Spatial predictions of soil moisture across a longitudinal gradient in semiarid ecosystems using UAVs and RGB sensors

Authors: Hernandez, Alexander; DUARTE, EFRAIN - Orise Fellow; Porter, Peter; Brecht, Holden

Submitted to: Geocarto International
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/28/2025
Publication Date: 3/2/2025
Citation: Hernandez, A.J., Duarte, E., Porter, P.W., Brecht, H.J. 2025. Spatial predictions of soil moisture across a longitudinal gradient in semiarid ecosystems using UAVs and RGB sensors. Geocarto International. https://doi.org/10.1080/10106049.2025.2461523.
DOI: https://doi.org/10.1080/10106049.2025.2461523

Interpretive Summary: Soil moisture mapping and monitoring play an essential role in the understanding of how ecosystems services are performing. This is crucial in semiarid lands where water is a limiting factor in agricultural crop development and in natural vegetation growth. Obtaining dependable maps of soil moisture for large landscapes is relatively straightforward with satellite data and technology. However, the grain or resolution of these maps usually exceeds several kilometers, and this type of scale is typically too coarse for precise assessments at the farm level. To obtain estimates of soil moisture at a reasonable scale of meters or centimeters, users require high end multispectral and thermal sensors which are usually not accessible for the general population due to their cost and complexity to use and process the data. We developed a protocol to obtain reliable estimates of soil moisture at a fraction of centimeter scale using readily available RGB sensors that can be mounted on most commercial unmanned aerial vehicles (“aka” drones). This protocol will provide rangers, farmers and managers of the land with a relatively pragmatic technology to derive their own soil moisture estimates at the farm or stand level. Such maps can be of extremely high resolution and will show the spatial variability of soil moisture at the time of the drone flight. This proof-of-concept can provide users of semiarid lands the opportunity to monitor changes in soil moisture as well as changes in vegetation with technology that is generally available to most users. This is the case of low-cost commercial RGB sensors and drones.

Technical Abstract: Unmanned aerial vehicles (UAVs) offer an efficient method for assessing and monitoring physical phenomena, including soil moisture (SM), particularly in semiarid regions. UAV-based RGB sensors were used to collect high-resolution imagery, and hundreds of SM samples were gathered concurrently with the UAV flights across nine study sites over a large latitudinal gradient in the western USA. We evaluated the predictive power of RGB bands, texture metrics, and vegetation indices for estimating SM using machine learning algorithms. The model showed a moderately acceptable predictive accuracy (R² = 0.63 using cross validation) and (R² = 0.53 using a fully independent validation). Texture metrics such as "mean" and "entropy," as well as the Excess Green Index (ExG) vegetation index showed the maximum predictive power while RGB bands showed minimal performance. The resulting spatial predictions showed high reliability (a < 0.01) for the States of Utah and California but exhibited a poorer performance for Idaho and Montana. We provide linear equations for the conversion of raw digital number (DN) values to reflectance, facilitating remote sensing applications that benefits from UAV simple and highly affordable RGB imagery. Our protocol provides a robust pathway to modelling SM with cost-effective solutions for monitoring semiarid ecosystems.