Various forms of microtopography are widely distributed and exist in the loess area of northern Shaanxi. Microtopography is also an important consideration for near-natural vegetation restoration models. The soil quality under the different microtopography types affects which near-natural vegetation configuration and restoration mode should be applied. Therefore, this paper took the microtopography of the Hegou watershed, which was restored by the natural exclosure method, and that of the Chaigou watershed, which was restored by artificial afforestation, in the loess region of northern Shaanxi as the research objects. A comprehensive evaluation of the soil quality in the study area was carried out by assessing 22 soil indicators and performing sensitivity analysis and principal component analysis. The results showed the following: (1) Soil available phosphorus, total nitrogen, organic matter, sucrase activity, urease activity, maximum water-holding capacity, field capacity, capillary water-holding capacity, noncapillary porosity, natural water content, available nitrogen, available potassium, cation exchange capacity, total phosphorus, total potassium, alkaline phosphatase and catalase constituted a sensitive-indicator data set for soil quality evaluation in the study area. (2) There was a significant linear fit between the sensitive-indicator data set and the all-indicator data set (y = 1.032x − 0.027, R² = 0.96 (P < 0.05)). (3) The microtopography types under natural restoration were ranked by soil quality as follows: collapse (0.672) > platform (0.565) > shallow gully (0.529) > gully (0.515) > undisturbed slope (0.422) > scarp (0.351). The microtopography types under artificial afforestation were ranked by soil quality as follows: gully (0.731) > platform (0.66) > undisturbed slope (0.431) > shallow gully (0.411) > scarp (0.395) > collapse (0.221).

陕北黄土区广泛分布着各种形态的微地形。微地形也是近自然植被恢复模型的重要考虑因素。不同微地形类型下的土壤质量会影响应采用哪种近自然植被配置和恢复模式。因此，本文以陕北黄土区自然围护法修复的河沟流域和人工造林修复的柴沟流域微地形为研究对象。通过对22项土壤指标进行评价，进行敏感性分析和主成分分析，对研究区土壤质量进行综合评价。结果显示如下：(1) 土壤速效磷、全氮、有机质、蔗糖酶活性、脲酶活性、最大持水量、田间持水量、毛细管持水量、非毛细管孔隙度、天然含水量、速效氮、速效钾、阳离子交换量、总磷、总钾、碱性磷酸酶和过氧化氢酶构成了研究区土壤质量评价的敏感指标数据集。(2) 敏感指标数据集和全指标数据集之间存在显着线性拟合（y = 1.032x − 0.027，R 阳离子交换量、总磷、总钾、碱性磷酸酶和过氧化氢酶构成了研究区土壤质量评价的敏感指标数据集。(2) 敏感指标数据集和全指标数据集之间存在显着线性拟合（y = 1.032x − 0.027，R 阳离子交换量、总磷、总钾、碱性磷酸酶和过氧化氢酶构成了研究区土壤质量评价的敏感指标数据集。(2) 敏感指标数据集和全指标数据集之间存在显着线性拟合（y = 1.032x − 0.027，R²  = 0.96 ( P  < 0.05))。(3)自然恢复下的微地形类型按土壤质量排序为：塌陷(0.672)>台地(0.565)>浅沟(0.529)>沟(0.515)>原坡(0.422)>陡坡(0.351)。人工造林微地形类型按土壤质量排序为：沟壑（0.731）>平台（0.66）>原坡（0.431）>浅沟（0.411）>陡坡（0.395）>塌陷（0.221）。