Photography measured-value magnification improves local correlation maximization-complementary superiority method of hyperspectral analysis of soil total nitrogen

Abstract Accurate remote retrieval of soil total nitrogen (TN) content is a challenging task due to many factors including soil color. The objective of this study was to investigate whether the photography measured-value magnification (PMM) method improves estimation accuracy by reducing the influence of soil color, when combined with the local correlation maximization-complementary superiority (LCMCS) method. Soil samples were collected from three areas of subsided land in Renqiu, Changzhou, and Fengfeng District, all located in Hebei Province, China. Soil spectral reflectance was measured using an ASD FieldSpec 3 spectrometer in a laboratory environment and the TN content was determined by the Kjeldahl method. After PMM analysis, the LCMCS method was used to build models for TN estimation. The results indicate that the LCMCS model of the PMM group produced lower prediction errors (Coefficient of determination [R 2 ] = 0.893, root mean square error of validation [RMSEV] = 0.090, mean relative error of validation [MREV] = 5.721%) when compared with the local correlation maximization-partial least squares regression (LCM-PLSR) model of the PMM group and the models of the conventional group. Overall, the PMM method combined with LCMCS has great potential to improve the estimation accuracy of TN content and enriches the choice of observation method.