Qualifying Stream Temperature Measurements of an Aerial Thermal Infrared camera While Gathering Evidence of Thermal Refugia for the City of Frederick, Maryland

Abstract

Thermal refugia, due to seepages and tributaries, provide mild-temperature sanctuaries relative to the more extreme water temperatures in the winter and summer. Exploring streams for thermal seepages via in-water probes can be a slow and time-consuming endeavor, especially when seasonal timing is of the essence. One way to offset this time-debt could be using a Thermal Infrared Camera (TIR) attached to a small Unmanned Aerial Vehicle (sUAV), yet in order to implement this new technology it is necessary to compare the measurements between the two devices to ensure that the drone can effectively replace the probes. This research compared the temperature measurements from the TIR against in-water temperature probes at three sites, one with a known seepage and two with suspected seepages, by analyzing agreement via Bland-Altman statistical analysis. The results of the analysis suggested little agreement between the two devices, and a broad range in the limits of agreement. Evidence of the TIR camera’s ability to differentiate temperatures but not measure them accurately was provided in the form of value-value plots and side-by-side TIR/RGB photographs. This research concluded that the sUAV’s TIR camera was useful for photographing and investigating temperature differentiation but true temperature needed to be evaluated with standardized, reliable ground-truthing methods. Evidence for the presence or absence of seepages at two sites designated by the City of Frederick Sustainability Department was provided in the form of side-by-side TIR/RGB photographs and value-value plots comparing the temperature difference along the stream reach as well as across the stream channel.