The inception of aerial photography as an aid to mapping in Haiti arose from the desire of the Travaux Publics to secure as quickly as possible an accurate map of the Arlibonite Valley. It is the largest single valley in Haiti, the most fertile, and the best adapted to irrigation as the largest river in the republic runs through one side of it. After bringing down silt for centuries, the river has gradually raised the land through which it flows until its course is now higher than the rest of the valley. The valley was intensively cultivated and partially irrigated by the French in colonial times. The long, straight and, in many cases, parallel roadways and section lines covering the valley today like a network, show the thoroughness or the colonists’ work. Now the valley lies dormant, grown to seed, yet with many little cultivated patches, scattered about here and there, with clusters of huts near by. There is a sharp contrast between the overgrown roadways with their jungle slothfulness, and the white man’s thrift and orderliness.
There are several methods of surveying territory by means of aerial photography. One is to estimate territory by oblique photographs which consists in exposing a succession of pictures obliquely over the side of the plane, and at the same angle and with parallel axes. From a computed scale the area covered by each picture is found. This method, for example, is used for estimating the amount of timber in a given area in heavily wooded sections of country.
There is another type of map which gives a fairly complete idea of a section of country but which is not absolutely accurate, made where large expanses of territory must be covered in a limited time. This map is made with a tri-lens camera. One of the lenses photographs vertically while each of the others takes a picture at an angle of about sixty-five degrees. All three are taken at the same time through the bottom of an airplane, with the axes of the lenses in a plane perpendicular to the surface of the earth. Fairly accurate vertical pictures are obtained from the two oblique negatives, and by means of a special printing machine, the tilt is taken out of the picture to a great degree. However, there is quite a little distortion in these two outside pictures, and when the prints are laid, successive strips do not match very well on this account. This camera has the advantage, in addition to its ability to take three pictures at a time, of covering a considerable area in each flight. The Alaskan Aerial Survey was made with this type of camera and furnishes a good example of the use to which it may be put.
Finally, there is a type of map which is produced by taking vertical pictures with a single lens camera of sufficiently long focus to practically eliminate all distortion. With this type of map it is possible to have an accurate scale, all prints match, even at the extreme edges, and the finished product is so exact as to satisfy the rigid requirements of civil engineers. It is easily seen from the above that this last type of map requires a greater amount of time to take and prepare than docs either of the other two.
It was this kind of map that the naval and civilian engineers required for their work in Haiti. The problem of obtaining it was rendered all the more difficult by the fact that no control data could be furnished as a means of checking the accuracy of the maps which were laid from the photographs. Thus the Republic of Haiti was a ripe field where aerial mapping could prove its value, as practically no maps existed, either of the plains or of the roads. Maps that could furnish contour lines, roadways, section lines, boundary lines of property to aid cadastral surveys, river courses, and other data were greatly needed.
Mapping by means of aerial photography is done by taking successive pictures of the surface of the earth as it passes beneath the airplane. In covering a given area, a number of lines of flight are necessary, each being made parallel to the last and at the same elevation as nearly as possible. The pictures are taken at such intervals during any one line of flight that each one covers half of the area in the preceding one, thus giving a 50 per cent overlap. This is a good average overlap, though more or less is often made, depending upon the mission. The successive lines of flight also should be flown so that a side overlap of 50 per cent is obtained in the resulting strips of prints. In taking the pictures, the observer makes an exposure, then waits a given number of seconds, depending upon the given altitude, speed of the plane and the wind velocity and direction, before making another, thus obtaining the required amount of overlap.
The photographic pilot and observer must act as a team if good results are to be obtained. Some work together is necessary for each to become thoroughly familiar with the other’s peculiarities. Many little difficulties arise continually, and if these are not mutually understood, and either anticipated or immediately corrected, they can easily prevent the mission from being successfully completed.
In the aerial mapping of Haiti the K-3 camera was used. It is the belief of the writer that this is the best mapping camera that has been developed up to the present time. It gives a clear-cut picture out to the very edge, which some other cameras do not do. The film used comes in rolls seventy-five feet in length and ten inches wide. Practically all of the work in Haiti was done at ten thousand feet, and at this elevation one negative on this film covers over a mile in length and a mile and a half in width. Several magazines may be carried on a single flight, or the film may be reloaded in the air, using only one magazine.
To obtain clear pictures at this, or any efficient working altitude, it is absolutely necessary to use a quality of film that allows the use of a filter so as to cut the haze. The strength of the filter depends on the density of the haze. If a filter is not used, or the film used cannot take a filter, the resulting pictures will be flat and hazy in tone, not showing anything with clarity or sharpness. The filter is simply a piece of stained gelatine of some degree of orange color and is placed between the elements of the lens.
Most of the exposed films had to be developed in Haiti at night when the air and water were coolest. The film is developed in the full seventy-five foot lengths by means of a special developing outfit. The film is first wound integrally within a celluloid apron on a reel, which is then placed in a twenty-gallon tank of developing solution. This tank has a water-tight cover, and, during the half-hour period of development, it is turned on its pivoted axis about once a minute, thus insuring the solution’s reaching all parts of the film evenly. The film is then passed into the usual fixing bath and later washed for an hour. All of these processes are carried on in twenty- gallon silver-lined tanks. By handling the film in this way the finished roll comes out unscratched, or otherwise blemished, and is evenly developed. After washing, the film is wound on a large drying reel, from which it comes ready for numbering and printing.
The printing machine is a contact one, the paper being laid on the negative and pressed on by a lid while light is turned on beneath the film. This printer was developed specially for handling mosaic film. It is very complete, having a pneumatic cushion on the under side of the lid to press the paper evenly and surely over the whole negative and an electric blower to dissipate the heat from the nine lights in the printer. It has special holders for the spools holding the film during printing and separate switches for each of the nine lights, so that some may be turned out under an uneven negative to make an even print.
In laying a mosaic map from a number of prints, a completed map is desired that looks like one picture of the area covered and not like a crazy quilt with any number of shades and tones where the different prints are joined. For this reason great care must be exercised in order to obtain the same tone on all prints going into a mosaic map. A light French blue-grey is preferable and copies best. In laying the mosaic, a print is first covered with an adhesive, preferably gum arabic, and then laid on the base for the map by wiping it down with a piece of wet cotton, beginning at one side of the print and gradually working across to the other edge. This eliminates air bubbles which otherwise would form under the print and cause bulges in its surface. The next print is then taken up and studied carefully, with reference to the territory on it that is covered by the preceding print. Let us say that it overlaps the first print 50 per cent. There is on it half of the area covered by the one already down. This means that we can cut away and discard at least half of it. The cut is made just beyond a roadway, fence line, hedge, river course, or any other natural break in the terrain in order to make the dividing line as nearly invisible as possible. The cut is made with a razor blade, and only half way through the paper. The part to be discarded is torn back and under the part to be retained so as to form a beveled edge. This edge is then sandpapered to a feather edge, and after the adhesive is applied, is laid over the first print, covering exactly all of the territory that is identical to each. The reason for making the cut beyond the hedge or roadways, and keeping it on the part retained, is that greater accuracy may be secured by having the line or object on the second print to lay directly over the like line or object on the under print. Print number three is then cut and laid down on number two in like manner and so on until the end of the strip is reached when the adjoining one is begun. In this way a map which is truly a mosaic emerges from the prints.
Aerial mosaic maps have been put to a very practical use in Haiti by the Travaux Publics. For a given plain they have made use of them in the following ways. From the film, one set of prints was made for laying into small mosaics, plane table sheet size, that could be used in the field; a set of prints was made for use under the stereoscope to be used in the field areas in connection with the small mosaics, and another set of prints was made for laying into a large mosaic to be retained in the laboratory for copying and for reference.
The small mosaics and stereoscopic prints proved to be of great value in the field, where they saved the engineers untold labor and time. The country in which they were working was tropical and hot, with jungle undergrowth which left the trails barely passable. Even on horseback they could not know what was on the other side of the thick cactus hedges bounding the road or trail. There were no knolls from which to make sights. But with the small mosaic of the area in which they were working, and by use of the stereoscope with the extra set of prints, they could see the whole terrain. They could know what to expect ahead on the trail they were following, which way to turn at a fork in the trail or road to bring them to the desired destination. As one engineer said, “With two of these prints under the stereoscope, I can more accurately trace the contours, roads, rivers, streams and section lines in fifteen minutes, than I could possibly do in a whole day by tearing my way through the bush.” One print covers an area a mile and a half square and by using the one before and after it under the stereoscope, all of it can be quickly covered. The engineers have from one to three known elevations that they have brought up from the previously worked area to the present one. Under the stereoscope they trace the contours through these points, and then fill in the others in proportion, judging the heights from what they can see between the known elevations.
After the required data is placed on one print, it is retraced on the like print in the small mosaic. Thus as the data is gradually secured from each print and placed on the mosaic, the mosaic soon has all the required information on it for the section of country it covers. Tracings are then made from the small mosaics until a complete contour map is had of a whole plain or valley.
The stereoscope mentioned above is of such great value when in mapping with prints, that it deserves a word here. During the last war the stereoscopic principle was used and to great advantage, but with the old time stereoscope with which every one is familiar. It was used as follows: Mapping prints must have overlap to qualify as such. Therefore they are similar to two pictures taken with a huge camera having two lenses half a mile apart, for example, the distance between positions of the plane at the time of each exposure. All prints with overlap have potential stereoscopic value, which is to say that all the area identical to each may be viewed in strong relief with a stereoscope. When any part of a mapping print was to be viewed with the old type of stereoscope, it was first necessary to decide what part of the print possessed military information. Then this portion, say two inches square, was cut from each of the two prints and pasted on card board two and a half inches apart (the average distance of the human eyes). Then this card was slipped into the rack on the old stereoscope and everything in the little square of the picture stood up in relief.
This system had many disadvantages. There was the delay incident to preparing the parts before they could be viewed, the restricted area that could be seen at one time, and the probability of making a poor decision in the part cut out. Good camouflaging of a position or road would enable it to pass unnoticed.
All of these disadvantages disappeared at once with the advent of this new stereoscope. It was brought out by the same firm that makes the K-3 camera and it is of inestimable value from a commercial and military standpoint. When mapping prints are viewed with this stereoscope all that is necessary is to place the two prints about a foot apart on the table on which it rests. Then with a little shifting for correct position, all of the country identical to each print immediately jumps up into strong relief, amazingly sharp and clear. On prints taken at ten thousand feet elevation, the differences in relative height between the upper and lower branches of a tree and the ground can readily be noticed.
As the smallest depression or rise in the ground shows clearly, it can be seen at once how rapidly a given section of country could be covered in this way. The camouflage over a road or position cannot hide it for it will be seen standing up over the road like a flat roof twenty feet high and the artillery can be trained on it at once.
The simplicity of the construction and use of this stereoscope is marked. It consists of a fairly heavy base supporting a horizontal bar a foot above it. In the middle of this bar two and a half inches apart, are two small mirrors an inch and a half by two inches in size, placed with their faces up, but tilted outboard at a forty-five degree angle. At each end of the bar, a foot away from the small mirrors, are two large mirrors, also placed at an angle of forty-five degrees. The large mirrors face downward and inboard, so that each little mirror faces each large mirror. A horizontal line would pass through the center of all of the mirrors.
To use the stereoscope the prints are placed approximately under the large mirrors, which pick up the image of each print.
Then one looks down into the small mirrors, with the eyes very near them, and in each sees the image in the large mirror. This gives the relief illusion, the principle of the stereoscope.
The kind of mapping described above has been done for the Travaux Publics of Haiti for the last two and a half years by Observation Squadron Two, the Marine Corps Aviation Unit at Port au Prince. The work began with the Artibonite Valley job and the results attained were so satisfactory to the naval officers of the Travaux Public, the American high commissioner and to the president of Haiti that they desired further mapping. Therefore, as rapidly as the photo officer and two men could furnish them, these mosaic maps were made. It took all of the men’s time developing the film, making the prints, and performing the incidental work necessary for the maintenance and operation of a mosaic laboratory. Whenever favorable flying weather for mapping occurred, the pilot and one of the men would have to drop everything and go.
During the above mentioned period, all of the valleys and plains of Haiti, except the inland plain not potentially irrigable, all of the roads, several of the rivers, and some special projects, were mapped. The areas mapped equalled over two thousand square miles, and the roads a thousand miles in length. Including the field maps, over four thousand square miles were laid in mosaic. Over eight thousand negatives were exposed, and from these about thirty-five thousand prints were made. All of these areas were laid into large maps and copied. Field maps were made of most of the areas and stereoscopic prints were furnished. All areas desired were mapped. Large maps of the Northern Plain were not made, the field maps of a third were furnished, and several areas laid twice. Only a few stereoscopic prints of the road maps were furnished up to the middle of April.
The photographic officer laid all of the mosaics, except some of the small field maps of the Artibonite Valley and a few of the road maps that were not completed when orders were received transferring the photographic officer and one photographer to Nicaragua.