UAV-Based Land Surface Temperatures and Vegetation Indices Explain and Predict Spatial Patterns of Soil Water Isotopes in a Tropical Dry Forest

Abstract

The spatial variation of soil water isotopes (SWI)—representing the baseline for investigating root water uptake (RWU) depths with water stable isotope techniques—has rarely been investigated. Here, we use spatial SWI depth profile sampling in combination with unmanned aerial vehicle (UAV) based land surface temperature estimates and vegetation indices (VI) in order to improving process understanding of the relationships between the spatial variability of soil water content and soil water isotope patterns with canopy status, represented in the form of VI. We carried out a spatial sampling of 10 SWI depth profiles in a tropical dry forest. UAV data were collected and analyzed to obtain detailed characterization of soil temperature and canopy status. We then performed a statistical analysis between the VI and land surface temperatures with soil water content and SWI values at different spatial resolutions (3 cm–5 m). Best relationships were used for generating soil water isoscapes for the entire study area. Results suggest that soil water content and SWI values are strongly mediated by canopy parameters (VI). Various VI correlate strongly with soil water content and SWI values across all depths. SWI at the surface depend on land surface temperature (R² of 0.66 for δ¹⁸O and 0.64 for δ²H). Strongest overall correlations were found at a spatial resolution of 0.5 m. We speculate that this might be the ideal resolution for spatially characterizing SWI patterns and investigate RWU in tropical dry forest environments. Supporting spatial analyses of SWI with UAV-based approaches might be a future avenue for improving the spatial representation and credibility of such studies.