Developing a fuzzy logic model for predicting soil infiltration rate based on soil texture properties

  • Ahmed Z Dewidar Agricultural Engineering Department, King Saud University, Riyadh 11451, Kingdom of Saudi Arabia
  • Hussein Al-Ghobari Agricultural Engineering Department, King Saud University, Riyadh 11451, Kingdom of Saudi Arabia
  • Abed Alataway PSIPW Chair, Prince Sultan Institute for Environmental, Water and Desert Research, King Saud University, Kingdom of Saudi Arabia
Keywords: FLS, infiltration rate, MLR, modelling, sensitivity analysis


The prediction of the soil infiltration rate is advantageous in hydrological design, watershed management, irrigation, and other agricultural studies. Various techniques have been widely used for this with the aim of developing more accurate models; however, the improvement of the prediction accuracy is still an acute problem faced by decision makers in many areas. In this paper, an intelligent model based on a fuzzy logic system (FLS) was developed to obtain a more accurate predictive model for the soil infiltration rate than that generated by conventional methods. The input variables that were considered in the fuzzy model included the silt and clay contents. The developed fuzzy model was tested against both the observed data and multiple linear regression (MLR). The comparison of the developed fuzzy model and MLR model indicated that the fuzzy model can simulate the infiltration process quite well. The coefficient of determination, root mean square error, mean absolute error, model efficiency, and overall index of the fuzzy model were 0.953, 1.53, 1.28, 0.953, and 0.954, respectively. The corresponding MLR model values were 0.913, 2.37, 1.92, 0.913, and 0.914, respectively. The sensitivity results indicated that the clay content is the most influential factor when the FLS-based modelling approach is used for predicting the soil infiltration rate.

  • Abstract 67
  • PDF 39
Views and downloads are with effect from 11 January 2018
Research paper