基于线性源法与图像处理的土壤饱和导水率快速测量方法

土壤饱和导水率是计算土壤剖面水通量以及设计灌溉和排水系统的重要参数，其测量准确与否直接影响各类水文和水动力学模型的预测精度。然而，现有土壤饱和导水率测定方法费时费力，给土壤水动力学研究工作带来了诸多不便。为此，该研究提出了一种基于线性源入流法与手机图像处理相结合的土壤饱和导水率快速测量方法。该方法首先利用手机拍照获取图像记录充分供水条件下线性水流在土壤表面扩散的过程，图像经处理后计算出土壤表面湿润面积及其随时间的变化关系，然后根据线性源入流法估算的土壤稳态入渗率来测得土壤饱和导水率，并与传统的定水头标准法测得的饱和导水率进行对比。结果表明：图像经畸变校正与二值化处理之后计算出栓皮栎林区土壤、油松林区土壤和砂壤土表面湿润面积与时间具有较好的幂指数关系，决定系数R2分别为0.994、0.995和0.998；在此基础上，采用线性源入流法测量栓皮栎林区土壤、油松林区土壤和砂壤土的稳态入渗率（即土壤饱和导水率）分别为23.40±1.21、23.86±1.83和22.99±2.26 mm/h，同时使用定水头标准法测量三种土样得到的饱和导水率分别为24.41±1.53、24.26±0.37和23.81±0.10 mm/h，与定水头标准法相比，该研究提出的土壤饱和导水率测量方法的相对误差分别为4.14%、1.64%和3.42%。可见，该研究提出的测定方法较为合理、简便、准确，可为获取土壤饱和导水率提供一种新的测量手段，后续研究会将该方法用于野外环境下土壤饱和导水率的原位测定，并验证该方法的准确性。

Rapid measurement of soil saturated hydraulic conductivity using linear source inflow method and image processing

Soil saturated hydraulic conductivity is an important parameter to calculate the water flux of soil profile in the irrigation and drainage system for agricultural production and resource utilization. An accurate and rapid measurement directly determines the prediction accuracy of various hydrological and hydrodynamic models. However, there are great differences in the saturated hydraulic conductivity in space, due to the different soil texture, structure, as well as physical and chemical properties. The current measurement approaches are time-consuming and laborious for the soil saturated hydraulic conductivity in the soil hydrodynamics. In this study, a rapid measurement was proposed for the soil saturated hydraulic conductivity combining the linear source inflow method and mobile phone image processing. Soil saturated hydraulic conductivity refers to the water flux per unit area of soil perpendicular to the water flow direction under the action of unit water potential gradient, when all pores of soil are filled with water. Soil infiltration rate refers to the rate at which water penetrates into the ground at any given time. The soil infiltration rate is close to a fixed value when the soil water is close to or reaches saturation in the process of infiltration. Therefore, this value is usually considered to be close to the actual soil saturated hydraulic conductivity. The specific procedure was taken as the core idea. Firstly, the soil infiltration rate was measured to consider the infiltration rate, when the water was saturated as the soil saturated hydraulic conductivity. The diffusion process of linear water flow on the soil surface was recorded by a mobile phone. The wetted area of the soil surface and the change curve with time were calculated, after the image processing operations, such as cutting, distortion correction, binarization, and count the number of target pixels in the binarized image. The water balance principle was selected to estimate the change process of soil steady-state infiltration rate using the linear source inflow method. The stable infiltration rate was taken as the soil saturated hydraulic conductivity, and then compared with the saturated hydraulic conductivity measured by the traditional constant head method. The results show that after the distortion correction and image processing, the surface wetted area of Quercus variabilis BL. forest soil, Pinus tabulaeformis Carr. forest soil and sandy loam presented an excellent power exponential relationship with the time, where the determination coefficients R2 were 0.994, 0.995, and 0.998 respectively. The steady-state infiltration rates (i.e. soil saturated hydraulic conductivity) of Quercus variabilis BL. forest soil, Pinus tabulaeformis Carr. forest soil, and sandy loam measured by the linear source inflow method were 23.40±1.21, 23.86±1.83, and 22.99±2.26 mm/h, respectively. At the same time, the values of saturated hydraulic conductivity for the three soil samples measured by the constant head standard method were 24.41±1.53, 24.26±0.37, and 23.81±0.10 mm/h respectively. The relative errors of the two methods were 4.14% 1.64% and 3.42%, respectively. It can be seen that the measurement was reasonable, simple, and accurate under the condition of sufficient water supply. The new measurement method can be widely expected to quickly acquire the soil saturated hydraulic conductivity.