Modeling the Vertical Soil Calcium Carbonate Equivalent Variation by Machine Learning Algorithms in Qazvin Plain

آب و خاک

Article 8, Volume 35, Issue 5 - Serial Number 79, November and December 2021, Pages 719-734   

Document Type: Research Article

DOI: 10.22067/jsw.2021.71748.1076

Authors

S.R. Mousavi¹; F. Sarmadian^{* 2}; M. Omid³; P. Bogaert⁴

¹Ph.D. Student of Soil Resources Management, Science and soil Engineering Department, Faculty of Agricultural Engineering and Technology, College of Agriculture and Natural Resources, University of Tehran, Karaj, Iran

²Professor of Soil and Science Engineering Department, Faculty of Agricultural Engineering and Technology, University College of Agriculture and Natural Resources, University of Tehran, Karaj, Iran

³Professor of Agricultural Machinery Engineering Department, Faculty of Agricultural Engineering and Technology, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, Iran.

⁴Professor of Earth and Life Institute, catholique de Louvain, Louvain-la-Neuve, Belgium.

Abstract

Introduction: Calcium Carbonate Equivalent (CCE) is one of the key soils properties in arid and semi-arid regions. The study of spatial variability of surface and subsurface layers is important in the sustainable land management of arable soils. This study aimed to model the spatial distribution of CCE percentage by using three machine learning algorithms including Random Forest (RF), Decision Tree regression (DTr) and k-Nearest Neighbor (k-NN) at five standard depths of 0-5, 5-15, 15-30, 30-60, and 60-100 cm.
Material and Methods: The study area with 60,000 ha includes the major part of the lands of Qazvin plain located on the border of Qazvin and Alborz provinces. Field and laboratory surveys included 278 representative profiles were excavated, described by the horizon, and determined physicochemical properties. The studied soils have a very high diversity in soil moisture (Aridic, Xeric, and Aquic) and temperature regimes (Thermic). These variations have led to the formation of eight great groups of soils in the region based in the USDA soil classification system with the three classes of Haploxerepts, Calcixerepts, and Haplocalcids were the dominant soil classes in the study area. A total of 22 environmental covariates, including 12 variables extracted from the primary and secondary derivation of digital elevation model (DEM), six remote sensing (RS) indicators, two climatic parameters, and two soil covariates were prepared, and then the most appropriate environmental covariates were selected using principal component analysis (PCA) and expert knowledge. The CCE percentage data were randomly divided into two parts, 80% for training and 20% for testing, which was then modeled by three machine learning algorithms RF, DTr, and k-NN, and were evaluated by some statistical indices as coefficient determination (R²), root mean square error (RMSE) and Bias.
Results and Discussion: The results of harmonizing the CCE values at the genetic horizons with the standard depths showed the high efficiency of the spline depth function in providing an acceptable estimate with minimum error and maximum agreement between observed and predicted values. The PCA method showed that the first to fifth components with the explanation of more than 80% of cumulative variance were Multi-Resolution Index of Valley Bottom Flatness (MrVBF), Mean Annual Temperature (MAT), Greenness index (Greenness), Probability of Calcic horizon (Cal.hr), and Wind Effect environmental covariates which had the highest eigenvalues. Besides, Clay was selected on expert knowledge-based. The relative importance (RI) of the environmental covariates showed the spatial distribution of CCE were affected by Clay with an explanation of more than 57%, 41.8% and 45% of its variance at three surface depths of 0-5, 5-15, and 15-30 cm, while the Cal.hr covariate had the highest impact in the spatial prediction of CCE compared to other predictors as auxiliary variables with 67.8% and 52.8% justification, respectively, at two depths of 30-60 and 60-100 cm. Hence, using the calcic horizon probability Map (Cal.hr) as a derivative soil factor made it possible to produce more appropriate final maps, while preventing the reduction of the accuracy of the modeling results in the subsoils. The auxiliary variable of remote sensing, i.e., Greenness, could not show a significant impact on the expression of the variation of CCE percentage at all studied depths. Unlike remote sensing indices, the topographic attribute of the MrVBF, at two standard depths of 0-5 and 5-15 cm, the MAT at a depth of 15-30 cm, and the Wind Effect at the standard depths 30-60 and 60-100 cm, after the soil covariates, were the most effective in justifying the spatial variations of CCE%. RF algorithm with a range of R² values of 0.83 - 0.76 and RMSE of 2.14% - 2.21% resulted in the highest accuracy and minimum error. Even though the DTr method presented R² values (0.52-0.39) weaker than the RF in the validation dataset, in general, the results of its spatial predictions were similar to the RF model from the surface to the subsurface and more stable than the k-NN. Against RF and DTr, k-NN couldn’t display acceptable performance in the prediction of CCE% at all standardized depths.
Conclusion: In general, it is necessary to understand the spatial distribution of CCE due to its effect on soil moisture accessibility and plant nutrient uptake. Therefore, in the present study, we tried to introduce the RF machine learning algorithm as a superior model with environmental variables that were selected by PCA and the expert knowledge variable selection method. The maps prepared by this approach have an acceptable level of reliability for agricultural and environmental management by managers, soil experts, and farmers.
 

Keywords

Digital soil mapping; Standard depth; Spline function; Soil forming factors

References

• Amirian C.A., Taghizadeh Mehrjardi R., Sarmadian F., and Mohammadi J. 2018. Study of lateral and vertical distribution of soil calcium carbonate using geostatistics and spline functions. (In Persian with English abstract)
• Arrouays D., Grundy M.G., Hartemink A.E., Hempel J.W., Heuvelink G.B., Hong S.Y., Lagacherie P., Lelyk G., McBratney A.B., McKenzie N.J., and dL Mendonca-Santos M. 2014. GlobalSoilMap: Toward a fine-resolution global grid of soil properties. Advances in Agronomy 125: 93-134.
• Asgari Hafshejani N., and Jafari S. 2017. The study of particle size distribution of calcium carbonate and its effects on some soil properties in Khuzestan province. Iran Agricultural Research 36(2): 71-80.
• Bouslihim Y., Rochdi A., and Paaza N.E.A. 2021. Machine learning approaches for the prediction of soil aggregate stability. Heliyon 7(3): e06480.
• Chakan A.A., Taghizadeh-Mehrjardi R., Kerry R., Kumar S., Khordehbin S., and Khanghah S.Y. 2017. Spatial 3D distribution of soil organic carbon under different land use types. Environmental Monitoring and Assessment 189(3): 131.
• Esfandiarpour Boroujeni I., ShahiniShamsabadi M., Shirani H., Mosleh Z., BagheriBodaghabadi M., and Salehi M.H. 2020. Assessment of different digital soil mapping methods for prediction of soil classes in the Shahrekord plain, Central Iran. Catena 193: 104648.
• Esmaeili E., Shahbazi F., Sarmadian F., Jafarzadeh A.A., and Hayati B. 2021. Land capability evaluation using NRCS agricultural land evaluation and site assessment (LESA) system in a semi-arid region of Iran. Environmental Earth Sciences 80(4): 1-14.
• 1973. Irrigation, Drainage and salinity. FAO/UNESCO.
• Hengl T., Mendes de Jesus J., Heuvelink G.B., Ruiperez Gonzalez M., Kilibarda M., Blagotić A., Shangguan W., Wright M.N., Geng X., Bauer-Marschallinger B., and Guevara M.A. 2017. SoilGrids250m: Global gridded soil information based on machine learning. PLoS one 12(2): e0169748.
• Hengl T., Miller M.A., Krizan J., Shepherd K.D., Sila A., Kilibarda M., Antonijevic O., Glušica L., Dobermann A., Haefele S.M., and McGrath S.P. 2021. African soil properties and nutrients mapped at 30 m spatial resolution using two-scale ensemble machine learning. Scientific Reports 11(1): 1-18.
• Keshavarzi A., Sarmadian F., Labbafi R., and Ahmadi A. 2011. Developing pedotransfer functions for estimating field capacity and permanent wilting point using fuzzy table look-up scheme. Computer and Information Science 4(1): 130.
• Khodaverdiloo H., Homaee M., van Genuchten M.T., and Dashtaki S.G. 2011. Deriving and validating pedotransfer functions for some calcareous soils. Journal of Hydrology 399(1-2): 93-99.
• Khaledian Y., and Miller B.A. 2020. Selecting appropriate machine learning methods for digital soil mapping. Applied Mathematical Modelling 81: 401-418.
• Kuhn M., and Johnson K. 2013. Applied predictive modeling (Vol. 26, p. 13). New York: Springer.
• Lacoste M., Minasny B., McBratney A., Michot D., Viaud V., and Walter C. 2014. High resolution 3D mapping of soil organic carbon in a heterogeneous agricultural landscape. Geoderma 213: 296-311.
• McBratney A.B., Santos M.M., and Minasny B. 2003. On digital soil mapping. Geoderma 117(1-2): 3-52.
• Mahmoudabadi E., Karimi A., Haghnia G.H., and Sepehr A. 2017. Digital soil mapping using remote sensing indices, terrain attributes, and vegetation features in the rangelands of northeastern Iran. Environmental Monitoring and Assessment 189(10): 1-20.
• Malone B.P., McBratney A.B., Minasny B., and Laslett G.M. 2009. Mapping continuous depth functions of soil carbon storage and available water capacity. Geoderma 154(1-2): 138-152.
• McDonald R.C., Isbell R.F., Speight J.G., Walker J., and Hopkins M.S. 1998. Australian soil and land survey: field handbook (No. Ed. 2). CSIRO publishing.
• Mosleh Z., Salehi M.H., Jafari A., Borujeni I.E., and Mehnatkesh A. 2016. The effectiveness of digital soil mapping to predict soil properties over low-relief areas. Environmental Monitoring and Assessment 188(3): 195.
• Mousavi S.R., Parsayi F., Rahmani A., Sedri, M.H., and Kohsar Bostani M. 2020. Spatial Prediction Some of the Surface Soil Properties Using Interpolation and Machine Learning Models. Journal of Soil Management and Sustainable Production 10(3): 27-49. (In Persian with English abstract).
• Mousavi S.R., Sarmadian F., Dehghani S., Sadikhani M.R., and Taati A. 2017. Evaluating inverse distance weighting and kriging methods in estimation of some physical and chemical properties of soil in Qazvin Plain. Eurasian Journal of Soil Science 6(4): 327-336.
• Mousavi S.R, Sarmadian F., Omid M., and Bogaert P. 2021. Digital modeling of three-dimensional soil salinity variation using machine learning algorithms in arid and semi-arid land of Qazvin plain. Iranian Journal of Soil and Water Research, doi: 10.22059/ijswr.2021.323030.668957. (In Persian with English abstract)
• Mulder V.L., Lacoste M., Richer-de-Forges A.C., Martin M.P., and Arrouays D. 2016. National versus global modelling the 3D distribution of soil organic carbon in mainland France. Geoderma 263: 16-34.
• Nelson R.E. 1982 Carbonate and gypsum. In: Page AL (ed) Methods of soil analysis. American Society of Agronomy, Madison, pp 181–197.
• Nemes A., Rawls W.J., and Pachepsky Y.A. 2006. Use of the nonparametric nearest neighbor approach to estimate soil hydraulic properties. Soil Science Society of America Journal 70(2): 327-336.
• Pahlavan-Rad M.R., and Akbarimoghaddam A. 2018. Spatial variability of soil texture fractions and pH in a flood plain (case study from eastern Iran). Catena 160: 275-281.
• Padarian J., Minasny B., and McBratney A.B. 2019. Using deep learning for digital soil Soil 5: 79–89.
• Parsaie F., Firouzi A.F., Mousavi S.R., Rahmani A., Sedri M.H., and Homaee M. 2021. Large-scale digital mapping of topsoil total nitrogen using machine learning models and associated uncertainty map. Environmental Monitoring and Assessment 193(4): 1-15.
• Presley D.R., Ransom M.D., Kluitenberg G.J., and Finnell P.R. 2004. Effects of thirty years of irrigation on the genesis and morphology of two semiarid soils in Kansas.
• Rahmani A., Sarmadian F., Mousavi S.R., and Khamoshi S.E. 2020. Application of Geomorphometric attributes in digital soil mapping by using of machine learning and fuzzy logic approaches. Journal of Range and Watershed Managment 73(1): 105-124. (In Persian)
• Rezapour S. 2014. Response of some soil attributes to different land use types in calcareous soils with Mediterranean type climate in north-west of Iran. Environmental Earth Sciences 71(5): 2199-2210.
• Rossel R.V., Chen C., Grundy M.J., Searle R., Clifford D., and Campbell P.H. 2015. The Australian three-dimensional soil grid: Australia’s contribution to the GlobalSoilMap project. Soil Research 53(8): 845-864.
• Rostaminia M., Nouri N., Keshavarzi A., and Rahmani A. 2019. Quantitative Evaluation and Zoning of Spatial Distribution of Soil Quality Index in Some Parts of Arid and Semi-Arid Lands of Western Iran (Case Study: Kane Sorkh Region, Ilam Province). Iranian Journal of Soil and Water Research 50(7): 1701-1719. (In Persian with English abstract)
• Sreenivas K., Dadhwal V.K., Kumar S., Harsha G.S., Mitran T., Sujatha G., Suresh G.J.R., Fyzee M.A., and Ravisankar T. 2016. Digital mapping of soil organic and inorganic carbon status in India. Geoderma 269: 160-173.
• Staff S.S. 2014. Keys to Soil Taxonomy, 12th Edn Washington. DC: Natural Resources Conservation Service, United States Department of Agriculture.
• Taghizadeh Mehrjardi R., Nabiollahi K., and Kerry R. 2016. Digital mapping of soil organic carbon at multiple depths using different data mining techniques in Baneh region, Iran. Geoderma 266: 98-110.
• Taghizadeh Mehrjardi R., Minasny B., Sarmadian F., and Malone P.B. 2014a. Digital mapping of soil salinity in Ardakan region, central Iran. Geoderma 213: 15-28.
• Taghizadeh-Mehrjerdi R., Amirin Chakan A., and Sarmadian F. 2014b. 3D digital mapping of soil cation exchange capacity in Dorud, Lorestan province. Journal of Water and Soil 28: 998-1010. (In Persian with English abstract)
• Tan W.F., Zhang R., Cao H., Huang C.Q., Yang Q.K., Wang M.K., and Koopal L.K. 2014. Soil inorganic carbon stock under different soil types and land uses on the Loess Plateau region of China. Catena 121: 22-30.
• Vargas R., Pankova E.I., Balyuk S.A., Krasilnikov P.V., and Khasankhanova G.M. 2018. Handbook for saline soil management. FAO/LMSU.
• Viscarra Rossel R.A., and McBratney A.B. 2008. Diffuse reflectance spectroscopy as a tool for digital soil mapping. In ‘Digital soil mapping with limited data’. Developments in Soil Science series. (Eds AE Hartemink, AB McBratney, L Mendonça-Santos) (Elsevier Science: Amsterdam).
• Wang Y., and Witten I.H. 1997. Inducing model trees for continuous classes. In Proceedings of the Ninth European Conference on Machine Learning. pp. 128–137.
• Wilding L.P. 1985. Spatial variability: its documentation, accomodation and implication to soil surveys. In Soil spatial variability, Las Vegas NV, 30 November-1 December 1984 (pp. 166-194).
• Zeraatpisheh M., Ayoubi S., Jafari A., Tajik S., and Finke P. 2019. Digital mapping of soil properties using multiple machine learning in a semi-arid region, central Iran. Geoderma 338: 445-452.
• Zhao W., Zhang R., Huang C., Wang B., Cao H., Koopal L.K., and Tan W. 2016. Effect of different vegetation cover on the vertical distribution of soil organic and inorganic carbon in the Zhifanggou Watershed on the loess plateau. Catena 139: 191-198.