Spectroscopy has been applied in monitoring soil nutrient concentrations. Two types of soil samples, sandy loam and silty loam, were selected as the research objects. The UV-visible near-infrared reflectance spectroscopy data and total carbon (TC), total nitrogen (TN), total phosphorus (TP), total potassium (TK), available nitrogen (AN), available phosphorus (AP), available potassium (FK), and slowly available potassium (SK) concentrations were measured. We compared the prediction results within and between two different types of soil with regard to the soil nutrient concentrations using four modeling methods, which were principal component regression (PCR), partial least squares regression (PLSR), least squares support vector machine (LS-SVM), and back propagation neural network (BPNN) models. In the prediction results within a given type of soil, LS-SVM and PLSR had better stability. In the prediction results of different types of soil, BPNN and LS-SVM had a high accuracy in most soil nutrient concentrations. By comparing different modeling methods, this study provides a basis for the subsequent selection of suitable models based on spectral technology to establish various soil nutrient models.
Similar content being viewed by others
References
D. Summers, M. Lewis, and B. Ostendorf, Ecol. Indic., 11, No. 1, 123–131 (2011).
M. Vohland, M. Ludwig, and S. Thiele-Bruhn, Geoderma, 223–225, No. 1, 88–96 (2014).
S. M. Tiquia, J. Lloyd, and D. A. Herms, Appl. Soil Ecol., 21, No. 1, 31–48 (2002).
H. Q. Ding and Q. P. Lu, Spectrosc. Spectr. Anal., 32, No. 1, 88–91 (2012).
L. Luan, Y. Wang, and X. Li, J. Near Infrared Spectrosc., 24, No. 4, 363–372 (2016).
A. M. Rady, D. E. Guyer, and W. Kirk, J. Food Eng., 135, No. 2, 11–25 (2014).
W. Ng, B. P. Malone, and B. Minasny, Geoderma, 289, 150–160 (2017).
A. Sakudo, Clin. Chim. Acta, 455, No. 3, 181–188 (2016).
S. Jia, H. Li, and Y. Wang, Geoderma, 268, 92–99 (2016).
D. J. Brown, K. D. Shepherd, and M. G. Walsh, Geoderma, 132, No. 3, 273–290 (2006).
A. V. Bilgili, H. M. V. Es, and F. Akbas, J. Arid. Environ., 74, No. 2, 229–238 (2010).
F. Feyziyev, M. Babayev, and S. Priori, Open J. Soil Sci., 06, No. 3, 52–58 (2016).
T. Naes and H. Martens, J. Chemom., 2, No. 2, 155–167 (1988).
S. Wold, M. Sjöström, and L. Eriksson, Chemom. Intell. Lab. Syst., 58, No. 2, 109–130 (2001).
J. A. K. Suykens, J. Vandewalle, Neural Process. Lett., 9, No. 3, 293–300 (1999).
R. Hecht-Nielsen, Neural Networks, 1, No. 1, 65–93 (1988).
K. Kawamura, Y. Tsujimoto, and M. Rabenarivo, Remote Sens., 9, No. 10, 1081 (2017).
M. Tatzber, F. Mutsch, and A. Mentler, Appl. Spectrosc., 64, No. 10, 1167–1175 (2010).
Cheng Wen Chang, D. A. Laird, and M. J. Mausbach, Soil Sci. Soc. Am. J., 65, No. 2, 480–490 (2001).
S. Wold, M. Sjöström, and L. Eriksson, Chemom. Intell. Lab. Syst., 58, No. 2, 109–130 (2001).
R. M. Balabin and E. I. Lomakina, Analyst, 136, No. 8, 1703–1712 (2011).
D. K. Ghose, S. S. Panda, and P. C. Swain, J. Hydrol., 394, Nos. 3–4, 296–304 (2010).
R. K. Douglas, S. Nawar, and M. C. Alamar, Sci. Total Environ., 616–617, 147–155 (2017).
D. Summers, M. Lewis, and B. Ostendorf, Ecol. Indic., 11, No. 1, 123–131 (2011).
L. J. Janik, S. T. Forrester, and A. Rawson, Chemom. Intell. Lab. Syst., 97, No. 2, 179–188 (2009).
H. Wang and D. Hu, Brain Neural Networks, 1, 279–283 (2006).
Zhi-biao, Zhao, and Yang, J. Bionic. Eng., 5, No. 3, 253–257 (2008).
Q. Zhang, Q. Li, and G. Zhang, Anal. Methods, 4, No. 7, 2039–2047 (2012).
Author information
Authors and Affiliations
Corresponding author
Additional information
Abstract of article is published in Zhurnal Prikladnoi Spektroskopii, Vol. 86, No. 4, p. 673, July–August, 2019.
Rights and permissions
About this article
Cite this article
Li, XY., Fan, PP., Liu, Y. et al. Prediction Results of Different Modeling Methods in Soil Nutrient Concentrations Based on Spectral Technology. J Appl Spectrosc 86, 765–770 (2019). https://doi.org/10.1007/s10812-019-00891-5
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10812-019-00891-5