Application of soil properties, auxiliary parameters, and their combination for prediction of soil classes using decision tree model

Document Type : Research Paper


1 Soil Science Department, Faculty of Agriculture, Vali-e-Asr University of Rafsanjan, Rafsanjan, Iran

2 Soil Science Department, Faculty of Agriculture, Shahrekord University, Shahrekord, Iran


Soil classification systems are very useful for a simple and fast summarization of soil properties. These systems indicate the method for data summarization and facilitate connections among researchers, engineers, and other users. One of the practical systems for soil classification is Soil Taxonomy (ST). As determining  soil classes for an  entire area is expensive, time-consuming, and almost impossible, this research has tried to predict the soil classes in each level of the ST system (up to family level) by using the data of 120 excavated pedons and some auxiliary parameters (such as derivatives of digital elevation model, i.e., DEM) in Shahrekord plain, central Iran. For this reason, the decision tree model was encoded and implemented in the MATLAB software for three conditions: use of soil properties, auxiliary parameters, and its combination. According to the results, soil class prediction error by using soil properties, auxiliary parameters, and its combination was estimated to be 0, 3.33 and 0% for order and suborder levels; 0.83, 15 and 0.83% for great group level; 3.33, 22.5 and 3.33% for subgroup level and 30, 52.5 and 30% for family level, respectively. In addition, the use of kriging maps of soil properties (instead of 120 observational points) decreased the prediction error of the modeling in all levels of the ST system. It seems that the effect of auxiliary parameters (in comparison to soil properties) is not very significant for predicting soil classes in low-relief areas. 


Main Subjects

Adhikari, K., A.E. Hartemink, 2016. Linking soils to
     ecosystem services - A global review. Geoderma, 262;
Bagheri Bodaghabadi, M., M.H. Salehi, J.A. Martínez-
     Casasnovas, J. Mohammadi, N. Toomanian, I.
     Esfandiarpoor Borujeni, 2011. Using Canonical
     Correspondence Analysis (CCA) to identify the most
     important DEM attributes for digital soil mapping
     applications. Catena, 86; 66–74.
Bagheri Bodaghabadi, M., J.A. Martínez-Casasnovas,
     M.H. Salehi, J. Mohammadi, I. Esfandiarpoor
     Borujeni, N. Toomanian, A. Gandomkar, 2015. Digital
     Soil Mapping using Artificial Neuronal Networks
     (ANN) and Terrain-Modelling Attributes. Pedosphere,
     25; 580-591.
Bockheim, J.G., A.E. Hartemink, 2013. Distribution and
     classification of soils with clay-enriched horizons in
     the USA. Geoderma, 209–210; 153–160.
Brungard, C.W., J.L. Boettinger, M.C. Duniway, S.A.
     Wills, T.C. Edwards, 2015. Machine learning for
     predicting soil classes in three semi-arid landscapes.
     Geoderma, 239-240; 68–83.
Das, M.D., 2009. Principles of Geotechnical Engineering
     (7th ed.), Cengage Learning, Stamford, CT.
Elliott, P.E., P.J. Drohan, 2009. Clay accumulation and
     argillic-horizon development as influenced by aeolian
     deposition vs. local parent material on quartzite and
     limestone-derived alluvial fans. Geoderma, 151; 98–
Esfandiarpoor Borujeni I., J. Mohammadi, M.H. Salehi,
     N. Toomanian, R.M. Poch, 2010. Assessing
     geopedological soil mapping approach by statistical
     and geostatistical methods: A case study in the
     Borujen region, Central Iran. Catena, 82; 1–14.
Geological survey and mineral exploration of Iran. 2017.
Goodman, J. M., P. R. Owens, 2012. Predicting soil
     organic carbon using mixed conceptual and
     geostatistical models. In: B. Minasny, B. P.Malone, A.
     B. McBratney (eds), Digital soil assessments and
     beyond (pp. 155–159). London: CRC Press.
Gunal, H., M.D. Ransom, 2006. Clay illuviation and
     calcium carbonate accumulation along a precipitation
     gradient in Kansas. Catena, 68; 59–69.
Heung, B., H.C. Ho, J. Zhang, A. Knudby, C.E. Bulmer,
     M.G. Schmidt, 2016. An overview and comparison of
     machine-learning techniques for classification
     purposes in digital soil mapping. Geoderma, 265; 62–
Holmes, K.W., E.A. Griffin, N.P. Odgers, 2015. Large-
     area spatial disaggregation of a mosaic of conventional
     soil maps: evaluation over Western Australia. Soil
     Research, 53; 865–880.
Jafari, A., P.A. Finke, J. VandeWauw, S. Ayoubi, H.
     Khademi, 2012. Spatial prediction of USDA- great soil
     groups in the arid Zarand region, Iran: comparing
     logistic regression approaches to predict diagnostic
     horizons and soil types. European Journal of Soil
     Science, 63; 284–298.
Kalavathi, K., P.V. Nimitha Safar, 2015. Performance
     Comparison between Naive Bayes, Decision Tree and
     k-Nearest Neighbor. International Journal of
     Emerging Research in Management and Technology,
     4; 152-161.
Khormali, F., S. Ghergherechi, M. Kehl, S. Ayoubi, 2012.
     Soil formation in loess-derived soils along a subhumid
     to humid climate gradient, Northeastern Iran.
     Geoderma, 179–180; 113–122.
Lagacherie, P., S. Holmes, 1997. Addressing
     geographical data errors in a classification tree for soil
     unit prediction. International Journal of Geographical
     Information Science, 11; 183–198.
Machado, I.R., E. Giasson, A.R. Campos, J. Janderson, F.
     Costa, E.B. Silva, B.R. Bonfatti, 2018. Spatial
     Disaggregation of Multi-Component Soil Map Units
     Using Legacy Data and a Tree-Based Algorithm in
     Southern Brazil. Rev Bras Cienc Solo; 42: e0170193.
Massawe, B.H.J., S.K. Subburayalu, A.K. Kaaya, L.
     Winowiecki, B.K. Slater, 2018. Mapping numerically
     classified soil taxa in Kilombero Valley, Tanzania
     using machine learning. Geoderma, 311; 143–148.
Maynard, J.J., M.G. Johnson, 2014. Scale-dependency of
     LiDAR derived terrain attributes in quantitative soil-
     landscape modeling: Effects of grid resolution vs.
     neighborhood extent. Geoderma, 230–231; 29–40.
McBratney, A. B., M. L. Mendonc, B. Minasny, 2003. On
     digital soil mapping. Geoderma, 117; 3–52.
Mirakzehi, K., M. Pahlavan-Rad, A. Shahriari, 2018.
     Digital soil mapping of deltaic soils: A   case of study
     from Hirmand (Helmand) river delta. Geoderma, 313;
Mohammadi, M., 1986. Semi-detailed soil studies report
     Chaharmahal-Va-Bakhtiari province (Shahrekord and
     Borujen area). Tehran, Iran. Iranian Soil and Water
     Research Institute.
Moore, ID., R.B. Grayson, A.R. Ladson, 1991. Digital
     terrain modelling: a review of hydrological.
     geomorphological and biological applications. Hydrol
     Process, 5; 3-30.
Mosleh, Z., M.H. Salehi, A. Jafari, I. Esfandiarpoor
     Borujeni, A. Mehnatkesh, 2017. Identifying sources of
     soil classes variations with digital soil mapping
     approaches in the Shahrekord plain, Iran. Environ
     Earth Sci, 76; 748p.
Odgers, N.P., W. Sun, A.B. McBratney, B. Minasny, D.
     Clifford, 2014. Disaggregating and harmonising soil
     map units through resampled classification trees.
     Geoderma, 214; 91–100.
Olaya, V. F., 2004. A gentle introduction to SAGA GIS.
     User Manual. Germany, DC; Gottingen.
Rossiter, D. G., 2000. Methodology for soil resource
     inventories. Lecture notes. 2nd revised version.
     Enschede, The Netherlands: Soil Science Division,
     International Institute for Aerospace Survey and Earth
     Science (ITC).
Rouse, J. W., R. H. Hass, J. A. Schell, D.W. Deering,
     1974. Monitoring vegetation systems in the Great
     Plains with ERTS. Proceedings of 3rd Earth Resource
     Technology Satellite (ERTS) Symposium, 1; 48-62.
Saunders, A. M., J. L. Boettinger, 2007. Incorporating
     classification trees into a pedogenic understanding
     raster classification methodology, Green River Basin,
     Wyoming, USA. In: P. Lagacherie McBratney A. B.,
     Voltz M. (ed.), Digital Soil Mapping: An introductory
     perspective. Developments in Soil Science. Elsevier,
     Amsterdam, 31; 389-399.
Scull, P., J. Franklin, O.A. Chadwick, 2005. The
     application of classification tree analysis to soil type
     prediction in a desert landscape. Ecological
     Modelling, 181; 1–15.
Soil Survey Division Staff, 1993. Soil Survey Manual.
     Soil Conservation Service, U.S. Department of
     Agriculture Handbook 18 (Chapter 3).
Soil Survey Staff, 2014. Soil taxonomy: a basic systems
     of soil classification for making and interpreting soil
     surveys (12th ed.). USDA; NRCS.
Taghizadeh-Mehrjardi, R., B. Minasny, J. Triantafilis, F.
     Sarmadian, M. Omid, 2014. Digital mapping of soil
     classes using decision tree and auxiliary data in
     theArdakan region, Iran. Arid Land Research and
     Management, 42; 225-237.
Taghizadeh-Mehrjardi, R., K. Nabiollahi, B. Minasny, J.
     Triantafilis, 2015. Comparing data mining classifiers
     to predict spatial distribution of USDA-family soil
     groups in Baneh region. Iran. Geoderma, 253–254;
Thompson, J.A., J.C. Bell, C.A. Butler, 2001. Digital
     elevation model resolution: effects on terrain attribute
     calculation and quantitative soil-landscape modeling.
     Geoderma, 100; 67–89.
US Geology Survey, 2016. images- from-
     USGS-2. (
Wilson, J.P. 2012. Digital terrain modeling.
     Geomorphology, 137; 107–121.
Wu, W., A.D. Li, X.H. He, R. Ma, H.B. Liu, J.K. Lv,
     2018. A comparison of support vector machines,
     artificial neural network and classification tree for
     identifying soil texture classes in southwest China.
     Computers and Electronics in Agriculture, 144; 86–93.
Zinck, J. A., 1989. Physiography and soils (Lecture notes
     for soil students. Soil Science Division, Soil survey
     courses subject matter, K6). Enschede, The
     Netherlands: ITC