Sun Dogs (Parhelion) in Fargo, ND on 18 Feb 2009, from "gopherboy6956" on Wikimedia Commons
If you have an aircraft ice detector you may have wondered why sometimes it goes off in clear air with no evidence of icing. Gulfstream gets that question now and then and here's the answer. (If you are flying a Gulfstream keep in mind things break and if you are getting ICE DETECT in clear air fairly regularly, it could be a bad detector.)
A sun dog, by the way, is a parhelion and is defined as "a bright spot in the sky appearing on either side of the sun, formed by refraction of sunlight through ice crystals high in the earth's atmosphere."
The sources for the information which follows can be found below under references. I've added a few comments shown in blue.
Some operators have reported situations where the aircraft ice detection system is displaying an amber Ice Detected (G450/G550) or ICE DETECTED (GV) Crew Alerting System (CAS) message while in an area where no visible moisture is reported by the pilots. The flight crews are concerned about the accuracy of the ice detect system and the possibility of erroneous CAS messages.
The ice detection system was tested and certified using an aircraft equipped with atmospheric characterization sensors during certification test flights. These sensors allow for atmospheric conditions to be tracked and recorded exactly. Goodrich, the ice detect probe manufacturer, used this data to certify the ice detect probe to a level that was much more accurate than the visual atmospheric observations of the flight crew.
There are situations where the pilots will be hard pressed to visually determine icing conditions. One condition that is extremely difficult and sometimes impossible for pilots to identify is Low Liquid Water Content (LLWC). Recent flight testing experience on an instrumented aircraft has shown good correlation between the Ice Detected message of the ice detector and the existence of moisture as reported by other meteorological instrumentation prior to experienced flight test pilots acknowledging that an icing cloud was present. Researchers who perform tests in an icing wind tunnel can verify that it is (nearly) impossible to see an icing cloud, even when you’re standing inches away and specifically looking for it.
Another way to think about this is as follows: A cloud forms one particle at a time. As droplets condense onto one nuclei and then another, a cloud slowly starts to build. At some point, there are sufficient particles that have condensed for this “cloud” to finally become visible. Prior to it becoming visible, was the cloud actually there? Well, it depends on your definition of a cloud, but the moisture was there regardless. The number of droplets, their size, and spatial distribution all play a role in determining whether or not a cloud is visible, as well as the LLWC in that cloud.
Further validation of the fact that non-visible moisture can exist in the atmosphere is the existence of “sun dogs” or “snow dogs.” Sun dogs show up as areas in the sky that exhibit the “rainbow” spectrum and are generally located at the periphery of the sun. These sun dogs are caused by the scattering of sunlight off ice crystals or moisture that is present in the atmosphere. In many cases, these sun dogs occur on a day when there is not a cloud in the sky. Thus, a pilot would consider the atmosphere to be void of moisture due to the lack of clouds, even though the presence of the sun dog indicates otherwise.
It should also be understood that “liquid” water is not the only type of moisture that can cause ice accretion. Non-visible moisture in the form of water vapor (humidity) can also cause icing. NASA has data to substantiate that large amounts of ice can form under certain conditions due to high humidity. This has been corroborated through icing wind tunnel tests with the ice detector where frost can form sufficiently to cause the ice detector to activate the Ice Detected signal.
It is important to understand what pilots use as a reference for identifying moisture. It can be the existence (or not) of a cloud or visible moisture (in the form of water or ice) on various parts of the aircraft such as the wing, nacelle, windshield wiper, etc. In some cases, the location where ice forms on the aircraft is on parts not visible to the pilot. Both icing wind tunnel tests and natural icing flight tests have demonstrated the ability for the ice detector to detect ice without accretion on the wing/nacelle under some very specific icing conditions at warm temperatures and certain flight configurations. This can occur due to the differences in these geometries and complexities in the aerodynamics/thermodynamics of the ice accretion physics. The ice accretion that forms on aircraft surfaces near these conditions is nearly impossible to see in flight, as well as proven during recent icing flight tests. Thus, even “visible” moisture can cause ice to form on aircraft surfaces that is not always visible to the pilots.
It is also important to remember that there are small differences that exist between the left and right probes on the aircraft. Subtle differences can cause one side to detect ice and not the other. Some of those differences are air flow caused by rudder trim, the side of the aircraft that the sun is on, and even the variance allowed in the internal calibration of the sensor. Assuming that no damage has occurred to the ice detector, flight testing and icing wind tunnel testing have shown that the Ice Detected signal of the ice detector is highly reliable and repeatable, and is more accurate than human assessments.
Kelly, Jeremy, "G450 / GV / G550 (ATA 30): "Ice Detected" Message - More Than Meets the Eye," The Gulfstream Journal, 8/19/2016.