基于多源遥感数据融合的土壤侵蚀强度分析

Analysis of soil erosion intensity based on multi-source remote sensing data fusion

单一遥感数据源存在空间与时间分辨率相互制约的问题,基于多源遥感数据融合生成高时空分辨率影像,可提高土壤侵蚀动态变化监测时效和精度。研究分析了多种遥感融合方法,并将最优融合结果应用于江苏省徐州市贾汪区土壤侵蚀监测,结果表明:1)先进行全色与多光谱影像融合,再进行时空自适应融合的方法(STARFM_NND)在目视评价、定量评估等方面表现最优;2)基于STARFM_NND融合得到的土壤侵蚀面积监测结果与真实高分影像反演结果相比,精度优于97%;3)贾汪区2021年水土流失总面积71.65 km^(2),轻度、中度和强烈侵蚀面积分别为64.25 km^(2)、6.35 km^(2)和1.05 km^(2)。研究提出的STARFM_NND融合策略能够实现低成本、高精度的地物类型变化预测和土壤侵蚀监测,推动了遥感技术在水土保持方面的应用。

The single remote sensing data has the problem of mutual restriction between spatial and temporal resolution.Based on multi-source remote sensing data fusion,a high spatial and temporal resolution image can be generated,which can improve the timeliness and accuracy of soil erosion dynamic change monitoring.A variety of remote sensing fusion methods were studied and analyzed,and the optimal fusion result were applied to soil erosion monitoring in Jiawang District,Xuzhou City,Jiangsu Province.The results showed that:1)The method of panchromatic and multispectral fusion first,then operating the Spatial and Temporal Adaptive Reflection Fusion Model(STARFM_NND)performs best in visual evaluation and quantitative evaluation;2)The precision of soil erosion area monitoring results obtained from STARFM_NND fusion result was exceed 97%compared with the real high-resolution image inversion results;3)The total area of soil loss in this area was 71.65 km^(2),accounting for 11.71%of Jiawang District,including 64.25 km^(2)of mild erosion,6.35 km^(2)of moderate erosion and 1.05 km^(2)of strong erosion.The STARFM_NND fusion strategy proposed in this study could realize low-cost and high-precision prediction of land use types change and soil erosion monitoring,and promoted the application of remote sensing technology in water and soil conservation.