Reading — IR/EO Fundamentals#
By the end of this lesson you should be able to:
Explain why every warm object radiates, and how temperature sets both how much it emits and where that energy peaks.
Identify the MWIR and LWIR atmospheric windows and explain why the IR fight lives inside them.
Distinguish reflective-band sensing (needs illumination) from emissive-band sensing (sees the target’s own heat).
Contrast photon and thermal detectors, and explain why high-performance IR sensors are cooled.
Up out of the radio spectrum#
Block 2 fought in the radio bands, and its central fact was that a radar must radiate to work — the emission is a weakness you can hear, geolocate, and jam. Block 3 climbs several decades up in frequency to the infrared (IR) and electro-optical (EO) domain, and the physics inverts in a useful way. Here the emission is not a choice the enemy makes. Every object warmer than absolute zero radiates on its own, all the time, whether it wants to or not. A jet cannot switch off its exhaust plume the way it can switch off its radar. That is the promise of IR: a passive, no-warning way to see heat. The catch, which the rest of this reading develops, is that the atmosphere and the weather get a vote.
Everything warm glows#
Any object above absolute zero emits thermal radiation across a spread of wavelengths. The idealized perfect radiator is a blackbody; real surfaces emit some fraction of that ideal, set by their emissivity. Two numbers pulled from an object’s temperature define its entire IR signature: how much it radiates and where that emission peaks.
The full shape of the emission curve — power versus wavelength — is given by Planck’s law. We treat it conceptually here; the point is not the constants but the shape: a single humped curve whose height and peak both climb with temperature. Two rules of thumb extract the operationally useful parts of that curve.
The first is Wien’s displacement law, which locates the peak:
As temperature rises, the peak wavelength shifts shorter — bluer. The Sun at roughly 5800 K peaks near 0.50 µm, in the visible. A jet plume near 1000 K peaks near 2.9 µm, at the edge of the MWIR. A tailpipe near 800 K peaks near 3.6 µm, squarely in the MWIR. Aero-heated skin near 325 K peaks near 8.9 µm, in the LWIR, right alongside a 290 K terrestrial background near 10 µm. Hot things peak in the mid-wave; cool things peak in the long-wave.
The second is the Stefan-Boltzmann law, which sets the total:
Total emitted power over all wavelengths scales with the fourth power of absolute temperature. Doubling the temperature radiates roughly 16 times as much power, and the same arithmetic applies between any two sources: a 1000 K plume out-radiates 325 K skin, per unit area, by \((1000/325)^4 \approx 90\). This fourth-power steepness is why a hot exhaust component can dominate an aircraft’s IR signature even though it is a small patch of surface area.
Key Concept
Two temperature facts define an IR signature. How much is emitted follows Stefan-Boltzmann (\(M \propto T^4\)); where it peaks follows Wien (\(\lambda_{\text{peak}} \approx 2898/T\)). Together they say: hotter sources radiate far more power and peak at shorter wavelengths.
The windows we get to fight in#
Radiating heat is only half the battle; that heat has to survive the trip through the atmosphere. Water vapor and carbon dioxide (CO₂) absorb broad stretches of the IR, closing whole regions to any sensor on the far side. What remains are a few relatively transparent atmospheric windows, and the entire IR fight is confined to them:
VIS / NIR / SWIR (~0.4–2.5 µm) — the reflected-light bands. This is where night-vision goggles and reflected-imaging systems work.
MWIR (3–5 µm) — the mid-wave window. Home of hot targets: plumes, engines, tailpipes.
LWIR (8–12 µm) — the long-wave window. Home of cool targets: skin, terrain, and most forward-looking infrared (FLIR) imagery.
Between the two IR windows sits a wall. The 5–8 µm region is largely opaque, blocked mainly by water vapor, and CO₂ cuts an additional absorption notch near 4.3 µm inside the MWIR. Because water vapor closes so much of the band, humidity and weather directly shrink IR range. The passive, no-warning advantage of IR is real, but it is emphatically not all-weather.
Two ways to see: reflected vs emitted#
The choice of window is really a choice between two sensing modes.
Reflective-band sensing (~0.4–3 µm) sees light reflected off the target — from the Sun, the Moon, or a laser. It carries fine spatial detail, but it goes dark the instant the illumination does. No source, no image.
Emissive-band sensing (~3–14 µm) sees the heat the target radiates on its own. It needs no external illuminator, so it works in total darkness. This is what lets an LWIR thermal imager form a picture on a moonless night.
There is a price for the emissive advantage: thermal sensing lives on contrast, the temperature difference between target and background. When that difference vanishes — cool skin against sun-warmed desert, for instance — the target can disappear entirely. That moment has a name, thermal crossover, and it is the subject of the next lesson.
Key Concept
Reflective bands see borrowed light and need a source; emissive bands see the target’s own heat and work in the dark. But an emissive sensor images only what stands out from its background — it runs on contrast, and at thermal crossover that contrast can go to zero.
Catching the photons: detectors#
A sensor that lives in a specific window needs a detector material tuned to that band, because the physics of absorbing a photon is band-selective. Silicon (Si) covers VIS/NIR; indium antimonide (InSb) covers the MWIR; and mercury cadmium telluride (HgCdTe, universally called MCT) covers the LWIR.
Detectors fall into two families with opposite trade-offs:
Photon (quantum) detectors free one charge carrier per absorbed photon. They are fast, sensitive, and band-selective — but warm, they drown in their own thermally generated noise. To reach their best-case performance, where the limit is the scene itself rather than the detector (the background-limited, or BLIP, regime), they must be cooled, often to about 77 K by liquid nitrogen or a Stirling cooler.
Thermal detectors, such as the microbolometer, sense the tiny temperature rise from absorbed energy. They are broadband and uncooled, which makes them cheaper, smaller, and lighter — at the cost of being slower and less sensitive.
That split — cooled, high-performance photon arrays versus uncooled, affordable microbolometers — shapes nearly every fielded IR system, from a cued missile seeker down to a handheld thermal monocular.
Where this is headed: the heat-seeker#
The weapon that turns this physics into a threat is the IR-guided missile — passive, so it gives the target no radar warning, and fire-and-forget once it is off the rail. Early reticle (spin-scan) seekers spin a patterned disc across the incoming IR, chopping it into a signal that encodes the direction of the brightest hot spot in the field of view. That simplicity is also the weakness: present a brighter source and the seeker chases it, which is exactly what a flare does. Modern imaging IR seekers read out a focal-plane array instead, forming an actual thermal picture and tracking the target’s shape — enough to tell an aircraft from a decoy.
L23 takes seekers, IRST, and FLIR apart, and L24 covers the countermeasures. For a penetrating bomber, this is the threat an RF jammer cannot reach.
Type-along
Put the lesson’s five sources through both rules in one pass. Type this in MATLAB — a temperature vector, Wien’s peak wavelength, the Stefan-Boltzmann total, and the window each source lands in:
T = [5800 1000 800 325 290]; % sun, plume, tailpipe, skin, bkgd
lam = 2898 ./ T; % Wien peak [um]
M = 5.670e-8 * T.^4; % Stefan-Boltzmann total [W/m^2]
win = strings(size(T));
win(lam < 0.75) = "visible";
win(lam >= 0.75 & lam < 3) = "MWIR edge";
win(lam >= 3 & lam <= 5) = "MWIR";
win(lam > 5 & lam < 8) = "H2O wall";
win(lam >= 8) = "LWIR";
for k = 1:numel(T)
fprintf('%5.0f K %6.2f um %s\n', T(k), lam(k), win(k));
end
fprintf('plume/skin power ratio = %.1f x\n', M(2)/M(4));
Then answer:
The plume peaks at 2.9 µm, outside the 3–5 µm window. Why hunt it in the MWIR?
The plume runs only about 3× hotter than the skin. Why is it about 90× brighter?
Solution
5800 K 0.50 um visible
1000 K 2.90 um MWIR edge
800 K 3.62 um MWIR
325 K 8.92 um LWIR
290 K 9.99 um LWIR
plume/skin power ratio = 89.6 x
Wien locates only the peak of a broad curve, and the curve is wide. Integrated across the band, the MWIR still collects about 36% of the plume’s total emission against about 9% for the LWIR — and the atmosphere is open there. The peak points you at a window; the in-band fraction decides what the sensor actually collects.
Stefan-Boltzmann: total exitance goes as \(T^4\), so \((1000/325)^4 \approx 90\). The fourth power is why a small patch of hot metal can dominate the IR signature of a whole airframe.
Quick problem — Nozzle and Skin at Altitude
Two sources on one notional aircraft at cruise altitude, where the thin, cold air holds the skin far below the reading’s 325 K.
Engine nozzle: \(T = 725\) K
Skin, cold-soaked at altitude: \(T = 250\) K
Treat both as blackbodies
Windows: MWIR 3–5 µm, LWIR 8–12 µm
(a) Each source’s Wien peak, and the window it lands in.
(b) Per unit area, how many times more total power does the nozzle radiate than the skin?
(c) Bypass-air mixing cools the nozzle to 610 K. Where does its peak move, and by what factor does its total output fall?
Solution
(a) Wien’s law, \(\lambda_{\text{peak}} \approx 2898/T\). Nozzle: \(2898/725 =\) 4.00 µm, inside the MWIR window. Skin: \(2898/250 =\) 11.6 µm, inside the LWIR window, near its long-wave edge.
(b) Stefan-Boltzmann: total exitance goes as \(T^4\), so the ratio is \((725/250)^4 = 2.9^4 =\) 70.7. Per square meter, the nozzle radiates about 71 times what the skin does.
(c) The new peak is \(2898/610 =\) 4.75 µm — 0.75 µm longer, but still in the MWIR window. The output falls by \((725/610)^4 = 1.189^4 =\) 2.00: cooling the nozzle by 115 K, about 16%, halves it.
So what: temperature sets how much far more steeply than it sets where — a 16% cut halves the nozzle’s output but leaves its peak inside the same 3–5 µm window, so cooling dims a hot part without moving it out of an MWIR sensor’s band.
Wrap-Up#
Every warm object radiates, and its temperature does two things at once: it sets how much it emits (Stefan-Boltzmann, \(T^4\)) and where that emission peaks (Wien, \(\lambda_{\text{peak}} \approx 2898/T\)). Hot parts land in the MWIR window (3–5 µm), cool parts in the LWIR (8–12 µm), and the water-vapor wall between 5 and 8 µm plus the CO₂ notch near 4.3 µm decide what the atmosphere will let through. Reflective bands borrow light; emissive bands see the target’s own heat and run on contrast. And the sensor is a choice too — a cooled photon array for sensitivity, an uncooled microbolometer for cost and size. Next, L22 — IR Emitters breaks a real airframe into its radiating components and finds the twice-daily moment of thermal crossover when it hides in its own background.