文献类型： 外文期刊

作者： Hong, Yongsheng ¹ ; Chen, Yiyun ² ; Chen, Songchao ³ ; Shen, Ruili ⁴ ; Hu, Bifeng ⁵ ; Peng, Jie ⁶ ; Wang, Nan ¹ ; Guo, Long ⁷ ; Zhuo, Zhiqing ⁸ ; Yang, Yuanyuan ⁹ ; Liu, Yaolin ² ; Mouazen, Abdul Mounem ¹⁰ ; Shi, Zhou ¹ ;

作者机构： 1.Zhejiang Univ, Inst Agr Remote Sensing & Informat Technol Applica, Coll Environm & Resource Sci, Hangzhou 310058, Peoples R China

2.Wuhan Univ, Sch Resource & Environm Sci, Wuhan 430079, Peoples R China

3.ZJU Hangzhou Global Sci & Technol Innovat Ctr, Hangzhou 311200, Peoples R China

4.Hubei Acad Environm Sci, Wuhan 430072, Peoples R China

5.Jiangxi Univ Finance & Econ, Sch Tourism & Urban Management, Nanchang 330013, Peoples R China

6.Tarim Univ, Coll Plant Sci, Alar 843300, Peoples R China

7.Huazhong Agr Univ, Coll Resources & Environm, Wuhan 430070, Peoples R China

8.Zhejiang Acad Agr Sci, Inst Digital Agr, Hangzhou 310021, Peoples R China

9.Zhejiang Univ City Coll, Sch Spatial Planning & Design, Hangzhou 310015, Peoples R China

10.Univ Ghent, Dept Environm, Coupure Links 653, B-9000 Ghent, Belgium

11.VYTAUTAS MAGNUS Univ, Fac Engn, Dept Agr Engn & Safety, Kaunas, Lithuania

关键词： Urban soil; Soil spectral library; Soil organic carbon; Stratified modeling

期刊名称：GEODERMA （ 影响因子：7.422； 五年影响因子：7.444 ）

ISSN： 0016-7061

年卷期： 2022 年 426 卷

页码：

收录情况： SCI

摘要： Accurate quantification of urban soil organic carbon (SOC) is essential for understanding anthropogenic changes and further guiding effective city managements. Visible and near infrared (vis-NIR) spectroscopy can monitor the SOC content in a time-and cost-effective manner. However, processes and mechanisms dominating the re-lationships between SOC and spectral data in urban soils remain unknown. The main objective of this paper was to evaluate whether multiple stratification strategies (i.e., based on land-use/land-cover [LULC], pH, and spectral clustering) resulted in better predicted performance for SOC compared to the non-stratified (global) models. Results showed that regarding the non-stratified models, the convolutional neural network (CNN) model exhibited the best performance (validation R2 = 0.73), followed by Cubist (validation R2 = 0.66) and memory-based learning (validation R2 = 0.65). After LULC stratification, Cubist model achieved the best prediction (validation R2 = 0.76), improving the value of ratio of performance to interquartile distance by 0.11 compared to the global CNN model. Areas with high SOC values were mainly located in the city center. Stratification by LULC class is a promising strategy for addressing the impact of the soil-landscape diversity and complexity on vis-NIR spectral estimation of SOC in urban soil spectral library.