Proceedings of the 4th World Conference on Climate Change and Global Warming
Year: 2024
DOI:
[PDF]
Innovative SelectKbest-based GEP Model to Develop Soil Water Characteristics for Sustainable Irrigation Scheduling in the Face of Climate Change
Aitazaz A. Farooque, Saad J. Cheema
ABSTRACT:
Soil water characteristics curves (SWCC) are critical to predict soil water retention and water movement that impacts agricultural productivity and crop yields. Efficient water resource management and environmental conservation require in-depth knowledge and understanding of soil and water dynamics in the face of climate change. This study aims to enhance the understanding of SWCC in Atlantic Canada and explore the suitability of analytically derived SWCC models using the pressure membrane apparatus. This research strives to develop a generalized machine learning (ML) model using a novel SelectKbest-based gene expression programming (GEP) technique. The suitability of the developed models was evaluated for three soil textures (loamy, sandy loam, and loamy sand soils). SWCC had various measured suction potential levels and model parameters to determine the optimum soil hydraulic parameters. HydroMe program was calibrated and validated with standard retention curves for selected regions and soil types. The coefficient of determination (R2) and root mean square error (RMSE) were employed to assess the robustness and accuracy of the developed models. In addition to a novel ML-based model, a SelectKbest-based GEP model was developed to generalize the mathematical SWCC model for Atlantic Canadian soils (R2=0.976 and RMSE=0.04). These novel approaches can significantly improve site-specific agricultural water management strategies in the changing climates. The methodology developed in this study can be adapted to other regions and soil types which can enhance global agriculture and environmental management efforts to promote sustainable agriculture.
keywords: Climate change; Machine learning; Soil moisture; Water retention curves; Sustainable agriculture