Abstract:
Soil samples from Ajay river basin has been collected and analyzed for its physical
properties (texture and bulk density). The water retention limits of soil at two points
(i) field capacity (-33 kPa) and (ii) permanent wilting point (-1500 kPa) has been
measured in laboratory. In this paper an effort has been made to estimate the soil
water limits from easily available soil texture and bulk density data of Ajay river
basin. Three models- Brooks-Corey (BC), Multilinear Regression (MLR) and Artificial
Neural Network (ANN) have been developed to estimate the soil water retention
limits at the above two points. Parameters of the Brooks-Corey model has been
estimated using texture and bulk density obtained from soil analysis. Multi linear
regression analysis has been performed to develop regression equation using texture
and bulk density as independent variables. ANN method has been applied taking
texture and bulk density as input and the measured water retention as output.
Comparison has been made between measured and estimated values of water retention
by the above three models. It is found that the soft computing technique of ANN has
given better output with minimum error with higher R2 value than BC and MLR
model.
