Machine Learning Pedotransfer

Saturated hydraulic conductivity changes with soil organic carbon content.

In this project, I worked with my Ph.D. advisor, Professor Teamrat Ghezzehei to develop machine learning models that were able to predict soil hydraulic conductivity–one of the key soil water properties–with significant improvement in accuracy compared to previous models. Details of this project can be found in our publication at Water Resource Research Journal.

I am committed to the principles of reproducible research and making scientific findings widely accessible. I have shared the entire project codes of my research in a public GitHub repository. There, I have also developed and shared an application with a graphical user interface to enable easier use of the models I developed. Pedotransfer App

Avatar
Samuel N Araya
Postdoctoral Research Fellow

My research interests include soil science, machine learning and spatial analysis.