A study, in the ISPRS Journal of Photogrammetry and Remote Sensing, presents a method and proves the effectiveness of a system for the direct measurement of the 3D spatial properties of a volcanic plume.
The shape of a plume is reconstructed in three dimensions using multi-view imagery collected from static ground cameras. Four similar NicAIR infrared camera systems were used.
Each system consists of a FLIR Photon 640 camera core with a pixel microbolometer matrix and thermal infrared sensitivity (7.5 to 13.5 μm). The cameras were equipped with 25mm or 35mm germanium lenses. The filter holder at the front contained up to four filters with narrow bandwidth (≈1μm wide, half maxi). The camera is calibrated in the laboratory using a black body source. It also contains a black-body shutter placed in the field of view before each scene view to allow temperature calibration in the field. The possibility of acquiring multispectral data on narrow-band channels makes it possible to estimate and mass-charge the ashes.
Fuego - (a) Example broadband thermal images shortly after an eruption with the plume visible above the summit. (b) The region of operation within 40 km of Guatemala City airport and near the often closed R644 airway. (c) The arrangement of the cameras during the November 2017 campaign. The summit vent is denoted by FGO. Major population centres are also shown. (d) The NicAIR multi-spectral IR camera. - Doc. Measurement of three dimensional volcanic feather fields using multiple ground based infrared cameras - K.Wood & al.
The method was developed from data collected during an expedition to Fuego, Guatemala, where four thermal infrared cameras were deployed to capture simultaneous images of ash-rich regular eruptions. A spatial partitioning method was applied to the problem to estimate the volume of the plume at any time.
By successively applying the method to sets of sequential images, other quantitative measures can be deduced such as the direction of drift, the ascent rate and the dispersion rate.
The entire process of the method is presented, including data capture, calibration processes, image processing, space cutting method and practical implementation issues. The method is sensitive to the alignment of the camera. A new technique for estimating the angles of orientation of the camera, usable if a model of terrain of high precision, is described. Other sources of error relating to the number, timing and resolution of cameras are also discussed. Preliminary results are presented using data collected on the Volcán de Fuego in November 2017 over a period of 1.25 hours, including three separate eruptions.
Fuego - A diagram showing the method of cutting space. A 3D space region (voxel) is projected into the 2D image planes (pixels) of the cameras surrounding the plume. If the location of the pixel is in the plume for all cameras, the voxel is kept (green). If the location of the pixel is outside the panache for one or more cameras, the voxel is refreshed (red). The process is systematically repeated on a search volume until only the voxels identified as plume remain. - Doc. Measurement of three dimensional volcanic feather fields using multiple ground based infrared cameras - K.Wood & al.
For each eruption, the number of voxels detected (a voxel is a 3D space region) initially increases due to the natural expansion of the plume. The volume then reaches a maximum before decreasing. This is because the plume dilutes below the 2D segmentation threshold, and also because it drifts beyond the camera view. The most accurate detection period is when the plume is visible to all cameras.
Total volume, plume height, horizontal drift velocity, drift heading can all be calculated.
This work represents the first step towards a three-dimensional recovery of ash mass and gas concentrations. For an ash concentration map to be estimated, a hypothesis must be made on the particle size distribution. This could be based on direct measurement (for example, using drones), or using methods developed from 2D image analysis (Lopez et al., 2015). This would significantly change the ability to monitor volcanic emissions and manage airspace more effectively.
Source and details : Science Direct - ISPRS Journal of Photogrammetry and Remote Sensing
Measurement of three dimensional volcanic feather fields using multiple ground based infrared cameras - K.Wood & al. - link