Scotland’s Land Cover

Contents

• introduction
• Origins of Land Cover Mapping
• Scottish Biogeography
• Climate
• Geology and Soils
• Time
• Land Cover Mapping
• The Land Cover Map of Scotland
• Future Landscapes of Scotland
• Support Materials
• About

6. Land Cover Mapping

What is land cover?

Land cover refers to features of land surface. These can be natural, semi-natural, managed or totally man-made. They are directly observable. Land use, on the other hand, refers to activities such as agriculture or recreation. Uses are not directly observable, although inferences about use can often be made from land cover (e.g. the distinctive features of a golf course). A land cover map is not a land use map.

The main reason for producing land cover maps is to give us a clear idea of the stock and state of our natural and built resources. A land cover map, or land cover inventory, is an essential component in developing a responsible attitude to environmental management. Would a business be well-run if its owners did not know what funds they held in the bank?”. The same applies to the management of the natural resources of the United Kingdom PLC.

Land cover maps are essential inputs to environmental and land use planning at local, regional, and national levels. They are being used directly in, for instance:

• resource inventories (e.g. woodland census); or
• state of the environment reporting (e.g. establishing the distribution, extent and quality of key habitats); as well as
• input to the development and monitoring of a range of land use and conservation policies (e.g. in ‘Environmentally Sensitive Areas’ or native pinewoods habitat restoration).

How do we map land cover?

The critical question is “what do you need to know?”, and its answer defines the best mapping option.

Since land cover is directly observable, the size of the features which you want to map drives the choice of mapping approach [see Poster]. Data for mapping may be acquired from:

• satellite imagery
• air photography
• ground survey

For example, mapping forests is quite feasible from satellite images. These currently have a spatial resolution of 10m at best (i.e. it is only possible to distinguish between clearly differentiated individual features of greater than 10m x 10m in size). New developments in imaging technology make resolutions of around 1m conceivable in the near future.

In terms of map scale, satellite imagery can provide useful small scale maps of land cover, equivalent to 1:250,000 or 1:100,000. In the UK, however, most planning and environmental agencies require finer detail and thus the better spatial resolution provided by aerial photographs. The scale, or spatial resolution of an aerial photograph, is determined by three variables:

• focal length of the camera lens (ƒ)
• flying height of the aircraft (H)
• ground height above sea level (h)

Medium scale aerial photography (i.e. 1:25,000) is routinely flown over the UK to provide coverage for the Ordnance Survey’s rolling programme of mapping and updating. Such photography can be used for land cover mapping and locally provides a valuable source of base material for studying historical land cover changes. However, as different areas are flown in different years it does not provide a standard time base. The Land Cover of Scotland 1988 (LCS88) therefore used specially commissioned aerial photography, the bulk of which was flown in 1988.

Interpretation is greatly assisted by stereo coverage. Both aerial photographs and satellite images can now be used for this. Stereo coverage is acquired by taking photos of the same area of the ground from two different viewpoints. The principles of stereo viewing are explained by comparing the use of two photographic viewpoints with the ways our own eyes allow us to see in three dimensions. The distance between the two different viewpoints in aerial photography is often called the giant eye base.

Whilst most land cover mapping is undertaken remotely, through the use of satellite imagery or aerial photography, sometimes greater detail is required, for instance on the species composition of vegetation. This can only be accomplished on the ground, by field surveys. The cost of providing such detailed information for the whole country would be prohibitive. Field surveys of this type are therefore undertaken on restricted areas, or they are incorporated within a sampling scheme to provide statistically valid estimates based upon a limited number of detailed field observations (e.g. using vegetation quadrats).

Some important land cover mapping initiatives

There are no statutory requirements for land cover surveys of the UK (contrast with the statutory requirements for the June Agricultural Census which farmers complete annually.) This has meant that there is no standard UK land cover classification system.

Because there is no statutory requirement for land cover surveys, there is no single agency with responsibility for them. Among numerous examples of land cover mapping in Scotland, four national-scale initiatives have been conducted by three different organisations/agencies. Each had its own objectives, reflected in the different methods used. This highlights the key point – how you choose to map depends on the answer to the question: “what do you need to know?”.

Countryside Survey: the Institute of Terrestrial Ecology have developed a sampling-based system for assessing land cover/habitat changes at a national level. First applied in 1976, it was repeated in 1984 and 1990. The next repeat will be in 1998 for reporting in the year 2000 (hence, it is referred to as the Countryside Survey 2000). It uses detailed field surveys/mapping of some 508 1 km grid squares selected throughout Great Britain on the basis of the ITE national land classification [a use of the 32 Land Classes is illustrated in Chapter 2]. It provides robust statistical estimates of land cover/habitat changes at a national level. However, because of the small number of samples, which cover about 0.25% of Scotland’s land area, it does not provide reliable estimates at a local level.

Land Cover Map of Great Britain (LCMGB): also undertaken by the Institute of Terrestrial Ecology and produced in 1993, this was the first ever national land cover map to be derived from satellite imagery. The map provides a census of land cover using LANDSAT imagery obtained between 1988 and 1990. Interpretation of land cover was computer-assisted and used images of the same areas taken at different times of the year to improve interpretation accuracy. Some 25 land cover classes were classified. The spatial resolution, according to the pixel size of the imagery, was 25m x 25m. With filtering of single pixels, the minimum mappable unit is taken to be 0.125 ha. Because it has produced an electronic land cover map for the whole of Great Britain, it has been utilised in the development of an integrated database on the natural environment of Europe.

Land Cover of Scotland 1988 (LCS88): completed in 1993 by the Macaulay Land Use Research Institute (MLURI) on behalf of a Scottish Office-led consortium. The survey was undertaken specifically to provide a baseline for future monitoring of land cover changes. It now provides a medium scale census (1:25,000) of the land cover of Scotland based upon purpose-flown aerial photography [see poster]. Manual interpretation of the aerial photos used 126 land cover classes with a variable minimum mapping unit of between 2 and 10 ha depending on land cover feature. Accuracy was field checked using rigorous statistical procedures. The maps were digitised and have been used in combination with other digital maps (e.g. soil, topography, climate) to explore a wide range of land use and environmental issues at local, regional and national levels in Scotland. [The map is illustrated in the Advances 4 poster].

National Countryside Monitoring Scheme (NCMS): this used historical aerial photography, within a sampling framework, to assess land cover change since 1947. The NCMS was initiated by the former Nature Conservancy Council and was completed in 1997 by Scottish Natural Heritage. It now provides estimates of land cover change from the late 1940s to the early 1970s and to the late 1980s. It does not provide total area mapping coverage: the 464 sample squares cover about 7.5% of the area of Scotland [See Advances 5 for NCMS details].