|
|||||
|
|
Development of a national VNIR soil-spectral library for soil classification and prediction of organic matter concentrations |
|
| Authors: | Zhou Shi QianLong Wang Jie Peng WenJun Ji HuanJun Liu Xi Li Raphael A Viscarra Rossel |
| Institution: | 1. College of Environmental and Resource Sciences, Zhejiang University, Hangzhou, 310058, China 5. Cyrus Tang Center for Sensor Materials and Application, Zhejiang University, Hangzhou, 310058, China 2. College of Plant Science, Tarim University, Alar, 843300, China 3. College of Resources and Environment, Northeast Agricultural University, Harbin, 150030, China 4. CSIRO Land & Water, Bruce E. Butler Laboratory, Canberra, ACT, 2601, Australia |
| Abstract: | Soil visible-near infrared diffuse reflectance spectroscopy (vis-NIR DRS) has become an important area of research in the fields of remote and proximal soil sensing. The technique is considered to be particularly useful for acquiring data for soil digital mapping, precision agriculture and soil survey. In this study, 1581 soil samples were collected from 14 provinces in China, including Tibet, Xinjiang, Heilongjiang, and Hainan. The samples represent 16 soil groups of the Genetic Soil Classification of China. After air-drying and sieving, the diffuse reflectance spectra of the samples were measured under laboratory conditions in the range between 350 and 2500 nm using a portable vis-NIR spectrometer. All the soil spectra were smoothed using the Savitzky-Golay method with first derivatives before performing multivariate data analyses. The spectra were compressed using principal components analysis and the fuzzy k-means method was used to calculate the optimal soil spectral classification. The scores of the principal component analyses were classified into five clusters that describe the mineral and organic composition of the soils. The results on the classification of the spectra are comparable to the results of other similar research. Spectroscopic predictions of soil organic matter concentrations used a combination of the soil spectral classification with multivariate calibration using partial least squares regression (PLSR). This combination significantly improved the predictions of soil organic matter (R 2 = 0.899; RPD = 3.158) compared with using PLSR alone (R 2 = 0.697; RPD = 1.817). |
| Keywords: | diffuse reflectance spectroscopy vis-NIR soil organic matter soil spectral library China |
| 本文献已被 CNKI SpringerLink 等数据库收录！ | |