# Spacecraft Horizon Sensors for Lunar orbit

Recall that "traditional" horizon sensors for Earth missions utilize the IR wavelength regime since this can reveal a good contrast between the cold space and the warm Earth-limb edge.

If we were to now be orbiting the moon, would there be a better set of wavelengths that can provide discernment of the moon horizon? Without the presence of an atmosphere, I was thinking of the UV regime since it is much more reflective off of the moon, whilst Earth's ozone absorbs most of this energy. Is there feasibility (capability) of a sensor with UV "vision"? Is there a better set of wavelengths to look out for?

At the temperatures of the lunar surface IR limb sensors would work just fine. (See the post by @BobJacobsen) The problem with a UV sensor is that it requires reflected UV. Relatively little UV energy is emitted from the moon, either by thermal emission or by impacts of natural radiation. This means that if the limb the instrument would see is not illuminated by the sun, you don't get much signal at all. At IR wavelengths the moon radiates plenty from thermal emission, with a nice, cold (3 K) background, so there's plenty of contrast for locating the limb without it being illuminated.

• "At the temperatures of the lunar surface IR limb sensors would work just fine." How do you know this? Can you add a source where this can be checked? The surface of the moon gets pretty cold after ten days of being exposed to Space without sunlight. What wavelength range are IR limb sensors sensitive to?
– uhoh
May 4 '18 at 20:07
• Wikipedia gives surface temperatures that range from 100K to 390K at the equator. At 100K there will be little IR and what there is will be at very long wavelength, so unless the sensors you are describing can pick these up, this limb will be invisible and undetectable.
– uhoh
May 4 '18 at 20:14
• The CIRS instrument on Cassini could measure temperatures all the way down to ~35-40 K, but of course they had a couple of fairly long-wavelength sensors. If you need to get a signal from a 100K object with a reliable SNR you wouldn't need CIRS's longest wavelengths. But you'd need a mid-IR sensor at one focal plane, and you'd probably also need a radiator to keep that focal plane cold. But it could be done, at not too much expense. May 4 '18 at 21:34
• A limb detector would have to recognize the location of a hot limb and a cold limb at the same time. Edge detectors don't normally have a dynamic range of 10^4 (emission of 100K vs 300K at 10 microns for example). Are you sure "IR limb sensors would work just fine" is really correct?
– uhoh
May 5 '18 at 5:09
• See the post by @Bob Jacobsen above. May 5 '18 at 5:16

Following up on @TomSpiker's answer, Lunar Prospector carried an IR lunar limb sensor made by Ithaco. It's described in the Lunar Prospector Mission Handbook as (page 4-15)

4.2.5.3 Earth/Moon Sensor (EMS)

The Lunar Prospector Earth/Moon Sensor (EMS) is manufactured by Ithaco Space Systems. The EMS (P/N P108SA12), which consists of optical elements, infrared filters, a detector/sensing element, and associated electronics, provides electrical signals representing the 30 to 100 micron radiometric profile of objects (earth, moon and sun) as the sensor optics/detector scans across them. The scanning motion is provided by the angular rotation of the Lunar Prospector spacecraft. The analog output signal, representing the radiometric profile, goes to the C&DH electronics where threshold detection defines the leading and trailing edge of the scanned body. The leading and trailing edge threshold transitions are time-tagged and telemetered to the ground for further processing to (a) determine the angle between the LP spin axis and the nadir vector to the scanned body and, in conjunction with data from the sun sensor, (b) determine the dihedral angle between the sun and the scanned body.

There's also information about power, size, launch configuration, etc in that document.

Via $$\lambda_\rm{peak}$$ (microns) = $$2900/T$$ (K), the "30 to 100 micron" IR wavelength band correspond (as peak emission) to temperatures down to about 100 to 30K, well below lunar surface temperature even at dawn.

Incidentally, if you want to see the computations needed to convert the sensor data into attitude information, there's a description starting on about page 34 of the Lunar Prospector Ground System Software manual.