当前位置: X-MOL 学术Geocarto Int. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Estimating and Mapping of Soil Organic Matter Content in a Typical River Basin of the Qinghai-Tibet Plateau
Geocarto International ( IF 3.8 ) Pub Date : 2021-01-12
Qing Yu, Hongwei Lu, Wei Feng, Tianci Yao

Abstract

Spectroscopy is a fast, non-destructive, and cheap method, which has been widely used in the estimation of soil organic matter (SOM) concentration. This study presented a methodology to estimate and map SOM content by crop canopy reflectance spectra combining with land parameters in the Yarlung Zangbo River (YZR) basin. The reflectance spectra of the oat canopy were collected in the field, and then were processed by savitzky-golay filtering (S-G), continuous removal (CR), and first derivative of reflectivity (FDR). The principal components were extracted from the processed spectral data. Land parameters such as elevation, slope, NDVI, and land surface temperature (LST) were selected to establish 18 models to estimate SOM content. The results showed that the estimation accuracy of SOM content could be improved by combining crop canopy reflectance spectra with land parameters. FDR-BPNN model with land parameters had the best estimation effect (R2 = 0.973, MAE =0.847 g·kg−1).



中文翻译:

青藏高原典型流域土壤有机质含量估算与测绘

摘要

光谱法是一种快速,无损且廉价的方法,已被广泛用于估算土壤有机质(SOM)浓度。这项研究提出了一种通过作物冠层反射光谱结合雅鲁藏布江流域土地参数估算和绘制SOM含量的方法。在野外收集燕麦冠层的反射光谱,然后通过Savitzky-Golay滤波(SG),连续去除(CR)和反射率一阶导数(FDR)进行处理。从处理后的光谱数据中提取出主要成分。选择诸如海拔,坡度,NDVI和地表温度(LST)之类的土地参数,以建立18个模型来估算SOM含量。结果表明,结合作物冠层反射光谱与土地参数可以提高土壤有机质含量的估算精度。具有土地参数的FDR-BPNN模型具有最佳估计效果(R2 = 0.973,MAE = 0.847 g·kg -1)。

更新日期:2021-01-12
down
wechat
bug