1. Thermal cameras on unoccupied aerial vehicles (UAVs) are increasingly being used in environmental and ecological research, including hydrology, wildfire detection and prediction, urban heat studies, precision agriculture, ecosystem functioning, wildlife monitoring and microclimate studies.
2. Converting raw thermal signals to quantitative land surface temperature (LST) values requires careful application of correction procedures. However, these steps are often overlooked or ignored—either due to limited expertise in thermal remote sensing or because of the technical complexity involved. Neglecting corrections for atmospheric effects and surface emissivity can lead to discrepancies of up to 5°C in the resulting LST estimate.
3. We introduce theRmalUAV, an R package that facilitates LST processing with two workflows: an orthomosaic-based and an image-based approach. The orthomosaic workflow applies a single function to the entire dataset, whereas the image-based workflow can account for variations in environmental conditions during the flight that affect surface temperature. The package corrects for atmospheric effects, background temperature, spatial emissivity and weather fluctuations, incorporating a novel method to handle rapid illumination changes. The package currently supports 11 common thermal sensors. It also includes tools for data cleaning, co-registration and reporting.
4. We demonstrate both the importance of the workflow and its implementation using two distinct case studies to highlight its versatility. The main text presents a detailed example using the research-grade TeAx ThermalCapture 2.0. A complementary example, featuring the more commercially oriented DJI Mavic 3T, is provided in the Supporting Information. For comprehensive guidance and tutorials, readers are directed to the package vignette and its companion website.