METHODS FOR CORRECTING GROUND-BASED TIME-LAPSE INFRARED IMAGERY

The use of infrared (IR) imagery to evaluate stream temperature patterns is increasing due to the greater spatial resolution this data provides compared to traditional methods. Traditional methods such as discrete data loggers, are capable of providing stream temperature data with high temporal resolution, whereas IR imagery from satellites, drones and other aircraft only yield snapshots of temperatures at a point in time. At suitable field sites, ground-based time-lapse IR imagery may provide a means of acquiring stream temperature data with both high spatial and temporal resolution, though few studies have used this technique. A main limitation of this technique is the ability to separate the reflected temperature from the stream surface temperature when the camera is at an oblique angle. In this study, we explored two methods of correcting ground-based IR data to obtain accurate stream temperatures so that the high spatial and temporal resolution temperature data can be used within an energy balance model.

We captured ~1200 time-lapse IR images at 10 minute intervals during two field excursions using a Jenoptik VarioCam HD IR camera. Both sets of images were taken from cliffs along the study reaches. The first reach was ~500 m long with an observation angle ~75° from perpendicular, while the second reach was ~80 m long with an observation angle ~70° from perpendicular. We deployed temperature sensors in the stream and land surface as control points. Meteorological data were recorded concurrently with the IR data. We explored two approaches to correct these sets of IR images, which displayed evidence of reflections. One was an analytical approach combining published atmospheric transmissivity and surface emissivity values. Using these data, we investigated the range of reflected temperatures and found reflected temperatures were typically cold, indicating the influence of clear sky reflections. The other was an empirical approach using in-stream temperature sensors to create point-in-time correction factors that could be applied to the sets of images. These approaches demonstrate two options for obtaining absolute stream temperatures from low angle ground-based time-lapse IR imagery, and the potential to obtain high spatial and temporal resolution IR temperature data for streams in appropriate settings.