Journal
GEODERMA
Volume 259, Issue -, Pages 1-11Publisher
ELSEVIER
DOI: 10.1016/j.geoderma.2015.05.002
Keywords
Soil physical quality indicators; Sustainable productivity; Modeling; Particle swarm optimization; Decision tree
Categories
Ask authors/readers for more resources
Soil physical quality indicators (SPQIs) are used to make complex information more accessible to decision makers; however, their site-specific nature may restrict their applicability only to a specific soil type and/or management system. Therefore, it would be advantageous if such indicators could be predicted indirectly from easily available soil properties using inexpensive methods. In this study, we introduce a hybrid algorithm, specifically designed to work with optimized decision tree with particle swarm optimization (PSO-DT), for the prediction of SPQIs (i.e., air capacity, AC; plant-available water capacity, PAWC; and relative field capacity, RFC). The potential power of using the PSO-DT algorithm in setting up a framework for identifying the most determinant parameters affecting the physical quality of agricultural soils in a semiarid region of Iran (Baft plain, 29 degrees 11' to 29 degrees 13' N and 56 degrees 34' to 56 degrees 38' E) was also investigated. An empirical multiple linear regression (MLR) model was constructed as benchmark for the comparison of performance. In results, a permutation of five input features, including soil organic matter (SUM), electrical conductivity (EC), clay, sand, and bulk density (BD), was introduced by the hybrid PSO-DT algorithm as explanatory variables. Using the PSO-DT method resulted in higher model efficiency and coefficient of determination (R-2) than the MLR approach. The obtained R2 values for the constructed PSO-DT model for the AC, PAWC, and RFC predictions were 0.91, 0.90, and 0.96, respectively, whereas they were 0.61, 0.16, and 0.47 for the MLR-model. The SOM, clay, and sand parameters were accounted as the discriminating variables of the models constructed for the prediction of AC, PAWC, and RFC indicators, respectively. This study provides a strong basis for the prediction of SPQls and identifying the most determinant parameters influencing the physical quality of agricultural soils in semiarid regions of Iran; however, its general analytical framework could be applied to other parts of the world with similar challenges. (C) 2015 Elsevier B.V. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available