The subject of patrol plane navigation is still in its infancy. A few people know something, but no one knows everything about it. Pensacola, with its crowded schedule, can give only a faint idea to the student. It does not claim to do more. The proverb, “Experience is the best teacher,” comes true in this case, but experience in small plane navigation is of no value at all. Some day the tremendous difference in the problems of seamanship, engineering, actual control, and navigation should force the adoption of a new system, in which small plane and patrol plane pilots will be divided into separate categories. At present, when a pilot comes to a patrol plane squadron after 3 years in small planes he must start at the bottom.
In a patrol plane squadron, scarcely a week passes without some new idea or instrument being submitted for trial. Most of them are rejected, it is true, but enough of them stand the test so that patrol plane navigation is a constantly changing picture.
Since the shipboard officer may not know the conditions under which aerial navigation is worked, I would like to explain them. About 4 out of 10 celestial sights taken from a patrol plane show a mathematical error. In ground work this error does not exist; still, aviators are no better and no worse than other officers when it comes to mathematics. The reason for the error is that the aviator, although sitting at a navigation table similar to the one in a destroyer, is always conscious of the fact that he has at the wheel of the ship someone with less experience than himself. He is constantly on the alert and ready to rush to the cockpit. Moreover, he is wearing radio phones, and has static forever in his ears. He is interrupted repeatedly in his work by details concerning the plane, radio, messages, and voice communication. He is frequently above the limit of sufficient oxygen, and often he is cold. In addition, after 10 hours or more, he is tired. These things are unavoidable and necessary, but they explain the estimated percentage of error. I believe that in a 1,200-mile trip the ship navigator will spend about 10 hours on the job. The patrol plane navigator’s problem is harder, because he has only 10 hours altogether. At the end of that time he has arrived—we hope!
The system of navigation subsequently outlined is not entirely original. The elimination of mathematical errors and the reduction of the number of lines used to give a clearer picture of the situation are the main contributions. Due to excellent co-operation from the squadron officers, the system has been used constantly for over a year with complete success in Patrol Squadron 10. It has also been adopted by other squadrons.
Group Navigation
The system which I am going to outline is for navigation of a squadron. Usually there will be from 6 to 12 planes in the flight. Quite naturally, one may wonder why we make any distinction between single plane work and group work. The answer is that a single plane cannot do a thorough job of navigation, and attempting to do everything puts such a work load on the officer concerned that his efficiency in all of his work is impaired. It must be remembered that he is also the commanding officer of the plane, and with each new type of patrol plane that comes out his duties more closely approach those of any surface commanding officer. Flights lasting as long as 20 hours happen frequently. Since the human body can do only so much, we try to save the pilots whenever we can, let everyone contribute his share toward the navigation, and get superior results with less individual work. Furthermore, there is far less decrease in efficiency toward the end of the flight.
(1) Celestial navigation.—For this purpose the squadron is divided into 3 groups, regardless of the plane’s number, or the number of planes making the flight. The group into which a plane falls depends solely upon the position the plane is flying. Group I consists of all section leaders, including the squadron commander; Group II is all planes flying #2 position in the individual sections; Group III is all planes flying #3 position in the sections. With this arrangement, no matter what the last minute changes in the number of planes or position of the planes (and such changes frequently occur), there is no chance of a mix-up, there is always a fairly equal number in all groups, and the further advantages will be self-evident as the system unfolds.
The navigator, flying his own plane, issues all instructions concerning navigation over the voice radio. During the day he uses one group at a time for sun sights, usually taking them hourly. That is, individual pilots take a sight once every 3 hours. About one-half hour prior to taking the sights he issues instructions, e.g.: “At 2000 GCT, Group III observe the sun Lat. 21-21, Long. 157-58.” This position is his estimated dead reckoning at the time indicated. All Group III pilots then pre-compute the sight, meaning that they work out their computed altitudes prior to taking the sights, having all of the information that is necessary.
The actual taking of the sights commences 5 minutes prior to the designated time. Sights are taken as rapidly as possible for 10 minutes, 3 sights a minute being a good average. However, using an octant is an art in itself, and no matter how much one knows, only practice will get a good sight. The bubble is about as hard to corner as a volunteer for a weekend watch, particularly when the air gets bumpy.
In order to find out the altitude at the time for which the sight was worked, all sights must be plotted and the altitude picked off of the resultant curve. Incidentally, when picking off the curve, if your watch is slow, you subtract from the time you took the sight, instead of adding, as you do when working a sight. This point, obvious if given a moment’s thought, may correct many an error. If it were possible for an experienced observer to take a sight every second, the plot would look something like the track of a drunken, double-jointed snake on a dark night. However, there is a mean line that can be drawn through this maze which, surprisingly enough, is correct, at least within 10 miles—close enough for an airplane. Inexperienced pilots all seem to go through a stage where they get very nice looking lines, but miss their position by 30 or 40 miles. This is due to the fact that after the first sight is taken, the observer waits until the sun gets to about the same altitude and then marks. The octant is very obliging, and will read almost anything, if you wait long enough. The result is a beautiful line when plotted, but, as is so often true, it is beautiful but useless. He either has the peaks or the lows of his curve (usually the case) or some other particular portion. If the curve happens to be somewhat of a sine, he may be right—but I do not believe that anyone could consistently find the proper mean by observation. To moralize for the sake of emphasis—make the octant find the sun, not the sun find the octant. The observer now has his sight plotted, his intercept figured out, and is ready to report his results to the navigator. He reports only his azimuth and intercept, both to the nearest minute.
The navigator now enters the picture. As the reports come in, his first thought is to compare not the intercepts but the azimuths. The reason is that, since all pilots precomputed the sight with identical data, the azimuths must necessarily be identical. Must is the wrong word. Experience has shown that about 40 per cent of the sights turned in do not agree in azimuth. If two of the group do agree, however, the navigator knows where to point the finger. If not, he asks for a rework. There is only one answer, and that is that the intercepts turned in are not correct until the azimuths agree, because if they do not agree there is a mathematical error in the computation, but the navigator does not necessarily wait for a correction. If he feels that he does not have enough sights he can do one of two things: he can wait until the correction comes in; or he can obtain the observed altitude from the pilots who have mathematical errors, and the computed altitude from a pilot with the correct azimuth, and compute the missing intercepts. The latter is faster, but the former helps to eliminate that same error the next time. Having the sights that are necessary, the navigator averages the intercepts and ends up with a single line of position which is as faithful as a four-year-old car. However, before this system was adopted, I would have said seven-year-old.
For night observations, all of the procedure outlined above holds good. However, all 3 groups are used simultaneously, and a fix is obtained only every 2 hours (approximately). Using the Rude Star Finder, the navigator picks out a body bearing just about ahead for Group I. Group II, planes being to port in their sections, is given one 30 to 60 degrees to port. For the same reason Group III is given a body to starboard. The reason for using bodies in these positions is that for anything abeam there is apt to be wing interference, and for anything aft—if the observation is taken from the after part of the ship—there is far more vibration, and we can’t forget that elusive bubble. The information is then broadcast, and each individual body is handled in the same way that the sun was handled. The result is 3 simultaneous lines of position which are intended to make a small triangle, “the same as aboard ship.” Perhaps I am taking too pessimistic a vein, however, because this squadron usually has small triangles, and points with pride to a point fix after 17 hours in the air while en route from San Diego to Honolulu.
(2) Dead reckoning.—Since dead reckoning is the only form of navigation used in a small plane, the pilots come to boats knowing (usually) how to use one or another of the various types of aircraft plotting boards. Since opinions differ so strongly on the subject of which board is best, it seems the best policy is to let each pilot use his own particular type. The part the board plays in patrol plane navigation is small, being used mainly for drift sights and picking off headings. The main dead reckoning is, of necessity, carried on the charts, so the small plane pilot must even change his dead reckoning setup when he comes to big boats. In cases of bad weather, where constant changes of course are made, I use a board and then transfer to the chart.
The wind is naturally the controlling factor in dead reckoning, and even in this land of the trades it is about as dependable as a room boy’s laundry list. There are several methods of finding the wind. At low altitude (below 1,000 feet) the problem is simplified, because the surface wind will probably hold, and the good old seaman’s eye for force, combined with a bearing on the streaks for direction, is probably the best wind data that can be obtained at present. Man being a stubborn critter, however, will never believe that the wind is blowing quite as hard as it is—it just can’t do that to him! A drift sight combined with the seaman’s eye is consequently a very good idea. However, that is only a pipe dream. It would be fine if we habitually flew at low altitude, but there is a catch. Present planes are designed to be flown most economically at about 9,000 feet, and it will cost plenty in the way of speed and performance to take advantage of this simple method of finding the wind. Even the grid type sight fails at this altitude, and we are right back in the twenties, using the old post sight with a smoke bomb.
In this squadron, the navigator designates a section to take drift sights. The section leader takes a sight on course, his #2 plane goes off about 30 degrees to port, and his #3 goes about the same distance to starboard. The drift obtained, together with the course and speed when taking it, is reported to the navigator from each individual plane. The navigator then plots the sights, and with one hand on a rabbit’s foot and the other on the microphone, he announces the wind to the squadron. From then on the navigator requires a new series of sights upon change of altitude, change of weather, passing fronts, changing of drift on course (which he must check constantly), or, as is sometimes suspected by other pilots, just out of sheer cussedness. Knowing—or thinking you know—the wind, a plotting board will readily give ground speed after it is fed with temperature, air speed, and variation. However, mistakes are just as natural to make as it is natural to straighten a tie when talking to the commander on the phone.
There is just one more catch to dead reckoning, and that is missing some of the data. For this purpose we have forms for the man at the wheel to keep filled out.
Single Plane Navigation
And now we give our pilot the workout we were saving (?) him for. In the midst of group navigation it is entirely possible for a plane to get lost from the rest of the squadron in bad weather, in which case the navigation data, wind, position, etc., are all on the pilot’s desk, because the navigator has given him all data and position reports. The pilot who does not keep these data up to date is about as safe as a turkey the day before Thanksgiving. Or a plane may be operating independently. For group navigation there has been a minimum of 30-man hours for each 10 hours of flight. The pilot acting singly cannot hope to accomplish the accurate navigation that a group does, but there are certain things he can do to help in that direction. Naturally, he must have a zig-zag track after all of his drift sights, for there is no other plane to pace. Consequently, dead reckoning is a little more tedious than that of the group navigator. But what is to be done about that fatal 40 per cent error? Realizing that 4 out of 10 celestial sights he works are probably wrong, each sight must be worked entirely separately by someone else. The answer is that both second and third pilots must know how to work sights by your system. By comparing 2 sights worked with the same data it is easy to find the error. Harder on all hands, yes, but the percentage is too high to do otherwise. It is surprising the number of times there are errors after “double checking.”
For night sights, if you have 2 octants have the other one used on a second body simultaneously with your own to prevent running up sights as much as possible. For sun sights, have the second pilot take sights simultaneously with you and average the intercepts. Of course, the joker is that the second pilot must be trained to take good sights.
After position lines are obtained, it becomes a matter of judgment and experience as to how to use them. Some of the surface navigation rules hold, others may cause trouble. Briefly, one should not run up lines more than is absolutely necessary. In using single lines without any other information, a pilot will have to use the computed point, but he must be sure it is the computed point from his correct dead reckoning. Remember, when the sight was computed, the forecast on the dead reckoning was about as accurate as a forecast of tomorrow’s market. Courses, speed, and wind have changed since then. Erect a perpendicular to the new dead reckoning to get the correct computed point. Quite frequently a pilot has additional information which must be weighed and used with caution, such as a belief that if anything the wind was a little stronger, a little weaker, a little right, or a little left. Radio bearings give an indication. The distance from the radio station, time of day, condition of the air when bearings, sights, and drift sights are taken, the time since it was possible to take a drift sight, and the time of your last landfall all have a bearing on what you choose as the best position. No one can lay down rules telling just how much or when to use these factors, but it is surprising how quickly one can learn to evaluate them more or less correctly. Do not be afraid of large intercepts. They are probably right.
A final warning: Until a pilot has been burned repeatedly, he will underestimate distance from altitude. The old bow and beam bearings are still good.
And so it ends—for the present. There are still clouds below which prevent drift sights, clouds above to prevent celestial sights, and inaccurate radios for bearings. There is still bumpy air to ruin what little accuracy there is in drift sights, celestial sights, and radio bearings. Our drift sights are inaccurate, so our winds are inaccurate. The Guba (the plane used by Archbold on his expedition to New Guinea) has a better type of drift sight, but it still is far from perfect. Our octants are inaccurate, so our sights are inaccurate. A new octant came out a short time ago with improvements, but it is still far from perfect. So, after I go to small planes for 3 years and return to big boats, I, too, will start at the bottom. In the meantime, when the Captain says, “Mike, that was a good job!” I have to reply, “Captain, I was lucky!”—and I really mean it!
★
Senior officers in the old days were habitually very reticent regarding information concerning the destination of a vessel sailing for a new port. Officers of lesser rank were wont to hover about the binnacle to get some inkling from the course. It was not uncommon for a captain or navigator to take a sailing vessel out of the harbor and then turn to the admiral or captain or commodore as the case might be and inquire, “which way shall I cast her, Sir?”—Rear Admiral A. Farenholt (M.C.), U. S. Navy (Retired).