Study on Relationship between Soil Calcium Carbonate Content and Inversion Effect Based on Visible Near-infrared Reflectance Spectra
LIN Ka,LI Decheng,LIU Feng,ZHANG Ganlin
DOI: https://doi.org/10.11766/trxb201707170240
2018-01-01
Abstract:[Objective] Soil visible near-infrared reflectance spectra contains large volumes of information on soil physical and chemical properties, which implies that it is feasible to use soil spectra to invert soil properties quantitatively. Is it the higher the property value, the higher the inversion accuracy? However, at present, it is still unclear how to relate quantitatively effects of inversions to soil property contents. [Method] Therefore, this study selected soil calcium carbonate content as the target attribute for exploration of quantitative relationship between spectral inversion effect and calcium carbonate content. A total of 292 soil samples were collected out of the genetic horizons of 69 typical Aridosols profiles in the Heihe River Basin, Northwest China, for analysis of calcium carbonate contents with the gasometric method and acquisition of visible near-infrared reflectance spectra with a Cary5000 spectrophotometer. Based on the characteristics of the distribution of calcium carbonate content in the typical study area, 11 identical sample size subsets(A)and 5 similar dispersion subsets(B)were established with sample size and data dispersion (coefficient of variation) as the criteria for dataset partitioning, and the partial least-squares regression (PLSR) method was used to invert calcium carbonate content from the spectral curves.[Result] Results show that calcium carbonate in the Aridosols of the Heihe River Basin varied in the range of 4.86 g kg-1~236.03 g kg-1in content with an average of 103.07 g kg-1. Soil samples with calcium carbonate content varying in the range of 30~60 g kg-1and of 120~150 g kg-1, were in dominancy, accounting for 21.4% and 32.6% of the total, respectively. As a whole, the soil is high in calcium carbonate content, which is consistent with the characteristics of Aridosols being rich in calcium carbonate. With the PLSR, modeling was performed for prediction of calcium carbonate contents of the soil samples in the 11 A subsets. RPD of the validation set of each subset ranged between 0.92 and 1.04, fluctuating around 1 with no obvious features of variation, which indicates that calcium carbonate content does not have much impact on prediction or inversion of soil calcium carbonate content, using visible near-infrared reflectance spectra. Modeling was also done for prediction of calcium carbonate content in 5 B subgroups, with a similar result. [Conclusion] Therefore, soil calcium carbonate content is not the main factor affecting the prediction using spectra, which is inconsistent with the qualitative knowledge the researchers already have in mind. Calcium carbonate can enhance spectral reflectance of visible near-infrared bands, but the effect is not so significantly reflected in using the visible near-infrared spectral reflectance to inverse soil calcium carbonate content. Therefore, it seems unnecessary to divide calcium carbonate samples by content of soil calcium carbonate when using spectra to predict calcium carbonate contents. Whether the conclusion is applicable to other soil properties needs to be further verified, and how to improve accuracy of the prediction of target attribute will be the focal point of the next phase of the study.