Using UAV-based Infrared Thermometry in the identification of gas seeps: a case study from Ciomadul dormant volcano (Eastern Carpathians, Romania)

In our study, we tested a UAV-based IRT and Structure from Motion (SfM) for the identification of CO2 rich gas emission areas at Ciomadul dormant volcanic area, Eastern Carpathians. Our aim is to demonstrate the efficiency of the identification method providing example from a case-study in the Eastern Carpathians.

The gas emissions from Ciomadul come with high flux and are of magmatic origin, associated with the volcanic activity in the past. We had the following assumptions before performing the measurements with the drone: the temperature of the gas vents is constant, as well as their flux, variability is represented only by the changes in ambient temperature. We had previous knowledge on the temperature of the gas emissions (6 °C), so we chose periods when the ambient temperature is either lower or higher than the temperature of the gas. We performed several field observations with the camera both at daytime and in the evening.

The acquisition of UAV photography was made using a DJI Mavic 2 Enterprise Dual drone. This device is equipped with a 12 MP visual camera (RGB) with a 1/2.3" CMOS sensor. The visual camera has a lens with field of view of approx. 85°, 24 mm (35 mm format equivalent) lens with an aperture of f/2.8. It was also equipped with an Integrated Radiometric FLIR® Thermal Sensor. It is an Uncooled VOx Microbolometer with a horizontal field of view of 57° and f/1.1 aperture, sensor resolution is 160x120 (640x480 image size) and a spectral band of 8-14 μm.

The gas vents were clearly visible on the thermal images, and we discovered additional seeps that were not identified before. Later we confirmed the presence of the gas emissions with in situ measurements on the concentrations of CO2. The visibility of the gas emissions was influenced by parameters like temperature, the orientation of the gas vent, the influence of sunlight, the flux of the gas vent, etc.