# How does an atmospheric probe measure temperature during descent?

I heard that the Galileo spacecraft sent Galileo Probe into Jupiter, and it reported temperatures reaching 153 degrees Celsius.

The question is, when you send a probe crashing into a planet that has an atmosphere (and gravity), wouldn't it heat up? So how would you actually measure the temperature of surrounding atmosphere?

• Yes, the heat shield got incredibly hot. However its job was to insulate the instruments from that temperature, which it did. It was thrown off after its job was done, leaving the relatively cool science instruments to descend on a parachute and do their job. If the heat shield were left on, then the heat would in fact soak back and increase the temperature of the instruments, which is one of the reasons it was thrown off. (Also because it was heavy, and would have increased the descent rate on the parachute.) – Mark Adler Apr 24 '15 at 15:55
• As an additional sub-question, I wonder if the temperature can be measured at some distance from the probe rather than within the immediate surrounding of the sensor. – mins Apr 25 '15 at 17:44

As described in extensive detail by Al Seiff and T.C.D. Knight in their May 1992 paper in Space Science Reviews, Vol 60; The Galileo Probe Atmosphere Structure instrument (which can be read without paywall by clicking "print this article"), the Galileo Probe measured atmospheric temperatures in Jupiter's troposphere with a platinum resistance thermometer (PRD). Unlike a thermocouple, a PRD generates no significant electrical potential (voltage) on its own. Instead, the instrument electronics puts a small voltage across a very thin platinum wire. The resistance of that platinum wire depends on its temperature. The instrument measures the current flowing through the wire as a result of the applied voltage. That can be converted to the wire's resistance (via E = IR), and from that to its temperature. Since a current flow can heat the platinum wire, the applied voltage is very low to keep the current and thus the heating low. But the wire is VERY thin, so any heat generated by the current flow quickly dissipates into the (rapidly) flowing atmosphere around the sensor. The thin wire also aids in coming quickly to equilibrium with the surrounding atmosphere, so there is little "thermal lag".

The radiometer on the Galileo Probe was to measure, as a function of depth in the atmosphere, how much radiant energy (sunlight) was propagating downward in Jupiter's atmosphere, vs. how much radiant energy was propagating upward (thermal radiation to space). This allowed atmospheric scientists to measure where in Jupiter's atmosphere the sun's energy was being deposited, how much energy was coming up from the interior, and then, correlating with the wind speed measurements of the Doppler Wind Experiment, determine whether Jupiter's winds are driven largely by solar energy or by energy welling up from the interior.

From the article:

After entering Jupiter's atmosphere at 170,700 kph (~47 Km/s) it used its ablative heat shield to decelerate. After 2 minutes, and reaching deceleration levels of 230g, it was traveling at only 430 kph. It then deployed its parachute, and descended 200 km during the next 57.6 minutes, at an average speed of 208 kph.

There are several ways of measuring the temperature of a gas stream; I am still investigating the design of the temperature sensing system, and will update this section with the results. Although it can be said that when the gas stream is of low velocity, a viable method is to use a thermocouple.

Thermocouples have a wide range of capabilities. Of course any effective temperature gauge quickly reaches the temperature of its surrounds, otherwise it would not be very useful.

Thermocouples work by generating a small electrical voltage from the junction of two different metals, called a bimetal junction.

As mentioned, updates to follow. Also references.

• The Galileo probe used an infrared radiometer, not a thermocouple. One of the most expensive parts of the Galileo project was a 200-carat diamond which was cut to make the window for this radiometer. I haven't been able to figure out yet how a radiometer can be used to measure atmospheric temperatures. – Hobbes Apr 24 '15 at 19:04
• @Hobbes Thanks for update. Still working on it ... – user8406 Apr 24 '15 at 23:45
• Haven't finished - I believe the radiometric technique measures the spread in the emission spectrum caused by the random spread of velocities in the gas. Not confirmed yet though. – user8406 Apr 28 '15 at 14:02
• Thermocouples generate a small voltage, the voltage depends on the temperature difference to be measured. The current flows if the circuit is closed. The current depends on the input resistance of the thermocuple amplifier connected to the thermoelement. The tables for different types of thermocuples list the voltage, not the current. – Uwe Nov 18 '16 at 13:27

With regard to the sub-question about measuring temperatures at some distance from the probe, the answer is a qualified yes; see below. Measurements made essentially in contact with the thing of interest, such as measuring the temperature of atmospheric gases with a sensor in contact with the gases, are called "in situ" measurements. Measurements made at a distance by means of something, usually electromagnetic radiation of one sort or another, propagating from the location of the thing to the location of the sensor, are called "remote sensing" measurements. You're asking if there are any remote sensing techniques that could be applied to the atmospheric entry probe, right?

Indeed there are, but you sacrifice both accuracy of the measurements and a lot of the mission's instrument mass allocation. The technique most often used for remote measurements of temperature is infrared radiometry. Or you might use microwave radiometry, if you're measuring really low temperatures. But this technique is most useful for measuring temperatures of solids. At infrared wavelengths most solids have emissivities very close to unity, so they behave pretty much as blackbody radiators. If you measure the intensity of the blackbody radiation being given off by an object at multiple wavelengths (at least two, preferably more) the shape of a blackbody spectrum will tell you what the temperature of the object must be to fit the pattern of the measurements.

As an example application for measuring the temperatures of solids, inexpensive hand-held radiometers that use this technique are commonly available in hardware stores. I have one at home. They are definitely cool, definitely useful when the home air conditioning is doing something wierd, and are moderately accurate. Note: they are not very accurate! For various materials the infrared emissivities do vary a little across the IR band, and this affects the assumption of an ideal blackbody radiator. So with one of these devices you can get temperatures typically accurate to 1 or 2 C, but not 0.1 C.

Remote measurements of temperatures of gases is a whole different ball of wax. The emissivity spectra of gases are wildly non-blackbody and are very different for different gases, so the assumption of a blackbody spectrum is invalid. It is possible to make radiometer measurements of IR intensity and infer the temperature of a mixture of gases, if you know exactly the composition of that mixture of gases. If you do know that composition, then you can make the radiometer measurements at carefully chosen wavelengths and back out the temperature. But the radiative behavior of gases is somewhat akin to the conductivity behavior of semiconductors: it only takes a tiny amount of something to change their behavior tremendously, especially at a single wavelength. So if you're trying to make radiometric measurements of a giant planet atmosphere's temperature, and you assume all the usual constituents in that atmosphere (hydrogen, helium, methane, water, ammonia, hydrogen sulfide) but if it turns out there's also a bit of phosphine, or carbon monoxide, or hydrogen chloride, or some such, then the radiometer's calibration is no longer valid and you can't trust the inferred temperatures. All it takes is one unexpected constituent in the atmosphere, with an IR emission line on top of or near one of the ones the radiometer uses, to drive the uncertainties in the retrieved temperatures through the roof, making them useless.

Another reason to use in situ measurements on an entry probe is that they are very low in mass and power use. Radiometers are more massive and use more power, and those are precious commodities on an entry probe mission.

If some of the science objectives you're trying to achieve at a planet are of such high priority, and are so difficult to do remotely, that you're sending an entry probe to do the required measurements, then in situ measurement of temperatures is the hands-down winner in that trade.

The NASA Deep Space Network uses a microwave radiometer at its big ground stations to measure not the temperature of the local air, but the amount of water vapor in the local air. The amount of water vapor in the air affects its refractive index, and this makes a difference in measurements of the distance to spacecraft they are tracking, of concern especially when they are trying to measure the gravity field of the planet where the spacecraft is. This kind of water vapor measurement is possible because we know the composition of Earth's atmosphere to a gnat's eyelash, with the primary variable quantity being water vapor.