The development of aerial photography

Aerial photography can be used for four broad purposes in archaeology: as illustration, for research as data in its own right (where the virtues of illustration are applied to create new knowledge), as an aid to field research and excavation, and finally as an aid in determining the rate of destruction of sites and measures needed to conserve them. ⁵

After being used in the First World War for reconnaissance in both the Near East and the European theatres, ⁶ aerial photography was quickly grasped by the United Kingdom Ordnance Survey as an essential too for archaeological survey and mapping. In 1924, concluding a paper on 'Air Survey and Archaeology', O.G.S Crawford noted:

It is difficult to express in suitable words my sense of the importance of air-photographs for archaeological study They provide a new instrument of research comparable only to that provided by excavation. . . . They are not a substitute for fieldwork, but they are the most powerfually of the field archaeologist. . . . As a revealer of almost vanished earthworks they are superior in every way to observation on the ground. . . . and it is to be hoped tha they will figure largely in illustration of future books which may be written about the earthworks of a country . . . ⁷

In a subsequent paper, ⁸ Crawford and Alexander Keiller reviewed the history of aerial photography, and first systematically described the distinction between page 17 archaeological sites indicated only by changes in crop patterns (stunting or abundance of growth known as 'crop marks') caused by archaeological features beneath the surface, and sites which could be recognised from their surface relief alone ⁹ —techniques to which we shall return shortly. Indeed from about this time, once the pioneering work of recording the surface relief of earthworks or monuments such as Stonehenge or Bronze and Iron Age fortifications had been completed, attention turned much more to crop marks. Over the years since, Cambridge University has collected aerial photographs which have been used for many publications on the historical landscapes and sites of the United Kingdom. ¹⁰

In New Zealand in this same era, an interest in recording and making sense of the mark of human beings in the landscape is apparent in the early work of W.H. Skinner and S. Percy Smith, ¹¹ both surveyors by profession, and in the writings of the ethnologist, Elsdon Best. ¹² All were great interpreters, and not just for a European audience, of the Māori traditions of that same landscape. ¹³ Elsdon Best was publishing on pā at the same time as O.G.S. Crawford in England was publishing aerial photographs of Iron and Bronze Age hill-forts, but there is not the slightest hint in Best's writings that aerial photography might have been of use, nor any apparent interest in the potential of aerial photography for historic site work.

By the late 1930s aerial photography was being widely used in New Zealand, ¹⁴ and by 1945 there was full coverage for military purposes of the coastal districts. In New Zealand, a comprehensive archive of aerial photographs was built up. ¹⁵ Aerial photography from the time of the Second World War is still filed and available from the Department of Survey and Land Information.

However, it was not until 1947 that a paper on New Zealand archaeology and air photography, covering sites such as Papamoa (Bay of Plenty), the Wairau 'canals', and pā at Pakipaki, central Hawke's Bay, was published by G. Blake-Palmer, a general practitioner and later public servant. ¹⁶ His paper usefully reviewed the English experience and discussed matters of scale, the then-existing corpus of vertical aerial photography, crop marks and optimum seasonal and diurnal lighting conditions. Shortly after, Leslie Kelly published a booklet on Marion du Fresne, the French explorer, which used aerial photographs to illustrate the historical landscapes of the Bay of Islands. ¹⁷ Overall, however, Blake-Palmer's paper does not appear to have been particularly influential. The Wellington Archaeological Society did have lectures by him, and the use of aerial photographs was common in the extensive surveys of the Wellington region conducted by the society in the early 1960s, and surveys elsewhere in the lower North Island, for example, Colin Smart's work in Whanganui. ¹⁸

Blake-Palmer's paper was, of course, before its time, since there were no full-time practising field archaeologists in New Zealand until a decade later. In 1957 Jack Golson discussed the use of aerial photography, ¹⁹ but it was not until the publication of Archaeology in North Taranaki by Alastair Buist in 1964, ²⁰ one of the more distinguished reports produced from that 'boom' decade in New Zealand amateur archaeology, that the potential of aerial photography for illustration and analysis began to be fully realised. Here for the first time, in a modestly produced book, were extraordinary images of the Taranaki historic landscape, emerging through the limitations of the printing. Subsequent books which have used aerial photographs for illustration have been J.D.H. Buchanan and David Simmons's Māori History and Place Names of Hawke's Bay, a superb historical landscape illustrated by black and white photographs by New Zealand Aerial Mapping, Ltd; ²¹ Barry Brailsford's The Tattooed Land, a study of pā in the South Island; ²² and some recent books on Taranaki and the Waikato, notably work by Nigel Prickett and F.L. Phillips's Ngā Tohu a Tainui ²³

These books use aerial photographs by way of illustration for a text which is more concerned with traditional and archaeological knowledge. With the exception of the Buist monograph, they are not strong attempts to use aerial photographs as a source of new knowledge. There have been attempts (unpublished) to map individual sites using aerial photography, notable examples are: Te Pōrere (discussed in chapter 9); a pre-European pā at Hawai in the eastern Bay of Plenty; Ruapekapeka (discussed in chapter 7); some of the Auckland volcanic cones; Pouēua in the Bay of Islands; and Tapui near Manūtūke, Gisborne. ²⁴ With the exception of the last two, all these exercises have produced results of indifferent archaeological value, mainly because they were done by technical photogrammetrists and not closely enough interpreted to yield informative archaeological results— for example, walls or banks constructed by human beings were not distinguished from naturally occurring high points or features. Aerial photographs are also used quite widely in the field to assist archaeological mapping and page 18site location, and these results are filed as site records with the New Zealand Archaeological Association. Such site records are a source of new knowledge, although they have seldom been gathered into a published work.

Hawke's Bay seems to have led the development and use of aerial photography in archaeology, because the principal company involved with aerial photography, New Zealand Aerial Mapping, Ltd, is based there, and because sites in the region are generally well sculpted into the ground and maintained in grass—ideal candidates for aerial photography. Les Groube, formerly a lecturer in archaeology at Auckland and Otago universities, was influential in the 1960s in guiding his students in the use of aerial photography. He also carried out aerial photography in Hawke's Bay in the early 1960s. ²⁵ Another of his important projects was the mapping of site locations in the Bay of Islands for a pioneering study of site distribution. ²⁶ Elsewhere, not far from the Hawke's Bay region, the Royal New Zealand Air Force photographed the Waiu pā complex, situated at an altitude of 1,100 m on the southern edge of the North Island's volcanic plateau. Te Pōrere (to be discussed in chapter 9) was also photographed from a helicopter at about this time by the late Ormond Wilson, to record its restoration. In the South Island, Hardwicke Knight also photographed a number of sites in the early 1960s (his photographs are in chapter 5).

The most successful attempt to date to generate new knowledge from stereoscopic aerial photographs is that of Foss and Helen Leach of the University of Otago, whose work was influenced by Les Groube. Helen Leach's published maps of garden plots and stone rows or walls in Palliser Bay ²⁷ were significant both in detail and in the extent of features covered. Many purpose-flown vertical aerial photographs at a large scale (approximately 1:1,600) were used to map the coastal strip, with most of the detail checked in the field. (At this scale, a wall 2 m wide on the ground is 1.25 mm wide on the photograph.) In the course of the work she encountered and solved problems arising from lighting (high overhead or strongly oblique) and its effect on her ability to map the walls; changes in specific forms of vegetation cover on the walls, allowing them to be recognised; the loss of data due to blanketing vegetation cover; and distortion of the image at the edge of the lens. Most importantly, Helen Leach noted the value of the aerial view in being able to detect pattern and functions in features that defeated clear interpretation in the ground view. ²⁸

There have also been important regional surveys using low-level oblique aerial photography, such as Nigel Prickett's work in Taranaki, already mentioned. Another source used in this book was a survey of the Waikato undertaken by Kees Sprenger and Steve Edson of the Waikato Museum of Art and History in 1980. Little published, Sprenger and Edson's is the finest low-level aerial photographic survey in New Zealand, covering a landscape of great historical interest.

In the early 1980s Auckland University's Audio-Visual Unit also began to use oblique aerial photographs on sites. In 1982 Reg Nichol, then a student at Auckland University, with Godfrey Boehnke taking the photographs 'hung on a strap out of the door of a high-winged light aircraft', carried out a low-level survey of Hamlin's Hill, just south of the Auckland urban area. Nichol had thought that a regular pattern of parallel lines showing on the surface of the hill, and which had long puzzled the student excavators, were pre-European cultivations. ²⁹ The wider picture of the hill, unavailable in the ground view, convinced him that the lines resulted from European ploughing. In 1986 also using the university's audio-visual unit, Doug Sutton, a lecturer, arranged comprehensive low-level vertical and oblique aerial photography of Pouērua, a heavily settled volcanic cone in the inland Bay of Islands.

Aerial photographic techniques

Vertical aerial photographs are most commonly used for survey purposes. They are taken with large cameras fixed in the underside of an aeroplane capable of maintaining a stable flight path. As the name implies, the view is straight down, providing an image in the horizontal plane. Most vertical aerial photography is done with 'large -format' cameras, with single negative images as large as 200 by 200 mm—60 times the area of a conventional 35-mm camera negative. Very fine detail may be recorded on the film, hence its value for military reconnaissance, surveying, archaeology, forest- and town-planning, and other scientific and technical purposes. These photographs are usually taken in a sequence, with each photograph overlapping part of the photographs taken before and after. This allows stereoscopic viewing, whereby three-dimensional views of the land surface can be obtained. Stereoscopic aerial photography, such as that used by Helen Leach in Palliser Bay, can- page 19not, unfortunately, be adequately displayed in a book intended for wide public readership. It requires the assistance of a small but expensive pair of stereoscopic glasses to achieve basic results in archaeology in the field.

In contrast, oblique photography (taken at an oblique angle to the ground surface) has more uses in illustration. It is more successful, as a single photograph, in showing the shape or relief of the surface of the ground. Usually taken closer to the subject, it has relatively limited use for conventional land survey and map-making. However, it is still possible to take accurate measurements from such photographs, so they can also be used for analytical purposes.

The difference between vertical and oblique photographs is illustrated in a unique photograph taken by Air Maps (New Zealand), Ltd, of the Kuaotunu Peninsula, Coromandel, and the setting of its wide variety of historical features. In the foreground are the goldfields of Bald Spur and Waitaia Ridge (more or less a vertical view), ³⁰ and in the distance an oblique view of Opito Beach and Sarah's Gully, where, in the late 1950s, Jack Golson first archaeologically documented the early Māori settlement of the North Island. This settlement consisted of a small whānau or hapū settlement at the foot of a ridge in the inner part of a small open bay. The site revealed storage pits (demonstrating horticulture), dated to the early thirteenth century. ³¹

To see the relief of the earth or archaeological features on the surface without the benefit of a stereoscope, it is essential to use the advantages of shade and oblique light in defining the site. ³² In an oblique view, with the camera oriented towards the sun, early or late in the day or in winter is generally best; the surface features of the site will cast long shadows and can be seen readily. Disturbances in the surface of the ground throw a shadow much longer than the feature is high. A pattern to features which cover very large areas may be revealed that cannot be seen walking about on the ground under any lighting conditions. Aspect—whether the slope of the site faces towards or away from the sun—is very important.

However, there are some potential difficulties. The sun should not shine into the lens either directly or indirectly from water surfaces—a potential problem with a wide-angle lens. There may also be problems with haze, water vapour, dust or smoke gathering and reflecting light. The closer the camera is to the feature, the less significant these problems will be. Other angles produce different results. With the sun behind the observer and at a low angle in the sky (i.e., with the camera viewpoint oriented with the sunlight), very spectacular and strong colour and shadow effects can be gained. The light will be bright on the lit faces, and the camera will pick up the very long, very dark shadows cast behind the features of interest (provided the angle of view is high enough). However, with the sun behind the observer, archaeological features are usually strongly and evenly lit, obscuring their own shadows, and tend to be indiscernible. My own practice has developed to take both medium-angle (at about 45° to the horizontal) and high-angle (near 90°) obliques, the latter at a higher altitude or with a wide-angle lens. High-angle oblique views require uncomfortably tight banks (steep turns) by the aeroplane to get a view unencumbered by its wheels.

To this point, we have been discussing archaeological sites that show on photographs because they have surface relief, i.e., there is a curve or depression in the surface of the land. Not all sites have such relief features. However, sites without relief can still show from the air either when they are exposed on bare surface-soil (when ploughed, for example), ³³ or, as is more common, when they lie under the vegetated ground-surface. ³⁴ The former, soil marks, show as dark or light marks in the soil. The latter, crop marks, may have been caused by hard surfaces, such as paving or the lines of a foundation or path: crops or grass will not thrive on such places. The site may have incorporated a ditch or a drain, which filled with soil more fertile, deeper or moister than the surrounding earth, promoting greener grass or better crops. The contrast may be accentuated in spring as conditions slowly become dry. Specialised film, such as infrared which is sensitive to variations in the light reflected from leaves of high or low water content, may enhance the contrast. Crop marks of course also show on sites with surface relief, where there are stony soils or where the slope directly faces the midday sun, as it does on north-facing banks. In these situations, lush spring or autumn growth may be burnt brown by the heat of the sun.

Soil marks depend on surface disturbance or exposure of the soil itself. When sites are ploughed, the ditches or other features below the surface will, for a set depth, have a different colour from the surrounding soil. An earth oven (umu) will be black and lie from 10 to 50 cm below the surface of the ground. When the surface is ploughed, the even brown colour of the topsoil layer is page break

Kuaotunu Peninsula from the west, illustrating vertical and oblique effects

The photograph has been taken with a wide-angle lens. The view is vertical (immediately below the camera viewpoint) in the foreground. An oblique view oriented to the east shows from the middle distance. Several gold-mines were in operation from 1889 on the Waitaia Ridge (across centre) and Bald Spur (bottom right). Left centre is Otama Beach. In the far distance are Opito Bay and the localities of the sites of Skippers Ridge and Sarah's Gully. Mt Tahanga, source of widely used adze-stone, is the rounded knob on the far right of the photograph.

page 21

Sarah's Gully, scene of the first North Island archaeology in the 1950s, where Jack Golson first documented the early Polynesian settlement of the North Island. On the far ridge (above the rocky shoreline) is the Sarah's Gully pā. Golson's site, in which early forms of storage pits were discovered, was on the near ridge and the slopes just above Cross Creek (foreground). Later excavations were on middens on the far beach frontage, and on the sandy area on the near side of Cross Creek. The earliest radiocarbon dates for these sites indicate settlement in the early thirteenth century.

turned under, and the surface, say 15 cm down, is revealed. The oven will then show as a black patch in the ploughed surface surrounded by the ordinary brown subsoil. An example of this effect was found at the Waitaki River mouth in the early 1960s. The ovens, part of a wider pattern of settlement (the existence of which was well known), were exposed by a combination of wind erosion, ploughing and discing. They are illustrated in chapter 5. Sometimes, when a bank is ploughed, the dark and light effect shows in reverse. If the fill of the bank was dug up originally from the subsoil it will show as a lighter colour than the surrounding topsoil surface when exposed.

New Zealand sites are relatively young and the materials of which they are made match the surrounding natural patterns of earth. There are no concrete or other solid foundations in Māori structures, for example. ³⁵ Sites tend to be concentrated on the coast or on rivers and lakes, and are not very frequent in the great arable regions of New Zealand such as the Canterbury plains. Instead, they lie under more or less permanent crops such as horticulture or the pasture of dairying land. Subsoils also tend to be soft, as soft as the topsoil, and on many occasions not dissimilar in colour. These factors all con- page break

Crop marks and soil marks

Left. Soil marks of a ploughed-out pā on the Waikohu River, near Pūhā, East Coast. The pā lies on the left of the river on the opposite bank to the prominent house, centre. There were two ditches and banks, described by Elsdon Best in The Pa Maori as 'bulky earthworks'. These enclosed the point formed by the sinous stream marked by willows (coming in from the left) and the main Waikohu River. From top to bottom the pattern is: light grey line, subsoil of the outer lip of the ditch; broad light grey line, subsoil fill of the main bank; and dark patches, the topsoil and probable ovens protected from the plough by the main bank. The river is 30 m across.

Right. Crop marks outline the double ditch and bank of a pā on the high terrace at Nukutaurua, Māhia Peninsula. Fragments of the ditches and banks only survive outside of the fenceline running up and down right of centre on the photograph. Traces of the outer ditch lie at right angles to the large surviving fragment, and can be seen as a dark line enclosing the point. A pattern of later ploughing can also be detected.

page break

Two oblique views of Te Wheao, a pā on low hill country at Te Haukē, west of Lake Poukawa, central Hawke's Bay

These photographs were taken in the course of circling the pā in the late afternoon, with sunlight coming from a low angle to the north-west. The first view (top) is taken from the east, with the main features lit from behind and showing clearly. The second view (bottom) is taken from the west, with features lit from the front (i.e., from behind the camera) and showing less well. However, the photograph is taken from a fairly steep oblique angle so that the shadows created by some features, and the hill itself, can still be seen. The site is about 100 m long.

page 25 trast with the great, complex historic landscapes of southern England and the Thames valley, the places where aerial photography has been so important. Six thousand years of extensive agricultural settlement, and continuous ploughing, on chalklands or glacial/alluvial gravels with relatively thin topsoils gives greater definition of complex soil and crop marks, and hence greater scope for aerial photographic applications in archaeology. ³⁶ By contrast, in New Zealand crop marks are rare and intact sites with distinct surface features are most commonly seen.

So far, it might appear that aerial photography is a tool without parallel in archaeological sciences, and that it may as well displace all other methods of surveying and observing sites. That is not at all true, because the following important limitations apply. First, although aerial photographs can sometimes show with exact clarity and better than a ground view the sequence in which parts of a site were laid down over large areas, that sequence is as a general rule established or confirmed by archaeological excavation or the close examination of sections that are cut through the site. Second, excavation is also necessary to determine the plan of buried archaeological features, although these do sometimes show in aerial photographs. Third, and self-evidently, archaeological excavation is essential to take samples for further analysis in laboratories.

To be seen effectively, then, New Zealand sites require strong oblique lighting conditions, typical of higher latitudes, winter, or early morning and late afternoon. Unfortunately, for perfectly valid reasons, it is typical practice in general aerial survey to take the photographs in summer and near midday, and archaeological features cannot be detected because of the lack of clear shadows and sometimes the effects of thistles or other seasonal vegetation growth. The problem is exacerbated in the far north where shadows almost disappear in summer at midday (because of the low latitude, 35°S).

The height at which the photograph is taken is very important in determining the size of the objects which can be seen on the ground: too far away and the site is imperceptible, too close and only the detail of a large site is visible. Archaeological sites vary in size: from a single spot where an artefact may have been found to a pā site that extends for more than a kilometre along a ridge. The plan of an individual storage pit may be as small as 2 by 1 m. ³⁷ Forty years ago, flying went as low as 1,500 m for aerial photographic survey purposes, and the negatives were at a scale (depending on the lens) of as much as 1:10,000. At this scale 1 cm on the photograph equals 100 m measured on the ground. A ditch and bank 8 m across would measure just under 1 mm on the negative, giving ample scope to detect and document archaeological sites. Now, heights are triple that, with scales of 1:25,000 being typical. Archaeological features are small enough to become indiscernible at this scale. In this book, most of the oblique photographs were taken at heights of from 250 to 500 m so that detail as small as individual thistles is clear.

The final problem is destruction of sites and photographs. The physical record of Māori settlement in New Zealand has been lost very rapidly in the last two decades because of land development. The increasing rarity of these sites makes them valuable. Although there is reason to be gloomy and indeed alarmist about this disappearing heritage, there are factors which reduce the loss. First, sites constructed in the earth itself are robust and will survive fire, if not bulldozing or erosion. An example of loss caused by river erosion is illustrated here in the site Marama Tāwhana on the Uawa River, Tolaga Bay. Second, and most important from the perspective of this book, the sites have long been recorded by aerial photography. Where destruction has occurred, earlier aerial photographs are the more valuable. These photographs lie in archives which have not been much researched but from which there is much to be learnt.

To sum up: aerial photography preserves what may otherwise be physically destroyed, illustrates with great power what is there to be seen, and can act as a source of carefully interpreted knowledge in its own right, a powerful tool of particular value in reconnaissance, mapping page 26 and illustration of historic landscapes and sites. The potential for aerial photographs to reveal information about the patterns of sites and human settlement is enormous, ³⁸ yet very little exploited. The positioning of sites in the landscape can reveal much about how people used the surrounding land, and about the relationships between the groups of people that lived on the sites. Even less explored is the potential of aerial photography to reveal pattern within individual settlements, ³⁹ although the pattern within garden areas has been closely analysed. ⁴⁰ It is not the primary technique of archaeology; excavation is essential to the closer understanding of the function of particular site types in a system of settlement and the sequence in which the site was used over time (the function and arrangement of a site may change over time).

Marama Tawhana, on the Uawa River, Tolaga Bay

These two short lengths of ditch once defended a pā on the river point extending as much as 150 m out to the left (east) of the photograph. The inner bank of the pā (on the left) has been heavily eroded by recent floods such as those caused by Cyclone Bola. Even in the 1940s, however, aerial photographs show that there was only a small area about 5 m across and 20 m long surviving within (left of) the ditch and bank defences. The river point has built up downstream (top right, to the south-west) while the upstream bank has eroded, abruptly terminating the defences. The site is about 350 years old.