Chapter 1: Introduction

Introduction

Remote Sensing and Air Photo Interpretation

This collection aims to show the geography of Manitoba using remote sensing, a process that involves obtaining information about the earth’s surface without coming into contact with it. A range of instruments can sense much of the electromagnetic spectrum, whereas the human eye can sense only a narrow band of wavelengths from 0.4 to 0.7 micrometres, the so-called visible part of the spectrum. Most people are now familiar with at least air photographs, obtained using cameras, and with radar images. Radar has the advantage of not requiring light to operate and some radar wavelengths can penetrate clouds. Many Manitobans are familiar with the spectacular radar images of the Red River Plain obtained during the height of the 1997 “flood of the century.”

In remote sensing the remoteness of the sensor can vary enormously. Photographs can be taken from a few metres above the ground using a camera mounted on a “cherry picker” but most air photographs are taken from planes flying between 5,000 and 30,000 feet[i] above sea level. Photographs and other forms of images are obtained from satellites at heights up to thousands of kilometres. Despite the development of many new remote-sensing techniques, air photographs are still the best known and, for the non-expert, the most easily interpreted of all images; consequently, most images in this collection are air photographs.

Air photographs come in several different forms, the most obvious division being into colour or black and white (panchromatic) photos. Colour photos are more impressive and more easily interpreted because they are more like what people are used to seeing. However, they have some drawbacks vis-à-vis “black and white” photos. Despite modern developments, colour photos are more demanding with respect to light conditions, a factor that limits their availability. Consequently, colour air photo coverage of the province is limited to a few areas and was usually taken for specialized purposes such as vegetation mapping. In contrast “black and white” images are available for every part of the province and often coverage of the same area for several different dates is available. These so called “sequential air photos” are particularly useful for the interpreter trying to detect changes over the years, for example, at the edges of urban areas and along meandering rivers (figures 1.1, 1.2, and 1.3).

Another standard sub division of air photos is into verticals, obliques, and mosaics. As the name suggests, verticals are taken with the camera pointing directly downwards from the plane[ii]. These are the most readily available of the categories; consequently, most images in this collection are verticals. Obliques are taken at an angle of less than 90 degrees to the earth’s surface. They are easily recognizable by most people because they resemble the view passengers get from a plane window. They are, though, much less readily available than verticals, and, because the scale varies from place to place on the photo, measurements on them are complicated. Mosaics are secondary photographs made by re-photographing a series of verticals placed in their correct position with respect to one another. The result is a good overview of an area. Mosaics are available for all of Agro Manitoba[iii] and were used extensively in trying to locate the exact position of verticals used in this collection.

Air Photographs vs Maps

Geography “seeks to understand where things are and how and why they got there by studying—with an emphasis on location—the connections and interactions among people, places, and environments."[iv] Maps are useful in showing locations, patterns and the interrelatedness of things. They are indispensable to the geographer, so much so that some claim that if you cannot show it on a map it is not geography. Photos also show locations, patterns and the interrelatedness of things and have some additional merits compared to maps. An air photograph shows the surface of the earth as an observer looking at the ground from a plane would see it and, as most people who use this collection will have flown, the photos will be familiar to them. An air photo shows everything “seen” by the camera at a specific instant; maps on the other hand are sets of symbols. Although people become familiar with the map symbols over time, it takes skill and experience to become a good map interpreter. Moreover, the cartographer—mapmaker—cannot show every aspect of the earth’s surface, as the resulting map would be cluttered. Consequently a map is selective in what is shown.

A point against the air photo is that its time specificity gives the interpreter no indication of seasonal changes of the landscape. Most photos are taken during the summer, when the sun is high in the sky and the days long. Consequently winter photos are rare and few were available for inclusion in this collection. Nor do air photos show abstract things such as political boundaries, although different land use on either side of a boundary may give it physical representation as for example along the Canada/U.S.A. border (figures 2.5 and 2.6).

One further difference between air photos and maps is that maps are drawn as if it were possible for an observer to be directly above each point on the earth’s surface. A map is therefore a true plan view or orthogonal projection. In contrast when a vertical air photo is taken, the camera is directly above only one point on the ground; all other points are seen slightly “sideways on.” So, vertical air photos are perspective projections and as a result some points on the ground may be hidden from the camera by tall objects or high ground. This is particularly troublesome in mountainous areas (so it is not a major problem in Manitoba) and in areas with many tall buildings (limited in Manitoba to Winnipeg only-- Figure 1.13).

Air Photo Interpretation: The Process

Although we aim to keep this description non-technical, the reader needs to know how air photos are interpreted, which involves the introduction of some terms. Interpreting air photos is like detective work; a series of clues is assembled in an orderly fashion and then a solution is arrived at which best fits the clues. The interpreter uses seven criteria in the interpretation of photo images: tone, texture, pattern, shape, size, shadow, and location (or association).

Tone is a measure of the amount of light reflected by an object and recorded on a panchromatic (black and white) photograph. Tones are shades of grey; strictly speaking they range from very dark grey to very light grey—not from black to white. Tones vary greatly depending on the time of day, which influences the sun’s angle; the type of film and filter used; and the development process. On the whole, dark coloured objects appear dark grey and light coloured objects, light grey. There are, though, exceptions; two water surfaces on the same photo can vary from almost black to almost white, depending on the relationship between the sun’s angle, the water surface and the camera (figure 1.4); tonal contrast is more important than the actual tone. The human eye is sensitive to very small tonal differences and minor variations can represent significant differences on the earth’s surface (figure 1.12). Tone is a characteristic of all the other interpretation criteria.

Texture is difficult to define and is best described by reference to the texture of cloth. The same sort of terminology is used, so photo surfaces may be rough (figure 1.9), mottled (figure 1.4), or silky (figure 1.7). The texture displayed on a photo results from the arrangement of tiny images expressed by the tone, shape, size, and pattern of the object displayed. The texture of an area does not always remain the same; for example, a lake surface in calm weather will be expressed as smooth or even textured (figure 1.6) whereas with a strong wind blowing, the surface will be rough and will be expressed as a rough texture on an air photo. Texture also depends on the scale of the photos (explained later). For example, rows of trees in an orchard will show up as a series of round objects arranged in lines on a large scale photo, but on a smaller scale photo, the individual images merge to produce a rough texture. At a very small scale, all variations may disappear to produce a smooth texture (figures 1.15a, 1.15b, and 1.15c).

Pattern refers to the orderly arrangement of features. Some elements of the natural landscape form patterns; for example, under the right circumstances, river channels may be arranged into distinctive drainage patterns. However, more obvious patterns are associated with human activity. In much of Agro-Manitoba, for example, the pattern of square fields resulting from the Dominion Lands Survey (DLS) system is distinctive (figures 1.14, 1.15a, 1.16), but land that was settled by people of French heritage exhibits the equally distinctive long lot system (figure 1.12).

Shape is the general form or configuration of an object. Some landforms have distinctive shapes, for example, the loops of meanders (figure 1.2a, 1.2b); however, as with patterns, human-made objects tend to have more clearly defined and easily interpreted shapes. The “fish-like” appearance of golf courses (figures 17.15, 17.16) is often seen near the edges of towns and cities, as are oval shaped racetracks (figure 1.9). Also, the fan shape of drive-in cinemas (figure 12.12)—now almost extinct—and the triangular shape of small prairie airfields (figures 18.45 to 18.48) point to their functions.

Size, both actual and relative, is another criterion used. Even though the interpreter may be unable to determine the exact size of an object, it is usually possible to obtain a comparative size with respect to something whose size is known. An object at the edge of a field may be wider than a nearby road so the object is clearly not a car. Also the relative size of objects on a road may indicate whether they are buses, trucks, or cars (figure 1.13).

Shadow: Photographs are usually taken when the sky is cloud-free so ground objects will throw shadows; only rarely do cloud shadows appear on photos (figure 1.14). The length of the shadow will depend on the time of day the photo was taken: a long shadow, early and late in the day; a short shadow, near midday. Often the shadow shape will point to the nature of the object; for example, coniferous trees have different shadow shapes from deciduous trees (figure 1.6, 1.13). Experienced interpreters can even determine tree species from their shadows. Human-made objects such as electricity pylons, TV towers, bridges and the old wooden grain elevators (figure 1.9) throw distinctive shadows. Also because shadows are thrown generally northward, rather than to the south, the orientation of the photograph can be determined if it is not known from other information.

Location (sometimes referred to as association) is another useful criterion. The interpreter is concerned with the location of an object with respect to other things. Is the object in the middle of a city or in an open field? Is it at the top of a hill or the bottom of a valley? Is it closely associated with railway tracks or roads? The answers to these questions can lead to a positive identification or at least the elimination of some possibilities (figures 1.8, 1.13).

Scale is a factor that has an influence on all the photo identification criteria. Scale is the ratio between a distance on the photo and the distance it represents on the ground. It is usually expressed in the form, 1:20,000, meaning that one unit of measurement on the photo represents 20,000 of the same units on the ground. Scale is influenced by the height from which the photos were taken. Assuming the focal length of the camera remains the same, as the plane carries the camera higher, more ground is covered on an individual photo but at the same time there is a loss of detail as the scale decreases. Conversely, photos taken from low flight heights cover less ground but more detail is shown at a larger scale (figures 1.15a, 1.15b, 1.15c).

Stereoscopic Vision

People with two good eyes see things in three dimensions; i.e., they see stereoscopically. This is because they see objects from two different viewpoints—the right eye and the left eye. The two sight lines converge on the object being viewed and the two images are processed in the brain to give a three dimensional impression. Taking photographs from two different plane positions simulates this situation. The most common situation is for a plane to fly over an area with a camera automatically taking photos so that photo two overlaps photo one by 60 percent and photo three overlaps photo two by 60 percent and so on. In this way each point on the ground is photographed from two different viewpoints, duplicating the position of the two eyes. Two photos with one overlapping the other are known as a stereopair and three as a stereotriplet (figure 1.16a, 1.16b, 1.16c). To view the photographs, one arranges them side by side with the correct orientation and observes them through a stereoscope. The stereoscope is merely a device to help the observer to look at the left-hand image with the left eye and the right-hand image with the right eye. The brain fuses the two images to give a three dimensional (stereoscopic) effect. Usually the relief is exaggerated over reality. If the photographs are taken closely together to more closely resemble the sight lines from the two eyes, the exaggeration is reduced. Conversely, decreasing the overlap increases the exaggeration. Usually the vertical exaggeration is a bonus in areas of “moderate” relief such as Manitoba where minor landforms with little amplitude of relief—vertical distance from top to bottom—are more easily detected and interpreted.

Air photos should be arranged so that shadows fall towards the viewer. Failure to do so can result in pseudoscopic vision; that is, the topography is reversed—valleys look like ridges and ridges look like valleys (figure 1.17).

Organization

This collection follows the same outline as the book, The Geography of Manitoba: Its Land and Its People[v] so that the book and air-photo collection can be used in conjunction with each other. However, it has to be acknowledged that while some aspects of Manitoba’s geography are well displayed on air photographs, others are not. For example, landforms are very well shown but climate is not. Consequently the relative weighting of topics varies from that given in the book. Nevertheless, the sequence of topics is essentially the same but in the collection we treat transport as a major topic. In the book it was subsumed under other headings.