willner@cfa.harvard.edu (Steve Willner) (12/21/89)
From: willner@cfa.harvard.edu (Steve Willner) >From article <12209@cbnews.ATT.COM>, by christ@sci.ccny.cuny.edu (Chris Thompson): > I'm a little confused about the IR, though. [...] Now I know > that there is a large IR signature associated with heat, but I think you > can have thermal imaging which is not dependant on IR. I've only worked > with available light image-intensifiers, so I'm not positive about this, > but surely it is related to the same technology which measures body > temperatures? I'm referring to the medical uses, which scan for cancers > by locating patches of high-temperature cells. While it's possible that the military might use some special terminology, in general "infrared imaging" and "thermal imaging" are the same. If a distinction had to be made, the former term would refer to any imaging by means of infrared light, while the latter term would refer specifically to imaging an object's own thermal emission. For those who are not up on the basic physics, a more detailed explanation follows. It leaves out lots of details, but the overall picture shouldn't be misleading. First of all, "infrared" is simply light, i.e., electromagnetic radiation. The distinction from visible light is that the wavelength is longer and so detection techniques have to be different (e.g. eyeballs don't work), but propagation and optics are much the same as for visible light (leaving out "details" like atmospheric and glass absorption at certain wavelengths). In particular, clouds and fog affect infrared almost exactly the same as visible light, since the water droplets are large compared to the wavelengths. Another physical fact is that _any object_ at a finite temperature emits electromagnetic radiation; this radiation is called "thermal emission." The wavelength of peak emission (in microns) is about 2900/T, where T is the object's absolute temperature in kelvin. (Composition can make some difference but is generally less important than temperature.) Thermal radiation is emitted over a broad range of wavelengths but falls off rapidly at wavelengths shorter than the peak. (A factor of 3 in wavelength drops the radiation emitted by a factor of 100.) Thus to see an object visually via its thermal emission, its temperature has to be about (2900/3)/0.7 kelvin or about 2000 degrees F, i.e. "red hot." (Visible light is roughly 0.4 to 0.7 microns.) "White hot" requires a temperature of about 2900/0.4=7250 K; the Sun is in fact just about this hot. (Really 6000 K, but that's "equal" to within the accuracy of this calculation.) One last fact is that thermal emission at _any_ wavelength increases as the emitting object gets hotter. Infrared sensors typically employed operate in three general ranges: "short wavelength infrared" (SWIR), 1.5-2.5 microns; "medium wavelength infrared" (MWIR), 3-5 microns; and (you guessed it!) "long wavelength infrared" (LWIR), 8-12 microns. From the formula above, it's easy to estimate the temperature that each can easily detect: 483 K (410 F); 242 K (-24 F), and 97 K (-285 F). The last two may seem absurd, but they're not; every "room temperature" object really is giving off copious amounts of infrared radiation at these wavelengths. The proverbial "black cat in a coal bin" is an easy target for MWIR or LWIR sensors as long as the cat is a bit warmer (or colder) than the surrounding coal. Exact capabilities depend, of course, on the sensors, the targets of interest, and the background against which those targets are observed. (Detecting an object against cold space is lots easier than detecting the same object against the warm Earth.) And detecting an object isn't necessarily the same as tracking it. (That's why flares are useful as decoys.) Still, the above information should provide a pretty good idea of what is possible. ------------------------------------------------------------------------- Steve Willner Phone 617-495-7123 Bitnet: willner@cfa 60 Garden St. FTS: 830-7123 UUCP: willner@cfa Cambridge, MA 02138 USA Internet: willner@cfa.harvard.edu