CHAPTER XXVI

MAP MAKING

=Technique of Negative Making.=—Stated in its simplest terms, the whole problem of making a photographic map from the air consists in taking a large number of slightly overlapping negatives, all from the same altitude, with the plane flying uniformly level. When trimmed and mounted in juxtaposition, or pasted together so as to overlap in their common portions, the prints from these negatives constitute a complete pictorial map. There is thus furnished by a few hours' labor topographic information which would be the work of months to obtain by other means.

The making of map photographs involves all the special technique of spotting, with much in addition. The pilot's task is not merely to go over one object; he must navigate a narrow path, at a constant altitude, on an even keel. If he is to make not merely a ribbon, but a map of considerable width, he must take successive trips parallel to the first, each displaced just far enough from the previous course to insure that no portion is missed—a difficult task indeed.

It is the observer's duty to so time the intervals between exposures that they overlap enough, but not so much as to be wasteful of plates or film. He must also change magazines or films so quickly as to miss no territory, or if some be missed, his is the task of directing the pilot back to the point of the last exposure, where they begin a new series.

Level flying is entirely a pilot's problem. Its importance will be realized when we consider the accompanying diagrams (Figs. 134 and 135), where the effect on the resultant picture is shown of climbing, gliding, or banking to either side. Prints from negatives distorted in this way neither will be true representations of the territory photographed, nor will they match when juxtaposed. In fact, they can be utilized only if special rectifying apparatus is available for printing. Flying at a constant altitude is similarly necessary if the prints are to be utilized without enlargement or reduction in order to make them fit.

Assuming a skilled pilot who will do his part, the next step is to calculate the exposure intervals in order to insure an adequate overlap. If a negative lens is installed which has been marked with a rectangle the size of the camera field, the simplest method is to estimate the proper instant for exposure by watching the progress of objects across the lens face. This of course requires constant attention, and it is easier to do this only occasionally, in order to determine the ground speed in terms of camera fields traversed per minute. Thereafter exposures are to be made by time, as determined by a watch or clock. Any desired degree of overlap can be chosen, and either estimated, or more or less accurately fixed by lines marked on the negative lens at a shorter distance apart than the edges of the field. The most usual overlap is 20 per cent., except for stereos, which call for 50 to 75 per cent.

In the absence of a negative lens or some other sight to show the whole camera field, it is necessary to resort to calculation from the speed and altitude of the plane, the focus of the lens and the dimensions of the plate. If _A_ is the altitude, _a_ the focal length of the lens, _d_ the diameter of the plate in the direction of travel (usually the short length is chosen for economy of flights to cover a given width), _f_ the fractional part by which one negative is desired to overlap the next, and _V_ the ground speed of the plane, then we have, by simple proportion, that the interval between exposures, _t_, must be—

_Ad_(1 - _f_) _t_ = ————————————— _aV_

If _A_ = 2000 meters, _d_ = 18 centimeters, _f_ = ⅕, _a_ = 50 centimeters, and _V_ = 200 kilometers per hour, this relation gives—

2000 × .18 × .8 × 3600 _t_ = —————————————————————— = 10.3 seconds .5 × 200,000

The principle of overlapping map exposures is shown in the accompanying diagram (Fig. 129), together with data calculated as above for a 4 × 5 inch plate.

It is particularly to be noted that it is the _ground speed_ of the plane that is used. This may be calculated by knowing the air speed and the wind velocity and direction. Fig. 136 shows the method of doing this graphically. First an arrow is drawn representing the direction it is desired to fly. Next a second arrow is drawn of length to represent the wind velocity. This must be inclined toward the first arrow in the direction of the wind, and its head is to touch the head of the first arrow. Then with the farther end of this second arrow as a center, describe a circle of such a length as to represent the air speed of the plane, in the same units as the wind velocity. Connect the point where this circle cuts the arrow of flight direction to the center of the circle by a straight line. This line constitutes the air speed arrow, giving the direction it is necessary to fly, at the given air speed, to make the course desired. The length of the flight direction arrow between its head and its point of intersection with the air speed arrow gives the ground speed.

When the wind is ahead or astern this calculation reduces to the simple subtraction or addition of the wind velocity to the air speed of the plane. Whenever possible, mapping should be done up and down the wind (Fig. 137). If the plane is “crabbing,” the above calculations for overlap are only valid if the camera can be turned normal to the direction of travel over the ground. If the camera cannot be so turned the corners of the successive pictures overlap instead of their sides, with quite unsatisfactory results (Fig. 138).

Calculation of the distance apart of the parallel flights necessary to make a map of any width is done by the use of a formula similar to the longitudinal overlap formula above, distance figuring instead of time. Using the same symbols, and denoting the distance by _D_, we have—

_Ad_(1 - _f_) _D_ = ————————————— _a_

With the same figures as before, but substituting 24 centimeters for the plate dimension, this relation gives—

2000 × .24 × .8 _D_ = ——————————————— = 768 meters .5

It is of course largely a pilot's problem to steer the plane over parallel courses at a given distance apart, although the observer, noting conspicuous objects through a properly marked negative lens, may direct the pilot by any of the means of communication already mentioned.

An alternative method of securing parallel strips, which is to be highly recommended where enough photographically equipped airplanes are available, is for several planes to fly side by side, maintaining their proper separation (Fig. 139).

=Cameras and Auxiliaries for Map Making.=—Mapping can be done quite satisfactorily by hand operated or semi-automatic cameras, provided the observer has not too many other duties. On the other hand, the operation of exposing at more or less definite intervals of time, irrespective of the object immediately presented to the camera, is a largely mechanical one. It naturally suggests the employment of an automatic mechanism, whose speed of operation only is it necessary to watch.

If a non-automatic camera is used the timing of exposures may be done by watching a negative lens, as described above, or by reference to a clock, assuming that the ground speed is known through calculation. A very practical advance over the ordinary use of a clock is to attach a stop-watch to the shutter release, so that it is turned back to zero and re-started at each exposure (Fig. 70). In passing, it may be noted that if the stop-watch hand makes an electric contact which throws the shutter release, then the device constitutes an attachment for turning any semi-automatic camera into an automatic. The most suitable cameras for mapping are unquestionably those of the entirely automatic type. The use of such cameras always demands a knowledge of the ground speed. This demand has led to many suggestions for _ground speed indicators_. The common idea of these is to provide a moving part on the plane—either a disc of large diameter, or a chain, or a revolving screw—whose speed may be varied until any point upon it appears to keep in coincidence with a point on the moving landscape below. The ground speed is then to be read off a properly calibrated dial. Or, as a further step, the frequency of the exposures may be directly controlled by the ground speed indicator mechanism. The entire control of the camera would then consist merely in occasional adjustment of the ground speed indicator.

While entirely possible in theory, these devices are not easy to work with in practice, because the plane is always subject to some pitching and rolling, which make it difficult to hold any object constantly on the moving point. This is especially true at high altitudes, where the apparent motion of the earth is quite slow compared to the swervings of the plane. This objection is in part removed if the ground speed indicator is carried by a gyro stabilizer.

Ordinary mapping does not demand such exquisite rendering of detail as does trench mapping. Nor is it necessary to fly in peace-time at such high altitudes as in war. In consequence, mapping cameras are preferably of the short focus, wide angle type, say, 25 centimeter focus for an 18 × 24 centimeter plate. Film is to be preferred over plates because of the greater number of exposures it is possible to make on a flight. The shutter of the mapping camera must be extremely uniform in its rate of travel so that the elements of the map may match in tone (Fig. 140). A mount which permits the camera to be turned normal to the direction of flight, such as the British turret mount (Fig. 87), is particularly desirable if flying across the wind is necessary, as will often be the case in mapping strips between towns or between flying fields. Devices to indicate compass direction and altitude are called for in new and poorly mapped territory, and may be expected to receive intensive study in the future. The question of their utility is, however, bound up with the whole question of the sphere of aerial photographic mapping. Up to the present this has been almost entirely a matter of filling in details on maps obtained by the regular surveying methods, or of making pictorial maps for aviators. To what extent primary mapping can be done by the airplane is yet to be determined.

At this point mention must be made of special cameras for securing extremely wide angle views, thereby minimizing the number of flights. The _Bagley camera_, devised by Major Bagley of the U. S. Engineers, is an example. It has three lenses, a middle one pointing directly downward, and one to either side at an angle of 35 degrees. The pictures obtained with the side cameras are of course greatly distorted, and must be rectified in a special rectifying camera. The resultant definition is not good, but as the maps are made on a much smaller scale than the original pictures, this is not a serious objection. It is a matter for the future to decide whether the additional labor on the ground necessary for the rectifying process is to be more expensive than the extra flights which must be made with the ordinary types of cameras covering a smaller angle.

=Printing and Mounting Mosaics.=—With an ordinary set of overlapping negatives the first step toward producing a map is to _scale_ the negatives. For this purpose one should be selected which by comparison with a map shows no distortion, and which is on the desired scale, or is known to have been made at the average altitude of flight. A sketch map of the territory should then be drawn, on this scale, based on available maps. This sketch is preferably made on a large ground glass illuminated from behind (Fig. 141). On this all the negatives should be laid, and their proper relative positions sought. When this is done it is evident at once whether all the territory has been covered, and whether there are any superfluous negatives. Each negative should then be examined as to its scale and distortion. If it can be made to fit the scale by simple enlargement or reduction, a line can be drawn on one edge of a length indicating its scale. This line will later be used as a guide in the enlarging camera. If the picture is badly distorted it must either be replaced by another negative, or if rectifying apparatus is available, it must be set aside for the making of a rectified print.

The next step is to make prints from the negatives, which may be done either by contact, or, necessarily if differences of scale must be compensated, in the enlarging camera. If prints to an exact scale are required the shrinkage of the paper must be determined and allowed for. The prints must all show the same tone, and must be uniform from edge to edge. If the focal-plane shutter is not uniform in its travel, as is frequently the case, this means that the print must be “dodged,” or exposed more at one edge than the other, by locally shielding the plate and paper during exposure. A case of the step-like effect caused by uneven shutter action is shown in Fig. 140. The effect due to uneven shutter action is of course absent with a between-the-lens shutter, which constitutes a strong argument in favor of that type for use in mapping cameras.

When the prints are made they must be mounted together on a large card or cloth background. For a very small mosaic they may be juxtaposed by simple examination, matching corresponding details in successive prints. For a mosaic of any size an accurate outline map must be drawn on the surface to which the prints are to be attached. The prints are then laid out on this outline, moved to their correct positions, and held down by pins (Fig. 142). When they are all arranged the final mounting may be begun. The excess paper, beyond what is necessary for safe overlaps, may be trimmed off, exercising judgment as to which print of each adjacent pair is of the better quality, and utilizing it for the top one at the overlapping junction. If one print shows serious distortion it may be placed under its fellows on all four edges, thus minimizing its weight. The edges are best made irregular by _tearing_. Straight edges are apt to force themselves on one's attention in the final mosaic and give an erroneous impression of the existence of straight roads or other features. Both forms of edging are shown in Figs. 124 and 143.

An alternative method of securing the final print mosaic, where film negatives are used, is to trim successive film negatives so that the trimmed sections will exactly juxtapose, instead of overlap. The sections are then mounted, by stickers at their edges, on a large sheet of glass, and printed together. Captured German prints show that this was the method commonly used with the German film camera (Fig. 62).

It will be noted that the procedure which has been described and illustrated by Figs. 142 and 143 assumes the previous existence of a map accurately placing at least the chief features of the country covered. This draws attention at once to the limitations and true sphere of aerial photographic mapping at the present time. With the cameras thus far it is not possible, nor is it attempted, to do primary mapping of unknown regions. Distortions due to lens, shutter, film warping and paper shrinkage considerably exceed the figures permitted in precision mapping. From the standpoint of geodetic accuracy the cumulative errors of deviations in direction, altitude and level, peculiar to flying, would soon become prohibitive.

The great field for aerial photographic mapping in the near future lies in filling in detail on maps heretofore completed as to general outlines, or, as in the war, on maps far out of date. The war-time procedure in country largely unknown, such as Mesopotamia, was probably closely that which will be necessary in peace. Conspicuous points in the landscape were first triangulated from friendly territory, and from these the outline map was drawn, whose details were to be supplied by aerial photographs. Much of the “mapping” of cross country aerial routes so far done is frankly of a pictorial nature, showing conspicuous landmarks and good landing fields—extremely valuable and useful, but not to be confused with precision mapping. In assembling mosaics of this kind the elaborate procedure described above is not followed. The process is the simple one of juxtaposing adjacent prints as accurately as possible by visual examination. Errors are of course cumulative, but as long as exact distances are not in question this is no matter.