Evaluating some pedotransfer functions for simulation of transient water flow in soil

Subject Areas : Farm water management with the aim of improving irrigation management indicators

Sanam Jafari Gilandeh ¹ , Ali Rasoulzadeh ² , Habib Khodaverdiloo ^{3 *}

1 - M.Sc. student, Department of Soil Science, Urmia University, Urmia 57135-165, Iran
2 - Assistant Professor, Department of Water Engineering, University of Mohaghegh Ardabili, Ardabil, Iran
3 - Assistant Professor, Department of Soil Science, Urmia University, Urmia 57135-165, Iran

Received: 2012-12-06 Accepted : 2013-05-14 Published : 2013-09-22

Keywords: functional evaluation, Pedotransfer functions, soil water flow, simulation models,

Abstract :

Quantitative description of soil hydraulic properties is necessary for any scheduling related to soil and water resources conservation. Soil hydraulic properties are important inputs for simulation of water flow and solute transport in soil. Since direct measurement of these properties is time consuming and costly, pedotransfer functions (PTFs) have been widely and successfully used for their prediction. Yet, little efforts have been made for functional evaluation of PTFs predictions for application in simulating transient soil water flow. In this study, soil water retention curve (SWRC) of a clay soil was predicted, using some selected local and global PTFs. SWRC was also measured in the laboratory by direct method. Validity of studied PTFs in terms of prediction of SWRC was examined. By applying both predicted and measured hydraulic parameters to HYDRUS-1D program for simulation of soil water flow, functional behavior of PTFs was quantitatively compared in terms of simulation of water flow in soil. The obtained results indicated that both the selected global PTFs (Rosetta) with root mean square error (RMSE) < 0.025 cm3cm-3 and some regional PTFs (Ghorbani-1) with RMSE < 0.014 cm3cm-3 can reasonably well predict SWRC of soil surface. For prediction of SWRC of subsurface soil, the RMSE value of global PTFs was larger than 0.107 cm3cm-3 and was ranged from 0.036 to 0.356 cm3cm-3 for the local PTFs. However, for simulation of soil water flow a different set of PTFs was most efficient. RMSE values of adjusted Ghorbani PTF (S-Gh&H (Adj.)) for simulation of water content of surface and subsurface soils were 0.025 and 0.055 cm3cm-3, respectively. RMSE value of Wosten (1997, 1999) PTFs for surface soil was larger than 0.149 cm3cm-3 and for subsurface soil was less than 0.058 cm3cm-3. Simulation with the measured SWRC had RMSE values of 0.013-0.040 cm3cm-3 for surface and subsurface soils, respectively. It can be concluded that when validating PTFs, one should consider the objective for which the PTFs are assessed. A PTF might be accurate enough for predicting SWRC, but not for particular q(h) range governing the soil water redistribution process. Using such PTFs might lead to large errors in simulating soil water content.