GPS & RAIM
It's common sense: make sure that your navigation equipment works before you fly. There are a few cases where this doesn't matter, like when you are flying locally using pilotage alone, but certainly if you are going to be flying at night, in the clouds, or over large bodies of water you want to know that the black boxes are telling you the right thing.
This is an old lesson. World War I ace and head of Eastern Air Lines Eddie Rickenbacker learned it the hard way, in 1942. This story comes from Seven Came Through, published by Doubleday in 1943.
It was the height of World War II, and Rickenbacker was on a secret mission to inspect all air combat groups. After a fact-finding tour of the European theatre, he headed to Hawai'i to begin inspection of the Pacific. His first destination was "1800 miles southwest" of Hawai'i; it was still wartime when the book was published, and Rickenbacker was a little coy about the destination. My best guess for the destination is Howland Island, which is famous because another prominent aviatrix, Amelia Earhart, never made it there, either.
The Army Air Corps assigned a B-17 and a crew to the flight, and they left Hickam Field in Honolulu on October 21, 1942. The departure didn't go well; a brake problem on the right main caused a groundloop.
Nobody was hurt, so the Army dug up another airplane and they launched again, crossing the ocean using dead reckoning and an octant. An octant works like a sextant, but uses one-eighth of a circle rather than one-sixth. The problem with dead reckoning over the ocean is that you can't determine drift or groundspeed, so you depend on the accuracy of the winds aloft forecast. We all know how that goes. But the octant and navigator were there for backup.
The octant had been loose on the navigation table aft of the cockpit during the groundloop, and had fallen. Damage from the fall made it inaccurate. With no way to find the destination island (whose direction finding equipment wasn't working), the B-17 eventually ditched. All eight made it into the rafts, but one crewmember died before they were rescued. Twenty one days later.
Only a few luddites use celestial navigation in airplanes anymore. I've tried, but every time I go up the hill and take sights I end up knocking on my neighbor's door and asking when did you move to Montana? Where is the lesson for us?
In the USA, the Federal Aviation Regulations (14CFR) require regular checks of pitot-static (14CFR91.411) and VOR navigation systems (14CFR91.171). The magnetic compass should be checked with each annual inspection. We all know people who don't do the checks, and I can't tell you with a straight face that skipping a VOR check will make you spend 21 days in a raft.
Besides, I hear people say, nobody uses that stuff anyway. Now it's GPS, GPS, and when that fails GPS. How do you check the GPS?
First, let's get this out of the way: GPS can fail. Satellites have broken down. The military test jamming equipment. Incorrect almanacs have probably been uploaded. Ionospheric storms hit with shocking regularity and little or no warning. Antenna connections break. Voltage spikes fry avionics. Batteries leak in handhelds. The system is guaranteed to be 99.999% ("five nines") reliable, but there are 100,000 North Atlantic crossings every year, so statistically one of them will have GPS failure.
What do we do about GPS failure?
The big expensive answer is WAAS, which I will leave to a future blog entry.
The simplest answer is that we cross check against other systems, usually pilotage and VOR/DME. (When I flew turboprops I carried WAC charts, partially for sightseeing and partially "just in case.")
The newest answer is you are required to check GPS integrity before every flight using RNAV (DPs, STARs, Q- and T-routes). This is called a RAIM check (Receiver Autonomous Integrity Monitoring). It requires at least five satellites (and barometric altimeter input). See this post for an explanation.
The requirement has always been there, but it was scattered through a lot of supplemental documents like TSOs and operating manuals. Now the FAA has clarified the requirement, in a new issue of Advisory Circular 90-100. They have even given you a website to help you do it. If worse comes to worse, you can get the information from Flight Service.
Even if it's not required for your operation, it sounds like a good idea.
Modesty aside, I feel strongly enough about the study of the art of navigation that I wrote a book: Understanding Mathematics for Aircraft Navigation. Some people say that it contains more than pilots need to know. I've already said how I feel about that...