To rely entirely on any one method of navigation is dangerous. In order to obtain the best results under all conditions it is necessary to use a combination of all methods.
Use should be made of pilotage whenever landmarks or beacons are visible, but its limitations should be realized. Radio navigation, though a valuable aid, has not reached the point where it may be depended upon as a sole means of navigation, and it is very doubtful if it ever will.
Much stress has been placed upon celestial navigation, during recent years, by writers who have not had the opportunity to try its practicability in the air. Up to the present time actual use has not been made of celestial navigation on more than twenty flights. It is of great advantage in flights of 500 miles or over, which as yet are not common. The accuracy of four or five miles, which can be obtained under fairly good conditions by an experienced man, does not justify its use on short flights. As the speed and range of aircraft are increased, celestial navigation, if still further simplified, will assume greater importance.
Dead reckoning is unquestionably the prime requisite in air navigation. If carried out in the proper manner, the results that may be obtained are surprisingly accurate.
The Army Air Corps has devoted a great deal of attention to the equipment of planes and the training of officers in dead reckoning, radio, and celestial navigation. Two training units were formed, one at Rockwell Field, California, and one at Langley Field, Virginia. Officers were brought from the various Air Corps posts throughout the country to take the course, which consisted of blind flying and blind landing instruction in addition to the advanced navigation training. The ground schoolwork was followed up with practical training over water. Twelve Douglas amphibians, equipped with the latest blind flying and navigation instruments, were used for this work.
Most of the courses were flown with the pilot under a hood and the navigator curtained off so that he was forced to rely on his aperiodic compass and drift and ground speed meter. The navigator’s equipment consisted of a large aperiodic compass, ground speed and drift meter, air speed meter, air temperature meter, sensitive altimeter and kit containing necessary charts and plotting equipment. For celestial navigation, a Pioneer bubble sextant and two second-setting watches, one set to G.C.T. and one to G.S.T., were used. Precomputed altitude curves were found to be more practical than the use of line of position tables in the air.
All courses were set by the navigator by interphone communication with the pilot. Remote controls are now being fitted in these planes so that the navigator may set the pilot’s directional gyro for him, making conversation unnecessary.
On completion of the training, the average error in maintaining a course, without reference to landmarks, was 0.5 degree. This means that as long as the surface is visible a plane would be no more than 0.5 mile off course in a distance of 60 miles. In other words, on a flight of 2,400 miles to Honolulu the error in dead reckoning would only be 20 miles.
In flying a course in this manner there are seven possible causes of error. The first error may be caused by the true course not being measured accurately from the chart. The mean variation must be carefully picked from the chart for each leg of the course to be flown. It is very often difficult to judge the variation with any degree of accuracy between the roses, especially around the edge of a chart, where one cannot tell the trend of the isogonic lines. I have often wondered why the use of isogonic lines has not been standardized on all charts in preference to variation roses.
The next possible source of error is in the installation and swinging of the compass. This is a very common error where proper facilities are not provided for the purpose, and where the swinging is left to people who have no appreciation for the need of extreme accuracy and do not thoroughly understand the operation.
Should care not be exercised in the alignment of the drift meter during installation, the number of degrees it is out of line will result in a corresponding error in course. Each change in drift, due to varying winds, must be caught instantly or the effect on his course is apt to be serious.
Finally, the pilot must be kept right on his course and not allowed to deviate one way more than another.
In the practice of dead reckoning, the Army Air Corps has found it entirely impractical to use wind direction and velocity in the determination of drift and ground speed.
It is hard to understand why writers of textbooks on air navigation devote the major part of their books to examples of dead reckoning, intercept, and search based on the use of aerological reports. The wind is so variable in force and direction with changes of time and altitude that an assumption of a constant wind would lead to very erroneous results.
Even at low altitudes the direction and velocity of the wind on the surface, estimated from the whitecaps or other disturbance on the water, cannot be assumed for the altitude at which the airplane is flying, and may be grossly in error.
The first flight of one group at the Air Corps Navigation Training Unit gave them very conclusive proof of the fallacy of using wind direction and velocity. To give a comparison between the results obtained from using aerological data and that obtained from direct readings with a drift meter, they were instructed to calculate the drift before taking off, and to check it with the drift read on the meter. The planes came over the point of departure at an altitude of 600 feet and set out on the first course over Chesapeake Bay. The drift calculated from the aerological report was minus 6 degrees, that obtained from the drift meter was plus 5 degrees. On looking over the side of the plane the whitecaps showed a strong wind from the left which seemed to substantiate the calculated drift. This led most of the students to believe that their drift meter was error, so they applied a minus 6-degree drift. In a distance of 30 miles this caused them to finish up 5 and 6 miles off course.
On many other occasions comparisons were made between the drift calculated from the apparent velocity and direction of the wind on the surface and that obtained with the drift meter which further showed the necessity for direct readings for drift.
With practice both drift and ground speed can be determined under practically any conditions as long as the surface is visible.
It is easier to make an accurate course in a strong wind as the more disturbance there is on the water the easier and more positive are the drift readings. In one instance, during a "Santa Ana" off the coast of Southern California, a 23.5-degree change in drift occurred in 20 minutes; each change was noted, and on the 60-mile leg the error in course amounted to only 0.5 mile.
It is not generally known that sunlight provides one of the easiest means of getting drift and ground speed. The sunlight is broken up into a multitude of specks on the water, and these specks being stationary are very easy to use for ground speed and drift.
Experiments are being made for the determination of ground speed and drift at night by installing a light in the wing which illuminates a small area on the water directly below the drift and ground speed meter. The tests so far indicate that this is practical and are being continued.
Strict attention is paid to the preparation of log sheets before take-off, and during flight each change in drift and ground speed, steering course, altitude, and estimated time of arrival, is recorded. The estimated time of arrival (E.T.A.) at each turning point is calculated and changed with a change in ground speed, so that closed courses over water are accurately flown.
The question is often brought up regarding the surface movement of the water affecting the accuracy of the ground speed.
Having devoted a great deal of thought to the matter, and having had a considerable number of years of sea experience to draw from, the writer believes this error in most cases to be negligible.
Experiments with sea anchors have shown that even in the strongest gales the actual movement of the water does not exceed 1 mile or 1.5 m.p.h. The wave motion is transferred, but the whitecaps and spume are left behind. The wave motion can be seen from the deck of a surface vessel, but is not visible to an observer looking straight down.
Of course, excessive currents would introduce an error, but the regions where they occur are generally known and are limited in number.
The ground speed and drift meter used by the Army Air Corps was devised by the author and used on the round the world flight with Wiley Post in 1931. It is the only instrument up to the present time that will enable ground speed to be determined from the apparent motion of the mass of the water.
The basic principle of this device is the solution of similar triangles. A grid, in the form of an endless metal band, is caused, by clockwork mechanism, to travel across the field of view at a set constant speed. Through an eyepiece, adjustable in height, the apparent movement of the ground or water is observed through the openings in the metal grid. The closer the eyepiece is moved towards the metal band the faster it appears to move, so it is moved up and down until the movement of the band is synchronized with the apparent movement of the surface below.
The ground speed can then be readily determined by dividing the altitude above the water by the figure indicated on the counter alongside the eyepiece slide. The latest type instrument uses three constant speeds and three counters to enable it to be used for a wider range of altitudes and speeds. The latter is the model used successfully by Colonel and Mrs. Lindbergh on their recent 30,000-mile flight and is also being used by Admiral Byrd on his South Polar expedition.
The angle of drift is obtained by aligning longitudinal wires in the field of view with the apparent motion of the surface below. The use of two prisms gives a periscopic view enabling the instrument to be mounted inside the airplane.
The disadvantages of this device are that it is necessary to be able to see the ground or water and also make an accurate determination of the altitude above the surface.
If the barometric altimeter is corrected for temperature the error over water is slight, and, if one is in doubt, it is always possible to fly down close to the water and reset the altimeter.
Some day we will have an automatic ground speed and drift meter which will simplify the problem of navigation for both surface vessels and aircraft, but until that time we must make the best use of the means at hand and get every bit of accuracy that is possible out of our dead reckoning.