils. Finally, a quick and non-destructive measurement method as well as a measurement standard for the salt contents of soils will be proposed, which can provide important data support for the amelioration of saline-alkali soils. The volume scattering from profiles of the cracked regions and the effects of salt mineral migration on the spectral relectance of cracked soil samples will be also studied in this project. At the same time, spectral response to salt contens of cracked soil samples will also be analysed based on hyperspectral mixed pixel decomposition, which will provide an appropriate methed for the experiment of soil salinity inersion based on hyperspectral remote sensing.

加强对松嫩平原盐碱土的监测力度，快速、准确地测量土壤的盐分含量，对于确定土壤盐碱化程度、指导盐碱化土壤改良、实现粮食增产等方面都具有重要的意义。土壤盐分矿物的成分及含量能够改变盐碱土的力学特性，并在很大程度上控制其失水收缩开裂的过程。本项目拟根据松嫩平原盐碱土收缩性较高的特点，通过对土壤样本开裂过程的动态观测和分析，揭示含水量及含盐量对土壤失水收缩开裂过程的影响。通过研究土壤开裂过程中水、盐及表面纹理特征间的定量关系，建立土壤盐分含量的预测模型。在此基础上发展土壤盐分含量的在线测量方法并确定测量标准，实现对盐碱土盐分含量的无损、快速、准确地测量，从而为盐碱土改良提供数据和决策上的有力支持。本项目还拟研究裂纹剖面的体散射及盐分矿物的运移状况等因素对开裂土壤光谱反射率的影响；同时还拟在干燥条件下，通过混合像元分解的手段分析开裂土壤的整体光谱响应，优化高光谱反演土壤含盐量的实验方法和研究手段，从而为利用遥感手段进一步提高苏打盐碱土含盐量的反演精度提供理论准备和实验支持。

土壤盐碱化严重影响了土壤结构，使土壤的密度、联结性、渗水性及土体强度等物理性质极度恶化，这不仅严重制约了农业的可持续发展，也降低了土壤抵御自然灾害的能力。因此，加强盐碱土的监测范围和监测力度，高效、快速、准确地测量土壤的含盐量水平就显得尤为迫切。. 松嫩平原土壤盐碱化程度非常严重，是中国主要的土壤盐碱化区域之一，同时也是世界三大苏打型盐碱土集中分布区域之一。本项目从松嫩平原苏打盐碱土膨胀收缩性较高这一特性出发，通过可控实验模拟了苏打盐碱土失水收缩开裂的过程。在土壤样本开裂过程中测量了裂纹长度、裂纹面积、裂纹率等几何参数，研究了裂纹图像灰度共生矩阵统计纹理特征提取方法。同时，选取了灰度共生矩阵的最佳计算参数及其对应的裂纹图像类型，并建立了开裂土壤表面裂纹的灰度共生矩阵统计纹理特征量提取标准。在定量分析了这些几何及纹理特征参数随土壤含水量的动态变化规律后，本项目建立了基于裂纹几何特征及统计纹理特征的苏打盐碱土含盐量预测模型，从而为判断盐碱土的可改良性提供了快速、有效的技术手段，为当地确定土壤的盐碱化程度、指导盐碱化土壤改良、开展农业生产和恢复生态提供必要的数据支持，具有重要的现实意义。. 本项目进行了不同表面状况下的盐碱土样本高光谱反射率测量，定量研究了土壤含盐量及表面状况对样本光谱特征的影响，同时考虑了裂纹处的体散射作用对电磁波传输过程的影响。利用对比度纹理特征量作为裂纹状况的表征参数，从混合像元分解的角度研究土壤裂纹与含盐量对光谱特征的综合影响，并在此基础上建立了土壤盐分参数的一元和多元线性预测模型，实现了对土壤含盐量等理化参数的光学遥感反演实验方法及光谱处理手段的优化，从而为将来利用航空和航天等大范围的土壤含盐量高光谱遥感检测提供理论准备和实验支持。.