基于源汇景观单元的流域土壤侵蚀风险格局识别

人类活动影响下的土地利用及其景观格局会在一定程度上影响流域土壤侵蚀的发生发展。选取位于三峡库区的綦江流域作为研究区,利用2015年航空影像数据、数字高程模型和土壤数据库,进行水文响应单元的划分,以此作为研究区的源汇景观单元。综合景观类型、土壤和坡度对土壤侵蚀影响的贡献,构建源汇景观单元权重,在此基础上,对景观空间负荷对比指数进行修正并进行土壤侵蚀风险格局识别,最后利用修正的通用土壤侵蚀方程进行土壤侵蚀的模拟,以此验证风险格局的合理性,并综合分析了源汇景观空间特征:源汇景观单元组成结构、权重和土壤侵蚀风险。结果表明:(1)源汇景观单元权重大的地区主要分布在中低山区向低丘缓坡区过渡的地带,坡度较大、土壤可蚀性较高,以及水田、旱地和居民点的源汇景观单元也较为集中分布。(2)各子流域的景观空间负荷对比指数与平均土壤侵蚀模数具有显著正相关关系,因此基于源汇景观单元并赋予其权重的景观空间负荷对比指数能较好地反映流域内部土壤侵蚀规律,可作为流域土壤侵蚀风险评价的有效方法之一。(3)依据各子流域的景观空间负荷对比指数特征可将库区綦江流域划分为五大土壤侵蚀景观风险区:北部沿江地区各子流域耕地分布较为集中且相对水流路径较短,以及林草地较少,土壤侵蚀风险大;中部丘陵地区源景观单元分布较为分散,景观空间分布不均衡,存在一定的土壤侵蚀风险;南部中低山区以林地汇景观分布为主,源景观分布相对较小,土壤侵蚀风险较低。

Identification of soil erosion risk patterns in a watershed based on source-sink landscape units

The coupling relationships between landscape patterns and ecological processes have long been in landscape ecology. Land use and landscape patterns under the impact of anthropogenic activity affect the occurrence and development of soil erosion in watersheds. In the present investigation, the Qijiang watershed, located in Jiangjin District of the Three Gorge Reservoir Region (TGRR), was selected as study area, and hydrological response units were divided as "source-sink" landscape units by using aerial image data for 2015, DEM data, and a soil dataset. The soil erosion weight of the "source-sink" landscape units was assigned by combining with landscape type weight, soil erodibility weight, and slope weight, and the location-weighted landscape index (LWLI) related to relative flow path was calculated to identify soil erosion risk patterns. Finally, the soil erosion modulus of the watershed was simulated using the revised universal soil loss equation (RUSLE) to verify the rationality of the soil erosion risk patterns. The results showed that paddy field and dry land in the source landscape and woodland in the sink landscape were dominant landscape in the spatial structure of the "source-sink" landscape units. Next is residential area in the source landscape and grassland in the sink landscape, and distinct areas characteristics were identified in the Qijiang watershed. Regions with higher soil erosion weight of landscape units were mainly located in the transition zones from the low mountains area to the hilly slope area, where slope and soil erodibility were higher. Source landscape units of paddy field, dry land, and residential area were also concentrated. The soil erosion modulus was positively correlated with the LWLI, which indicated that the LWLI could correctly reflect the effect of source and sink landscapes on soil erosion, and could be used as a valid tool to evaluate the potential risk of soil erosion. Five landscape risk zones for soil erosion were divided according to the characteristics of the LWLI. In the northern region along rivers, there was more cultivated land than woodland and grassland, with a shorter relative flow path, the result of which was a greater soil erosion risk. In the central hilly region, the distribution of source landscape units was more decentralized and unbalanced, which indicated that there was a certain soil erosion risk in sub-watersheds. In the southern low mountain region, woodland was the dominant landscape, the amount of source landscape was relatively less, and the soil erosion risk was lower.