Application of Random Forest method for predicting soil classes in low relief lands (case study: Hirmand county)

Document Type : Complete scientific research article

Authors

Abstract

Abstract
Background and Objectives: Base of soil information for environmental modeling is soil survey and mapping as a way to determine soil distribution patterns, describe and display it to understood and interpreted for different users. Digital soil mapping creates link between classes or soil characteristics and environmental factors affected soil formation and development by using mathematical models which can provide more precise and accurate soil maps and reducing cost of soil survey and mapping projects. This study was done to mapping soil great groups and subgroups by using Random Forest technique in the Hirmand county lands in Sistan plain.
Materials and Methods: In this study 108 soil profiles were dug on about 60.000 hectares of Hirmand county lands. Sixteen environmental variables were used as estimator for soil mapping including land properties, salinity and vegetation index. After classification of soil profiles to great groups and subgroups, soil classes map provided by using random forest (RF) method. It should be mentioned 80 percent of data was used for model training and 20 percent for independent validation.
Results: Pedological studies showed soils that formed in Sistan plain haven’t high development and most of them are Entisol and Aridisol. Most soil profiles classified as Torrifluvents on great groups and Typic Torrifluvents as subgroups. Also the result of RF showed the lowest estimation error of out of bag (OOB) samples in soil great groups and subgroups were 43.53 and 50.59 respectively. Independent validation results showed the best accuracy obtained for soil great groups and subgroups were 48 and 53 percent respectively. Grooves depth, convergence index, channel network between and salinity in soil great groups and grooves depth, elevation and catchment area in soil subgroups were the most important environmental variables to estimate soil classes.
Conclusion: The results showed most soils are young in the low relief lands in arid regions and these regions have also low soil diversity. Soil digital mapping and random forest technique could be useful for soil classes prediction and soil mapping in this kind of regions.
Results: Pedological studies showed soils that formed in Sistan plain haven’t high development and most of them are Entisol and Aridisol. Most soil profiles classified as Torrifluvents on great groups and Typic Torrifluvents as subgroups. Also the result of RF showed the lowest estimation error of out of bag (OOB) samples in soil great groups and subgroups were 43.53 and 50.59 respectively. Independent validation results showed the best accuracy obtained for soil great groups and subgroups were 48 and 53 percent respectively. Grooves depth, convergence index, channel network between and salinity in soil great groups and grooves depth, elevation and catchment area in soil subgroups were the most important environmental variables to estimate soil classes.
Conclusion: The results showed most soils are young in the low relief lands in arid regions and these regions have also low soil diversity. Soil digital mapping and random forest technique could be useful for soil classes prediction and soil mapping in this kind of regions.
Keywords: Soil digital mapping, Random forest technique, Map accuracy, Arid regions, Sistan plain

Keywords


ebkit-text-size-adjust: auto; -webkit-1.Al-Masrahy, M.A., and Mountney, N.P. 2015. A classification scheme for fluvial–aeolian
system interaction in desert-margin settings. Aeolian Research. 17: 67-88.
2.Barthold, F.K., Wiesmeier, M., Breuer, L., Frede, H.G., Wu, J., and Blank, F.B. 2013. Land
use and climate control the spatial distribution of soil types in the grasslands of Inner
Mongolia. J. Arid Environ. 88: 194-205.
3.Behrens, T., Förster, H., Scholten, T., Steinrücken, U., Spies, E.D., and Goldschmitt, M. 2005.
Digital soil mapping using artificial neural networks. J. Plant Nutr. Soil Sci. 168: 21-33.
4.Behrens, T., Schmidt, K., Zhu, A.X., and Scholten, T. 2010. The ConMap approach for
terrainbased digital soil mapping. Eur. J. Soil Sci. 61: 133-143.
5.Boer, M., DelBarrio, G., and Puigdefabregas, J. 1996. Mapping soil depth classes in dry
Mediterranean areas using terrain attributes derived from a digital elevationmodel.
Geoderma. 72: 99-118.
6.Breiman, L., and Cutler, A. 2004. Random Forests. Department of Statistics, University of
Berkeley. http://www.stat.berkeley.edu/~breiman/RandomForests/cc_home.htm.
7.Brungard, C.B., and Boettinger, J.L. 2012. Spatial prediction of biological soil crust classes;
value added DSM from soil survey. P 57-60, In: B. Minasny, B.P. Malone and A.
McBratney (Eds.), Digital Soil Assessments and Beyond Proceedings of the 5th
GlobalWorkshop on Digital Soil Mapping. CRC Press, Sydney.
8.Brungard, C.W. 2009. Alternative Sampling and Analysis Methods for Digital Soil Mapping
in Southwestern Utah. Thesis for Master of Science, Utah State University. USA.
9.Brungard, C.W., Boettinger, J.L., Duniway, M.C., Wills, S.A., and Edwards Jr., T.C. 2015.
Machine learning for predicting soil classes in three semi-arid landscapes. Geoderma.
239-240: 68-83.
10.Buol, S.W., Southard, R.J., Graham, R.C., and McDaniel, P.A. 2011. Soil genesis and
classification. 6th edition. Iowa State Univ. Press. Ames. Iowa, 556p.
11.Campling, P., Gobin, A., and Feyen, J. 2002. Logisticmodeling to spatially predict the
probability of soil drainage classes. Soil Sci. Soc. Am. J. 66: 1390-1401.
12.Cook, S.E., Jarvis, A., and Gonzalez, J.P. 2008. A New Global Demand for Digital Soil
Information. P 31-43, In: A.E. Hartemink, A. McBratney and M.L. Mendonca-Santos (Eds.),
Digital Soil Mapping with Limited Data. Springer, Dordrecht Heidelberg London New York.
13.Grunwald, S. 2010. Current State of Digital Soil Mapping and What Is Next. P 3-12, In: J.L.
Boettinger, D.W. Howel, A.C. Moore, A.E. Hartemink and S. Kienast-Brown (Eds.), Digital
Soil Mapping: Bridging Research, Environmental Application and Operation. Springer.
Dordrecht Heidelberg London New York.
14.Hastie, T., Tibshirani, R., and Friedman, J.H. 2001. The Elements of Statistical Learning:
Data Mining, Inference and Prediction. Springer, New York.
15.Hengl, T., and Reuter, H.I. 2008. Geomorphometry. Concepts, Software, Applications.
Developments in Soil Science. Elsevier, Amsterdam.
16.Hengl, T., Toomanian, N., Reuter, H.I., and Malakouti, M.J. 2007. Methods to interpolate soil
categorical variables from profile observations: lessons from Iran. Geoderma. 140: 417-427.
17.Heung, B., Bulmer, C.E., and Schmidt, M.G. 2014. Predictive soil parent material mapping
at a regional–Scale: A random forest approach. Geoderma. 214-215: 141-154.
18.Jafari, A., Ayoubi, S., Khademi, H., Finke, P.A., and Toomanian, N. 2013. Selection of a
taxonomic level for soil mapping using diversity and map purity indices: a case study from
an Iranian arid region. Geomorphology. 201: 86-97.
19.Jafari, A., Finke, P.A., Van deWauw, J., Ayoubi, S., and Khademi, H. 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. Eur. J. Soil Sci. 63: 284-298.
20.Jenny, H. 1941. Factors of Soil Formation: A System of Quantitative Pedology. McGrawHill, New York.
21.Lieb, M., Glaser, B., and Huwe, B. 2012. Uncertainty in the spatial prediction of soil texture
comparison of regression tree and random forest models. Geoderma. 170: 70-79.
22.Liu, J., Pattey, E., Nolin, M.C., Miller, J.R., and Ka, O. 2008. Mapping within-field soil
drainage using remote sensing, DEM and apparent soil electrical conductivity. Geoderma.
143: 261-272.
23.McBratney, A.B., Mendonça Santos, M.L., and Minasny, B. 2003. On digital soil mapping.
Geoderma. 117: 1-2. 3-52.
24.Minasny, B., McBratney, A.B., and Hartemink, A.E. 2010. Global pedodiversity, taxonomic
distance and the World Reference Base. Geoderma. 155: 132-139.
25.Moonjun, R., Farshad, A., Shrestha, D.P., and Vaiphasa, C. 2010. Artificial neural network
and decision tree in predictive soil mapping of Hoi Num Rin sub-watershed, Thailand.
P 151-164, In: J.L. Boettinger, D.W. Howell, A.C. Moore, A.E. Hartemink and S.
Kienast-Brown (Eds.), Digital Soil Mapping: Bridging Research, Environmental Application
and Operation. Springer, Dordrecht.
26.National soil survey center. 2012. Field book for describing and sampling soils, Ver. 3. U.S.
department of agriculture, Natural resources conservation service.
27.Pahlavan Rad, M.R., Toomanian, N., Khormali, F., Brungard, C.W., Komaki, C.B., and
Bogaert, P. 2014. Updating soil survey maps using random forest and conditional latin
hypercube sampling in the loess soil of northern Iran. Geoderma. 232-234: 97-106.
28.Pahlavan Rad, M.R., Toomanian, N., Khormali, F., Brungard, C.W., Komaki, C.B., and
Bogaert, P. 2016. Legacy soil maps as a covariate in digital soil mapping: A case study from
northern Iran. Geoderma. 279: 141-148.
29.Pahlavan Rad, M.R. 2014. Mapping and Updating Soil Map Using Random Forest and
Multinomial Logistic Regression in Golestan Province. Phd Thesis, Gorgan University of
Agricultural Sciences and Natural Resources, 114p.
30.Pahlavan Rad, M.R., Toomanian, N., Khormali, F., Brungard, C.W., Komaki, C.B., and
Bogaert, P. 2014. Digital soil mapping using random decision tree models in Golestan
province. J. Water Soil Cons. 21: 6. 73-93. (In Persian)
31.Poggio, L., Gimona, A., and Brewer, M.J. 2013. Regional scale mapping of soil properties
and their uncertainty with a large number of satellite-derived covariates. Geoderma.
209-210: 1–14.
32.Roecker, S.M., Howell, D.W., Haydu-Houdeshell, C.A., and Blinn, C. 2010. A Qualitative
Comparison of Conventional SoilSurvey and Digital Soil Mapping Approaches. P 369-384,
In: J.L. Boettinger, D.W. Howell, A.C. Moore, E.A. Hartemink and S. Kienast-Brown
(Eds.), Digital Soil Mapping: Bridging Research, Environmental Application and Operation.
Progress in Soil Science. Springer, New York.
33.Schaetzl, R.J., and Anderson, S. 2005. Soils: Genesis and Geomorphology. Cambridge
University Press, 833p.
34.Soil Survey Staff. 2014. Keys to soil Taxonomy, 12th ed. U.S. department of agriculture,
Natural resources conservation service.
35.Stum, A.K., Boettinger, J.L., White, M.A., and Ramsey, R.D. 2010. Random Forests applied
as a soil spatialpredictive model in arid Utah. P 179-189, In: J.L. Boettinger, D. Howell,
A.C. Moore, A. Hartemink and E.S. Kienast-Brown (Eds.), Digital SoilMapping:Bridging
Research, Environmental Application and Operation. Progress in Soil Science. Springer,
Logan, USA.
36.Taghizadeh-Mehrjardi, R., Minasny, B., Sarmadian, F., and Malone, B.P. 2014. Digital
mapping of soil salinity in Ardakan region, central Iran. Geoderma. 213: 15-28.
37.Were, K., Bui, D.T., Disk, B., and Singl, B.R. 2015. A comparative assessment of support
vector regression, artificial neural networks and random forest for predicting soil organic
carbon stocks across an afromonkane land scape. Ecological indicator. Pp: 394-403.
38.Wilson, J.P., and Gallant, J.C. 2000. Terrain Analysis: Principles and Applications. In: G.J.
Wilson JP (Ed.), Digital terrain analysis. John Wiley, New York, 478p.
39.Xiong, X., Grunwald, S., Myers, D.B., Kim, J., Harris, W.G., and Comerford, N.B. 2012.
Which soil, environmental and anthropogenic covariates for soil carbon models in Florida
are needed? P 335-339, In: B. Minasny, B.P. Malone and A. McBratney (Eds.), Digital Soil
Assessments and Beyond: Proceedings of the 5th Global Workshop on Digital SoilMapping.
CRC Press, Sydney.
40.Yokoyama, R., Shirasawa, M., and Pike, R.J. 2002. Visualizing topography by openness: a
new application of image processing to digital elevation models. Photogramm. Eng. Remote
Sens. 68: 257-266.