Fun with thermal cameras

We spent the day doing some tests to see if color had an impact on the surface temperature of horses.

A few summers ago, I wrote a post about how the heat generated by the horse's circulatory system produces the dappling pattern on some greys ("Dapples are hot!"). I used this image to show how the temperature variation matched the light and dark portions of the coat. 

A reader asked if the horse in the image couldn't be dapple grey. Was it possible that the temperature difference was because the light and dark areas of the coat absorbed and reflected heat differently? I knew the normal procedure for thermal images was to get the horse warm from work and then bring it into a cool, dark area.

It was unlikely that color and light had anything to do with the image, but I didn't take the picture. Although I doubted the horse was dapple grey, I did not know its color. To test this, I needed a pair of images of an ordinary bay or chestnut horse: one traditional photo and one thermal image.

It took longer than I expected, but a few weeks ago, all the stars aligned. Alan and I were both available. It was chilly enough that my thermal camera would register the heat on the horses well, but the weather was clear, so my digital camera would get good shots. There was a hunter show nearby, which meant most of the horses would likely be clipped. (The camera cannot "see" past a heavy winter coat.)

Here is my ordinary bay horse.

Here is the thermal image of the same horse taken at the same time.

Although the resolution isn't as good as the earlier image, the dappling is visible. 

I should also mention that the original image had the order of increasing heat as yellow-orange-red. (See the color bar to the right of the horse.) I have the palette set differently on my thermal camera, which is why I didn't have to inverse the image to make the dapples read correctly. 

In hindsight, I wish I had asked Alan to borrow one of the higher-resolution cameras. I have the FLIR TG165. I got the camera to teach seventh-graders about heat transfer, which it does well. However, to get dappling, we had to get closer to the horses than was practical. We also couldn't remove the horse to a dark area for the picture, which would have maximized the contrast we could get. 

I did realize that we had another opportunity to show that heat from the horse—and not light absorbed or reflected by the colors on the horse—was what the camera registered. We didn't need a high-resolution camera to pick up the pattern on a black tobiano! You can see Alan holding the FLIR camera behind the pony at the top of this post. If the difference between light and dark fur influenced the camera, we should see the pattern in the thermal image. 

I took the image the thermal camera took and then duplicated it with an overlay of the patterning on the hindquarters. 

For my camera's palette, red areas are cooler than yellow. That means there are two cooler (red) patches on the pony's rump: one near the dock of his tail and one just below the point of his buttocks. Looking at the overlay image, the coat near the dock is black, while the area of the buttock is white. Meanwhile, the clear line of color down the hindquarter isn't reflected in the temperature map at all. The tobiano pattern does not show in the thermal image. All the camera is registering is the heat from within the horse, not anything absorbed or reflected by the coat.

We were able to confirm that the dappling in the original thermal image came from differences in body heat and not the color of the coat. Another interesting fact was that even when we were close to the horse, it took looking at thermal images of several different horses before we found one with a visible dappling pattern. That natural variation in their circulatory system might explain why some greys go through a dappling phase while others do not.