Author: Steve Stampfli, White Salmon, WA, USA, firstname.lastname@example.org
Last summer’s Eagle Creek fire, plus a huge construction project along Interstate 84, greatly hampered on-site investigation of Shellrock Mountain’s thermal features and pikas. Without access for installing sub-surface and surface temperature loggers, I had to look for a more remote means of investigating the behavior of Shellrock’s unique (possible ice-cored) talus mantles. This was probably fortunate, since it eventually led to consideration of the use of long wavelength infrared (LWIR) imaging for documenting variations in slope temperature, chimney effect thermal processes, and ultimately the identification of potential American pika habitat at Shellrock and other places in the Columbia River Gorge.
Having never used infrared imagery, I initially wondered whether anyone had attempted similar use. Geological applications are certainly not billed front-and-center by any of the major manufacturers. Nor did it appear that talus slope investigators had used it for analyzing the chimney effect, periglacial features, or other thermal slope processes. (I learned, however, that at least one investigator (Aaron Johnston of the USGS) is beginning to use a drone-mounted infrared camera for understanding slope-specific habitat processes related to pikas). The only other real geologic and ecologic applications I came across were for the study of glacier behavior, soil temperature, evapotranspiration, location of wildlife, crop and plant community health, and maybe a few others.
In July 2018, I purchased a non-cooled, 640×480 pixel resolution LWIR camera from the South Korean company, i3system, Inc. Like most makers, the bulk of this company’s research and production is geared toward military and police products for surveillance, industrial product inspection, medical health diagnosis, building inspection, and various night vision purposes. And as with many high technology gadgets, infrared equipment specifications are getting progressively better, even as costs decline. Today’s 640×480 resolution (0.3 megapixel) thermal cameras might cost several thousand dollars, whereas only 2-3 years ago the price was tens of thousands.
Why the high cost of thermography cameras? First, there isn’t a broad demand, so companies can’t sell enough units to justify the high initial research and production costs. This is exacerbated by differences in the silicon chip-mounted (light vs heat) sensors. Standard photography camera sensors are based on a matrix of photo-detectors that transduce visible light intensities and wavelengths into electric signals. Thermography camera sensors, on the other hand, use a different technology based on a matrix of chip-mounted thermometers that actually measure the temperature of each pixel that passes through the lens (an amazing 307,200 such thermometers in the case of a 640×480 resolution chip). This data is then processed and interpreted as the surface temperatures of the objects being charted. A final cost factor relates to the use of the rare and costly element known as germanium (instead of silicon) for the camera’s primary optics. Interestingly, while visible light is freely transmitted through silicon dioxide glass, it largely blocks the transmission of heat radiation. Instead, elemental germanium or Ge02 (germanium based glass) are used in the manufacture of LWIR camera optics, materials that are incidentally opaque to visible light and oddly have the appearance of polished metal.
Soon after beginning to research the use of thermal imaging for assessing Gorge talus slopes, I learned there were many limitations that could result in failure. Foremost were the relatively great distances I would be forced to photograph the slopes from. Oregon-side targets would need to be photographed from the Washington shoreline, and depending on river width and talus slope location, the distances would generally range between ½ to 1 ½ miles. Such distances could be concerning, given the relatively low pixel resolution of affordable thermal cameras, and likely inability to detect even relatively large targets at such distances. Adding to this was the fact that heat radiation from distant objects is attenuated by the atmosphere and certain atmospheric conditions. Many times, however, the specific project goals have less to do with measuring the actual temperature of distant target, and more to do with detecting apparent temperature differences between objects in the scene.
Given these limitations, I decided that the moderately high resolution (640×480) i3System Inc. camera mounted with a relatively narrow angle 35 mm lens, would most likely provide decent resolution of small targets at distance. Plus, given the fact that I was mainly interested in detecting relative temperature differences related to chimney effect venting on talus slopes, the possibility of inaccurate spot temperatures was of little concern.
After about one month of experimenting with the new camera, it is obvious that the technology can be very valuable in the study of how mountain slopes interact with (and influence) their physical and biological surroundings. The header image is a fun example of the technology, showing a Union Pacific train on the Oregon side as it passes in front of a patch of cold ground at the base of Shellrock Mountain. The photo was taken at 10:00pm, August 8, 2018, at a distance of 0.55 miles, or 2900 feet. What does the image show in regard to eventual project capability? First, the camera’s resolution allows interpretation of objects as small as the 40” diameter railroad car wheels from 0.55 miles. Good detection, in this case, is due to resolution plus the significant temperature difference between the train’s hot wheels and cool slope background. Zooming-in on the image even shows some details of the new rock-fall fence along I-84. Second, the image allows fairly easy interpretation of trees growing on the slope, and other background features. Finally, and most important, the image allows easy understanding of relative slope temperature differences. What looks like a white plume being emitted from the train’s locomotive is in reality cold ground in the slope behind the train.
A soon coming article will highlight the emerging results of using thermography for the study of talus features pertaining to pika habitat.