为了从取样、测量到评估分析全流程提高土地砾化程度监测指标——地表砾石覆盖度的监测评估精度，以内蒙古高原西部的阿拉善盟和巴彦淖尔市为研究区，设计不同尺寸的样方，开展二次平均状态取样方法研究。采用地面测量法和测量盘法对比确定小样方中砾石覆盖度的最佳测量方法，通过大、小样方平均数和中位数的统计特征分析确定大样方及样地砾石覆盖度的最佳评估方式。同时，探讨了构建“仿生态学”的必要性和初步思路。结果表明：（1） 二次平均状态法通过缩减样方尺寸、增加样本数量的方式，使样本更具代表性，且样方尺寸的减小可使测量盘的尺寸随之缩减，有助于进一步提高地表砾石覆盖度测量精度。（2） 由小样方到大样方再到样地，数据的离散程度越来越小，右偏及尖峰型分布越来越明显，研究区砾化程度的总体特征越来越凸显，同时大样方及样地的砾石覆盖度、地表单位面积砾石质量相关性较高，说明基于二次平均状态取样的土地砾化程度监测结果具有较高的精确度和稳定性，可用于改进土地砾化程度监测流程及精度，并提高野外工作效率。

To enhance the accuracy of monitoring and assessing surface gravel coverage throughout the sampling, measurement, and evaluation process, Alagxa League and Bayannur City in the western Inner Mongolian Plateau, China, were selected as the study area. Various sample sizes were designed, and research on the sampling method of the secondary average state was conducted. The optimal measurement method for small sample gravel coverage was determined by comparing the measurement disk method with the ground measurement method. The most suitable evaluation method for large sample gravel coverage was identified by analyzing the statistical characteristics of the average and median gravel coverage values in small samples. The same methodology was applied to evaluate the gravel coverage in sample plots. Additionally, the necessity and preliminary considerations for constructing imitative ecology are discussed. The results showed that: (1) The secondary average state method enhances sample representativeness through two pathways—reducing sample sizes while increasing sample quantities. Consequently, the size of the measurement disk is reduced, which further improves the accuracy of surface gravel coverage measurements. (2) Data dispersion decreases, and the characteristics of right-skewed and peaked distributions become more prominent progressively for small samples, large samples, and sample plots. This indicates that the overall characteristics of gravelized land in the study area become increasingly distinct. Gravel coverage is strongly correlated with the surface gravel mass per unit area, demonstrating that the secondary average state sampling yields highly accurate and stable results. These findings can significantly improve the procedures and accuracy of land gravelization measurement and enhance the efficiency of fieldwork.