Abstract:

The loess hilly area in China is ecologically fragile, and the soil erosion situation is grim. Accurately assessing the spatial and temporal variation of soil erosion in loess hilly areas and deep exploration of the underlying driving mechanisms can provide important theoretical support for local soil and water conservation. This study investigated the Zuli River Basin as the research objective, based on the RUSLE soil erosion assessment model, combining the transfer matrix, spatial autocorrelation, geographic detector, and other geostatistical analysis tools, and evaluated the spatial distribution pattern of the soil erosion modulus and driving force between 2000 and 2020. The results showed that 1) the soil erosion intensity class in the Zuili River Basin from 2000 to 2020 was dominated by slight and mild, and the average soil erosion modulus during the 2020, a period decreased from 1.299×10³ t∙km⁻²∙a⁻¹ to 635 t∙km⁻²∙a⁻¹, i.e., the erosion intensity class was reduced from mild erosion to slight erosion. 2) From 2000 to 2010, approximately 1.968×10³ square kilometers of the soil erosion area developed in a benign direction, accounting for 18.4% of the total watershed area. From 2010 to 2020, the soil erosion area increased by approximately 3.089×10³ square kilometers, accounting for 28.9% of the total watershed area, and the effects of soil and water erosion control were remarkable. 3) There is an obvious spatial aggregation of soil erosion in the Zuli River Basin, in which the “high-high” aggregation area is mainly distributed in the junction of Huining County and Anding District, and the mountainous hills in the northwest of the basin, and the “low-low” aggregation area is mainly distributed in the basin, and the “low-low” aggregation area is mainly distributed in the basin. The “high-high” gathering area is mainly distributed in the northern and southwestern coastal plain area of the watershed, and the area of the “high-high” and “low-low” gathering types has decreased annually over the past 20 years. The area of the “high-high” and “low-low” aggregation types has decreased annually over the past 20 years. 4) Precipitation is the primary factor that affects soil erosion. Terraces play an important role in improving soil erosion, and the joint action of natural factors and human activities has shaped the spatial distribution patterns of soil erosion. In the future, soil and water conservation management should focus on at-risk areas with low temperature and drought, high altitude, steep terrain, low vegetation cover, a high degree of soil sand pulverization, and few terraced fields.

Key words: RUSLE modle, Zuli River Basin, transfer matrix, spatial autocorrelation, geodetector

摘要：

黄土丘陵区作为中国生态脆弱区，水土流失形势严峻，准确评估其土壤侵蚀时空变化规律，深入探究其背后的驱动机理，对于指导当地水土保持工作具有重要科学意义。以祖厉河流域为研究对象，基于RUSLE土壤侵蚀评估模型，结合转移矩阵、空间自相关和地理探测器等地学统计分析工具，对2000—2020年间土壤侵蚀模数的空间分布格局及背后驱动力进行评估。结果显示，1）2000—2020年祖厉河流域土壤侵蚀强度等级以微度和轻度为主，20年间平均土壤侵蚀模数由1.299×10³ t∙km⁻²∙a⁻¹下降至635 t∙km⁻²∙a⁻¹，即侵蚀强度等级由轻度侵蚀降为微度侵蚀。2）2000—2010年期间，约有1.968×10³ km²的区域土壤侵蚀状况向良性方向发展，占流域总面积的18.4%；而在2010—2020年间，约有3.089×10³ km²的区域土壤侵蚀情况得到改善，占流域总面积的28.9%，水土流失治理效果显著。3）祖厉河流域土壤侵蚀存在明显的空间聚集性，其中“高-高”聚集区主要分布在会宁县与安定区交界处以及流域西北部的山地丘陵地带，“低-低”聚集主要分布在流域北部和西南部沿岸平原地区，20年来该“高-高”和“低-低”聚集类型面积逐年减少。4）降水是影响土壤侵蚀的首要因素，梯田在改善土壤侵蚀状况中发挥了重要作用，自然因子与人类活动因子之间的共同作用塑造了土壤侵蚀空间分布格局。未来水土保持治理重点应聚焦于低温干旱、高海拔、地势陡峭、植被覆盖低、土壤砂粉化程度高和梯田分布少等条件的风险区。

关键词: RUSLE模型, 祖厉河流域, 转移矩阵, 空间自相关, 地理探测器