[Research on the Method for Rapid Detection of Soil Moisture Content Using Spectral Data].

Song Tao,Bao Yi-dan,He Yong
DOI: https://doi.org/10.3964/j.issn.1000-0593(2009)03-0675-03
2009-01-01
Abstract:Spectroscopy technique is one of the qualitative and quantitative analytical techniques developed quickly in recent years. The spectral analysis is a fast and non-destructive method and has been used in many fields such as oil industry, food industry and so on. In the present paper, the spectral band sensitive to soil moisture content was found from the visible/near infrared spectra and a monadic linear regression model based on the data of sensitive spectral band was applied to develop a method for rapid detection of soil moisture content. The spectral data of 52 soil samples were collected by using FieldSpec HandHeld spectroradiometer made by ASD (Analytical Spectral Device) company in the US, and the data of soil moisture content were obtained by experiment. The spectral band sensitive to soil moisture content was achieved by correlation coefficient method. Then, the data of sensitive spectral band were used to build monadic linear regression model of soil moisture content. Finally, the model was employed for the prediction of soil moisture content. Correlation coefficient (r) of prediction and root mean square error of prediction (RMSEP) were used as the evaluation standards. The results indicated that the r and RMSEP for the prediction of soil moisture content were 0.966 5 and 0.012 1 respectively. Thus, it is concluded that the method used in this paper is an available method for the rapid detection of soil moisture content based on the visible/near-infrared spectra.
What problem does this paper attempt to address?