基于冻土四相高精度辨识的黑土孔隙结构与相变耦合

项目名称： 基于冻土四相高精度辨识的黑土孔隙结构与相变耦合

项目编号： No.41501283

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 地质学

项目作者： 赵玥

作者单位： 北京林业大学

项目金额： 20万元

中文摘要： 冻融交替是改变黑土结构、加剧该区土壤侵蚀的关键因子，但关于冻融过程中冰-水两相的交替转化如何改变土壤孔隙数量与形态特征的研究鲜有报道，尤其是冻土研究过程中因图像分割精度差，缺乏有效、定量辨识冰-水相态，导致土壤孔隙结构与相变间关系无法实现定量表达。为此，本项目以季节性冻融发生频率与强度不同的黑土耕作区与非耕作区土壤为研究对象，基于不同冻结条件处理下同位CT扫描图像以及Canny等边缘提取算子的多层级图像分割手段，确定冻土CT图像内各组成介质的边界范围，实现冻土四相的高精度识别；在此基础上，通过黑土孔隙结构与孔隙内冰、水形态与含量等指标参数的测定与计算，量化冻土孔隙结构与冰-水相变过程，进而构建冻融过程中冰-水相变与土壤孔隙结构特征间耦合关系模型，定量阐明冻土冰-水相变对黑土耕作土壤孔隙结构变化与重构的影响，为进一步揭示黑土结构与功能对季节性冻融的响应机制奠定理论基础。

中文关键词： 黑土；土壤孔隙结构；冰-水相变；CT扫描图像；图像分割

英文摘要： Freezing and thawing alternation is critical to alter black soil structure and accelerate soil erosion in black soil region of Northeast China. However, lack of understanding how ice-water phase transformation changes quantity and morphology of soil pore leads to an impossibility of quantifying the relationship between soil pore structure and phase transformation, with the limit of image segmentation accuracy. To fill the gap, top soil and subsoil from farmland was sampled and was subjected to different simulative freezing conditions. After hierarchical image segmentation based on Canny edge detection, the four phase boundary of frozen soil was identified with higher precision, and the quantity and morphology of soil pore and ice was measured. Based on the quantification of soil structure and ice-water phase transformation, coupling relationship of them was found. The research founding will help to understanding the effect of phase change when soil freezing on soil pore structure and its reconstruction; and lay a theoretical foundation for reveal the mechanism that seasonal freezing and thawing cycles affect the structure and function of black soil.

英文关键词： black soil;soil pore structure;ice-water phase transformation;computed tomograph scanning image;image segmentation