Soil moisture observation from a drone
In this project, I worked with my Ph.D. advisor, Professor Teamrat Ghezzehei, and Professor Josh Viers, Anna Fryjoff-Hung and Andreas Anderson to develop a powerful machine learning model to interpret soil moisture at a high spatial resolution based on multispectral imagery captured by small unmanned aircraft system (UAS).
Using photogrammetry from the images, we generated a high resolution (10 cm) digital elevation model and calculated several topographic parameters at multiple scales. The top most important variables for predicting soil moisture were precipitation and evapotranspiration variables, reflectance in the red band, and topographic position indices (TPI). Our results demonstrate that the dynamics of soil water status across heterogeneous terrain may be adequately described and predicted by UAS remote sensing data and machine learning. Our modeling approach and the variable importance and relationships we have assessed in this study should be useful for management and environmental modeling tasks where spatially explicit soil moisture information is essential.
This project is partly published in chapter four of my dissertation and is currently in preparation for publication in a peer-reviewed journal.