Spatial Prediction of Soil Saturated Hydraulic Conductivity by Integrating Soil Properties and Environmental Covariates

Document Type : Research Paper

Authors

1 Department of Soil Science, Faculty of Agriculture, University of Kurdistan, Sanandaj, Iran

2 Department of Soil Sciences, Faculty of Agriculture, University of Kurdistan, Sanandaj, Iran

Abstract

Soil hydraulic properties, particularly saturated hydraulic conductivity (Ks), play a crucial role in addressing problems related to soil and water management in agricultural, ecological, and environmental systems. Direct measurement of these properties is often difficult, costly, and time-consuming; hence, indirect estimation methods are commonly employed. In this study, the efficiency of multiple linear regression (MLR), decision tree (DT), and artificial neural network (ANN) methods was evaluated for estimating and mapping the spatial distribution of Ks in parts of the Cherdawel and Chamshir sub-basins (Ilam Province, Iran), using readily measurable soil properties along with environmental covariates (terrain attributes and remote sensing data). For this purpose, Ks was measured at 95 sampling points using a Guelph permeameter, and several readily measurable soil properties along with environmental covariates were also obtained at the same locations. The validity of the derived models for Ks estimation was assessed using the coefficient of determination (R²), root mean square error (RMSE), concordance correlation coefficient (rc), and mean error (ME). The results demonstrated that the ANN model outperformed both MLR and DT models in estimating Ks. While the MLR and DT tended to underestimate Ks, the ANN model produced more accurate and reliable predictions. Furthermore, the spatial variability map of Ks could be successfully generated by integrating soil properties with environmental covariates, suggesting the usefulness of this approach for developing agro-hydrological models in data-limited regions. Overall, the findings suggest that incorporating environmental covariates alongside soil properties can significantly enhance the accuracy of Ks estimation, particularly when using the ANN model.

Keywords

Main Subjects

Introduction:

Soil and water are two vital natural resources on which human life strongly depends. Soil hydraulic properties, particularly saturated hydraulic conductivity (Ks), play a crucial role in addressing problems related to soil and water management in agricultural, ecological, and environmental systems. Ks directly affects key hydrological processes such as surface runoff, soil erosion, and deep percolation. However, soil and biotic factors influencing Ks vary greatly in both space and time. Hence, it is also expected that Ks will vary in both space and time. Moreover, Ks is one of the most difficult, costly, and time-consuming soil hydraulic properties to measure directly; Therefore, indirect estimation methods are commonly employed.

Terrains play a fundamental role in modulating the earth's surface and atmospheric processes. Landform features generally control the movement of water and materials across the landscape, thereby influencing watershed hydrology at the topographic scale. Using terrain attributes to model Ks may serve as a suitable alternative, as terrain data are relatively easy to obtain compared to intensive soil sampling. In previous soil and landscape studies, the relationship between terrain attributes, remote sensing data, and Ks has not been thoroughly investigated. The main question is whether integrating these data with soil properties can enhance the accuracy of Ks estimation. Accordingly, this study evaluated the efficiency of multiple linear regression (MLR), decision tree (DT), and artificial neural network (ANN) methods for estimating and mapping the spatial distribution of Ks in parts of the Chardawel and Chamshir sub‑basins (Ilam Province, Iran), using readily measurable soil properties along with environmental covariates (terrain attributes and remote sensing data).

Materials and Methods

The study area extends from 46° 25' 53.08" to 46° 50' 13.52" E longitude and from 33° 31' 23.11" to 33° 50' 25.59" N latitude, located in the Chardawel and Chamshir sub-basins, Ilam Province, Iran. A total of 95 sampling sites were selected using a stratified random sampling approach, ensuring representation of land capability classes, land use categories, and geological and topographical features. At each site, Ks was measured using a Guelph permeameter. Disturbed and undisturbed soil samples to a depth of 30 cm were collected from the same sites for laboratory analysis. All samples were air-dried, gently crushed, and sieved through a 2-mm opening. The samples were analyzed for particle size distribution (PSD), particle density (Dp), soil moisture at field capacity (FC), total porosity, effective porosity, and soil organic carbon (SOC). The geometric mean diameter (dg) and geometric standard deviation (σg) of soil particles were calculated based on the method proposed by Shirazi and Boersma (1984). Undisturbed soil samples were analyzed for bulk density (Db) using the cylindrical core method.

A set of 16 terrain features was extracted from a Digital Elevation Model (DEM) with 12.512.5 m spatial resolution (ALOS PALSAR database) using SAGA GIS software (Version 9.3.2). Additionally, Landsat 8 satellite imagery (OLI/TIRS sensors) with a 3030 m spatial resolution (sourced from the USGS) was employed to extract remote sensing variables. Principal Component Analysis (PCA) was performed using the XLSTAT toolbox in Microsoft Excel 2019 to identify and select the most influential variables. To estimate Ks values, MLR, DT, and ANN models were developed using R software (version 4.5.1). The validity of the derived models was assessed using the coefficient of determination (R²), root mean square error (RMSE), concordance correlation coefficient (rc), and mean error (ME).

Results and Discussion

Based on the PCA results, among the 40 soil properties and environmental covariates used for spatial prediction of Ks, 12 variables including percentage slope, topographic wetness index (TWI), LS factor, analytical hillshade, total catchment area, and channel network distance were selected. From the remote sensing indices and reflectance values, bands 4 and 5 of Landsat 8 OLI/TIRS (as the biophysical variables), along with silt percentage, sand percentage, d60, and soil particle density, were identified as the most effective covariates.

The results demonstrated that the ANN model, with a greater coefficient of determination (R2train= 0.80, R2test= 0.49), lower root mean square error (RMSEtrain= 0.008 m/d, RMSEtest= 0.25 m/d), and higher concordance correlation coefficient (rc-train= 0.89, (rc-test= 0.70), outperformed both the MLR and DT models in estimating Ks. While the MLR and DT models tended to underestimate Ks, the ANN model produced more accurate and reliable predictions. Overall, the ANN model yielded reasonable estimates of Ks for the study area's soils.

A spatial variability map of Ks was also successfully generated by integrating soil properties with environmental covariates, suggesting the usefulness of this approach for developing agro-hydrological models in data-limited regions.

Sensitivity analysis of the ANN model highlighted the importance of two soil properties, sand percentage and soil particle density, along with slope percentage, analytical hillshade, and topographic wetness index terrain variables in the spatial modeling and distribution of Ks in the studied lands.

Conclusion

Terrain and remote sensing data, along with soil properties, were integrated using stepwise multiple linear regression (SMLR), decision tree (DT), and artificial neural network (ANN) models to predict the spatial variability of soil saturated hydraulic conductivity (Ks) in parts of the Chardawel and Chamshir sub-basins, Ilam Province, Iran. Our findings indicated that incorporating environmental covariates with soil properties can significantly enhance the accuracy of Ks estimation, particularly when using the ANN model; thus, this model can be used to develop a digital Ks map with acceptable accuracy to support sustainable land planning.

Funding

This study was funded by the University of Kurdistan, Sanandaj, Iran.

Authorship contribution

Conceptualization, M. Davari, M.M. Mahmoodi and K. Nabiollahi; methodology, M. Davari, M.M. Mahmoodi and K. Nabiollahi; software, M. Davari and A. Hekmatzad; validation, M. Davari, A. Hekmatzad, M.M. Mahmoodi and K. Nabiollahi; formal analysis, M. Davari and A. Hekmatzad; investigation, M. Davari, A. Hekmatzad, M.M. Mahmoodi and K. Nabiollahi; resources, M. Davari and A. Hekmatzad; data curation, A. Hekmatzad; writing—original draft preparation, M. Davari; writing—review and editing, M. Davari; visualization, M. Davari and A. Hekmatzad; supervision, M. Davari, M.M. Mahmoodi and K. Nabiollahi; project administration, M. Davari, A. Hekmatzad; funding acquisition, M. Davari and A. Hekmatzad, All authors have read and agreed to the published version of the manuscript.

Declaration of Generative AI and AI-assisted technologies in the writing process:

The authors state that they have not used generative AI or AI-assisted technologies in their manuscript.

Data availability statement

All necessary information and data layers have been included in the manuscript. However, any further data will be available from the corresponding author upon reasonable request.

Acknowledgements

The authors express their sincere appreciation to the University of Kurdistan, Sanandaj, Iran, for providing laboratory facilities and supporting the conduct of this study.

Ethical considerations

The authors avoided data fabrication, falsification, plagiarism, and misconduct.

Conflict of interest

The authors declare no competing interests.

References

Agyare, W., Park, S., & Vlek, P. (2007). Artificial neural network estimation of saturated hydraulic conductivity. Vadose Zone Journal, 6, 423-431. https://doi.org/10.2136/vzj2006.0131.
Alletto, L., & Coquet, Y. (2009). Temporal and spatial variability of soil bulk density and near-saturated hydraulic conductivity under two contrasted tillage management systems. Geoderma, 152:85–94. https:// doi. org/ 10. 1016/j. geoderma. 2009. 05. 023
Cheng, Q., Xu, Q., Cheng, X., Li, C., & Wang, X. (2019). In‑situ estimation of unsaturated hydraulic conductivity in freezing soil using improved field data and inverse numerical modeling. Agricultural and Forest Meteorology, 279, 107746. https://doi.org/10.1016/j.agrformet.2019.107746.
Dane, J.H. & Topp, C.G., (eds.). 2002. Methods of Soil Analysis: Part 4 Physical Methods Madison, WI: Soil Science Society of America, Soil Science Society of America Book Series Number 5, ISBN 0-89118-810-X, 1692 p.
Dehghani, B.S., Ghorbani Dashtaki, Sh., Mohammadi, J. & Khodaverdilou, H. (2012). Comparing the performance of multiple linear regression and regression tree to predict saturated hydraulic conductivity and the inverse of macroscopic capillary length. Iranian Water Research Journal, (9): 193-203. (In Persian)
Farzadmehr, M., Dastourani, M., & Khashei-Siuki, A. (2019). Comparing decision tree and instance-based learning models to estimate soil saturated hydraulic conductivity. Journal of Water and Soil Conservation, 25(5), 167-184. http://doi.org/ 10.22069/jwsc.2018.12334.2691. (In Persian)
Fashi, F.H., Gorji, M., & Shorafa, M. (2016) Estimation of soil hydraulic parameters for different land-uses. Modeling Earth Systems and Environment ,2, 1–7. https://doi.org/10.1007/s40808-016-0229-0.
Fitts, C.R. (2002). Groundwater science. Academic Press, New York.
Gabriel, J.L., Quemada, M., Martín-Lammerding, D., & Vanclooster, M. (2019). Assessing the cover crop effect on soil hydraulic properties by inverse modelling in a 10-year field trial. Agricultural Water Management, 222:62–71. https:// doi. org/ 10. 1016/j. agwat. 2019. 05. 034
Gumiere, S. J., Lafond, J. A., Hallema, D. W., Périard, Y., & Caron, J. (2014). Mapping soil hydraulic conductivity and matric potential for water management of cranberry: Characterization and spatial interpolation methods. Biosystems Engineering, 128, 29–40. https://doi.org/10.1016/j.biosystemseng. 2014.09.002
Hekmatzad, A., Davari, M., Mahmoudi, M.A., and Nabiollahi, K. (2018). Estimation of saturated hydraulic conductivity in some soils of Ilam province using artificial neural networks and regression methods. Journal of Water and Soil Resources Conservation, 7(3), 131-149. (in Persian)
Igel, C., & Hüsken, M. (2003). Empirical evaluation of the improved Rprop learning algorithms. Neurocomputing, 50, 105-123. https://doi.org/10.1016/S0925-2312(01)00700-7.
Jarvis, N., Koestel, J., Messing, I., Moeys, J., & Lindahl, A. (2013). Influence of soil, land use and climatic factors on the hydraulic conductivity of soil. Hydrology and Earth System Sciences, 17, 5185–5195. https://doi.org/10.5194/hess-17-5185-2013.
Khaledian, Y., Kiani, F., Ebrahimi, S., Brevik, E.C., & Aitkenhead-Peterson, J. (2017). Assessment and monitoring of soil degradation during land use change using multivariate analysis. Land Degradation and Development, 28, 128–141. https://doi.org/10.1002/ldr.2541.  
Khodaverdiloo, H. Cheraghabdal, H.K., Bagarello, V., Iovino, M., Asgarzadeh, H., Ghorbani Dashtaki, Sh. (2017). Ring diameter effects on determination of field-saturated hydraulic conductivity of different loam soils, Geoderma, 303, 60-69. https://doi.org/10.1016/j.geoderma.2017.04.031.
Kirkham, M. B. (2005). Principles of Soil and Plant Water Relations. Elsevier Academic Press: Burlington, MA. pp. 145 ‐ 172.
Leij, F. J., Romano, N., Palladino, M., Schaap, M. G., & Coppola, A. (2004). Topographical attributes to predict soil hydraulic properties along a hillslope transect. Water Resources Research, 40, W02407. https://doi.org/10.1029/2002WR001641
McBratney, A. B., Odeh, I. O. A., Bishop, T. F. A., Dunbar, M. S., & Shatar, T. M. 2000. An overview of pedometric techniques for use in soil survey. Geoderma, 97, 293–327. https://doi.org/10.1016/S0016-7061(00)00043-4
McKenzie N.J., & Jacquier D.W. (1997). Improving the Field Estimation of Saturated Hydraulic Conductivity in Soil Survey. Australian Journal of Soil Research, 35, 803-825. https://doi.org/10.1071/S96093
Mesri, S., Ghorbani Dashtaki, Sh., Shirani, H., Kamkar Rohani, A., & Motaghian, H.R. (2020). Separating effective soil properties on moisture characteristic curve using Decision Tree. Journal of Soil Management and Sustainable Production, 10(3), 135-150. http://doi.org/ 10.22069/ejsms.2021.17673.1928. (In Persian).
Moncada, M.P., Gabriels, D., and Cornelis, W.M. (2014). Data-driven analysis of soil quality indicators using limited data. Geoderma, 235, 271-278. https://doi.org/10.1016/j.geoderma.2014.07.014.
Moore, I.D., Gessler, P.E., Nielsen, G.A.E., & Peterson, G.A. (1993) Soil attribute prediction using terrain analysis. Soil Science Society of America Journal, 57, 443–452. https://doi.org/10.2136/sssaj1993.0361599500 57000 20026x
Mosleh, Z., Salehi, M.H., Jafari, A., and Esfandiarpoor Borujeni, I., & Mehnatkesh, A. (2016). The effectiveness of digital soil mapping to predict soil properties over low-relief areas. Environmental Monitoring and Assessment, 188, 195. https://doi.org/10.1007/s10661-016-5204-8.
Motaghian, H.R., & Mohammadi, J. (2011) Spatial estimation of saturated hydraulic conductivity from terrain attributes using regression, kriging, and artificial neural networks. Pedosphere, 21, 170–177. https://doi.org/10. 1016/S1002-0160(11)60115-X.
Mustafa, S. M. T., Hasan, M. M., Saha, A. K., Ahmed, S., Kabir, M. R., & Rahman, M. M. (2019). Multi‑model approach to quantify groundwater‑level prediction uncertainty using an ensemble of global climate models and multiple abstraction scenarios. Hydrology and Earth System Sciences, 23, 2279–2303. https://doi.org/10.5194/ hess-23-2279-2019
Neuzil, C.E. (1994) How permeable are clays and shales? Water Resources Research, 30,145–150. https://doi.org/10.1029/93WR02930.
Norouzian Azizi, Z., Ghajar Sepanlou, M., Emadi, S.M., & Sadeqzade, F. (2017). Evaluation of regression and Artificial Neural Network models to estimate the saturated hydraulic conductivity in Mazandaran Province. Journal of Soil Research, 31(1), 75-87. http://doi.org/10.22092/ijsr.2017.110346. (In Persian).
Odeh, I. O. A., & McBratney, A. B. (2000). Using AVHRR images for spatial prediction of clay content in the lower Namoi Valley of eastern Australia. Geoderma, 97, 237–254. https://doi.org/10.1016/S0016-7061(00)00041-0.
Pan, Y., Wang, D., Lyu, Sh., Li, X., Li, S., & Su, Y. (2026). Evaluating saturated hydraulic conductivity schemes: impacts on soil moisture simulations and soil texture-dependent applicability. Journal of Hydrology, 671, 135129. https://doi.org/10.1016/j.jhydrol.2026.135129.
Park, S.J., & Vlek, P.L.G. 2002. Environmental correlation of three-dimensional spatial soil variability: A comparison of three adaptive techniques. Geoderma, 109, 117–140. https://doi.org/10.1016/S0016-7061(02)00146-5.
Qaderi, K., Jalali, V., Etminan, S., Masoumi Shahr-babak, M. & Homaee, M. (2018). Estimating soil hydraulic conductivity using different data-driven models of ANN, GMDH and GMDH-HS. Paddy and Water Environment, 16, 823–833. https://doi.org/10.1007/s10333-018-0672-9.
Rezaei, M., Mousavi, S. R., Rahmani, A., Zeraatpisheh, M., Rahmati, M., Pakparvar, M. & Cornelis, W. (2023). Incorporating machine learning models and remote sensing to assess the spatial distribution of saturated hydraulic conductivity in a light-textured soil. Computers and Electronics in Agriculture, 209, 107821. https://doi.org/10.1016/j.compag.2023.107821.
Romano, N., & Palladino, P. (2002). Prediction of soil water using soil physical data and terrain attributes. Journal of Hydrology, 265, 56–75. https://doi.org/10.1016/S0022-1694(02)00094-X.
Russo, S.L., Taddia, G., & Abdin, E.C. (2018) Modeling the effects of the variability of temperature-related dynamic viscosity on the thermal affected zone of groundwater heat-pump systems. Hydrogeology Journal, 26, 1239–1247. https:// doi. org/ 10. 1007/ s10040- 017- 1714-x.
Saha, A. K., Hasan, M. M., Khan, M. R. A., & Al‑Amin, M. (2021). Prediction of spatial saturated hydraulic conductivity at the upper soil layer using soil class and terrain attributes. Modeling Earth Systems and Environment. https://doi.org/10.1007/s40808-021-01317-y
Schaap, M.G., Leij, F.J., & van Genuchten, M.Th. (1999). A bootstrap-neural networks approach to predict soil hydraulic parameters. p. 1237–1250. In M.Th. van Genuchten et al. (ed.) Proc. Int. Workshop on characterization and measurement of the hydraulic properties of unsaturated porous media. Univ. of California, Riverside.
Shahabi, A., Nabiollahi, K., Davari, M., Zeraatpisheh, M., Heung, B., Scholten, T. and Taghizadeh-Mehrjardi, R. (2022). Spatial prediction of soil properties through hybridized random forest model and combination of reflectance spectroscopy and environmental covariates. Gecarto International, 37(27), 18172–18195. https://doi.org/10.1080/10106049.2022.2138565.
Shirazi, M.A., & Boersma, L. (1984). A unifying quantitative analysis of soil texture. Soil Science Society of America Journal, 48, 142-147. https://doi.org/10.2136/sssaj1984.03615995004800010026x
Sihag, P. (2018). Prediction of unsaturated hydraulic conductivity using fuzzy logic and artificial neural network. Modeling Earth Systems and Environment, 4, 189–198. https://doi.org/10.1007/s40808-018-0434-0
Souza, C.M.P.D., Thomazini, A., Schaefer, C.E.G.R., Veloso. G.V., Moreira, G.M., & Fernandes Filho. E.I. (2018) Multivariate analysis and machine learning in properties of ultisols (Argissolos) of Brazilian Amazon. Revista Brasileira de Ciência do Solo, 42, 1–20. https://doi.org/10.1590/18069657rbcs20170419
Tian, Z., Kool, D., Ren, T., Horton, R., & Heitman, J.L. (2018). Determining in-situ unsaturated soil hydraulic conductivity at a fine depth scale with heat pulse and water potential sensors. Journal of Hydrology, 564, 802–810. https://doi.org/10.1016/j.jhydrol.2018.07.052
Wilson, J.P., & J.C. Gallant. (2000). Digital terrain Analysis. p. 1–27. In J.P. Wilson and J.C. Gallant (ed.) Terrain analysis: Principles and application. John Wiley & Sons, New York.
Yang, T., Jia, X., Niu, L., Huang, L., Zhao, Ch., Li, X., Ren, X., Wang, R., & Shao, M. (2026). High-resolution mapping of saturated soil hydraulic conductivity across China’s drylands. International Journal of Applied Earth Observation and Geoinformation, 148, 105176. https://doi.org/10.1016/j.jag.2026.105176.
Zeraatpisheh, M., Ayoubi, S., Sulieman, M., & Rodrigo-Comino, J. (2019). Determining the spatial distribution of soil properties using the environmental covariates and multivariate statistical analysis: a case study in semi-arid regions of Iran. Journal of Arid Land, 11(4), 551–566. https://doi.org/10.1007/s40333-019-0059-9
Iranian Journal of Soil and Water Research
Volume 57, Issue 4
June 2026
Pages 847-867

Files

Share

How to cite

Statistics