POWRE: Soil Hydraulic Properties as Affected by Soil Solution Chemical Composition

0074841Lebron该项目解决了量化溶液组成对土壤水力性质的影响的问题。 了解这些特性对于评估污染物在土壤中的迁移和随后的补救、预测CO2的释放和固存以及解释和改进对遥感土壤数据的使用，例如根据电阻率数据确定盐度和根据卫星信息确定含水量，都是必要的。 土壤中水分和溶质的流动和运移受土壤孔隙的大小、几何形状和特征的控制。 土壤孔隙的大多数特征是微观的，如粗糙度和圆形度。 传统的液体分布、流动和溶质输运模型仅依赖于圆柱形孔，而忽略了表面积、棱角性和连通性的作用。 神经网络已被用于预测土壤持水性能的宏观和微观参数。 这些模型具有巨大的潜力，以获得土壤传递函数（PTFs）预测水力参数。 然而，文献中的PTF没有考虑土壤溶液的化学成分，也没有考虑土壤中粘土矿物的矿物学数据。 粘土矿物和化学成分是膨胀性粘土土壤水力特性的关键决定因素，特别是当它们受到碱性影响时。 该项目的主要目标是量化的盐度，碱度和pH值的几何形状，大小和分布的土壤孔隙空间的影响。 这种量化将提供信息开发PTFs使用神经网络和自助方法，旨在提高我们的能力，预测土壤水力特性。

0074841LebronThis project addresses the problem of quantifying the influence of solution composition on soil hydraulic properties. The knowledge of these properties is necessary to assess transport of contaminants in soil and subsequent remediation, to predict release and sequestration of CO2, and to interpret and improve the use of remotely sensed soil data such as determination of salinity from electrical resistivity data and determination of water content from satellite information. Flow and transport of water and solutes in soils are controlled by size, geometry and characteristics of the soil porosity. Most of the characteristics of the soil pores are microscopic, such as roughness and circularity. Conventional models of liquid distribution, flow and solute transport rely solely on cylindrical pores, while ignoring the role of surface area, angularity, and connectivity. Neural networks have been used to predict water retention properties in soils using macro and microscopic parameters. These models have tremendous potential to derive pedotransfer functions (PTFs) to predict hydraulic parameters. However the PTFs available in the literature do not consider the chemical composition of the soil solution nor data considering the mineralogy of the clay minerals in the soil. Clay mineral and chemical composition are known to be critical determinants of the hydraulic properties in soils with swelling clays, especially if they are affected by sodicity. The main objective of this project is to quantify the influence of salinity, sodicity and pH on the geometry, size, and distribution of the soil pore space. This quantification will provide the information for developing PTFs using neural network and bootstrap methodology and is intended to improve our capability to predict soil hydraulic properties.