Measuring and Mapping Wilderness—A Brief Review of Metrics and Methods
Wilderness has been mapped and analysed across scales, from global to local level. The methodologies generally make use of available spatial data on human infrastructures, land cover, area size of ecologically intact regions, etc. as proxies for wilderness quality, but also employ expert knowledge on degree of naturalness and ecosystem modification. Despite the obvious challenges of mapping a multidimensional concept such as wilderness, studies using relevant indicators at a similar extent and resolution offer highly congruent results, likely because they share a common perception of the attributes and values of wilderness.
At the global level, Mittermeier et al. (2003) used a combination of human population density, intactness, and area size of the intact areas to define wilderness areas. Much of their assessment was based on literature and expert opinions. The wilderness areas identified coincided with the areas of the lowest human footprint identified by Sanderson et al. (2002) although the two studies used largely different metrics. The map of the human footprint at the global level used human population density, the transformation of land through the building of settlements, roads and railroads, and measures of human access. Power infrastructures were also quantified, using satellite night maps (Sanderson et al. 2002). Despite data limitations, these global studies reveal a fairly consistent big picture of the overall pattern and magnitude of human impact on the biosphere, both for terrestrial and marine ecosystems (Halpern et al. 2008).
In Australia, the Heritage Commission's National Wilderness Inventory used four metrics for defining wilderness: remoteness from settlements, remoteness from access, biophysical naturalness and apparent naturalness (Lesslie et al. 1995). In this case, thresholds were defined for minimum levels of these metrics that would characterize wilderness. Other approaches emphasize a wilderness continuum across the landscape (Fritz et al. 2000). Building on the Heritage Commission's National Wilderness Inventory research, Carver et al. (2002) added remoteness from national population centres and altitude in order to map wilderness in the United Kingdom.
Remoteness from national population centres was a measure of the accessibility to the whole British population in addition to the accessibility to the local population in the calculation of wilderness. The authors used multicriteria evaluation (MCE) and explored public perceptions of wilderness through the use of interactive tools by allowing the user to change the weights of the wilderness metrics. As expected, resulting wilderness maps were not radically different, but allowed for insights on what affects the perceptions of wilderness (Carver et al. 2002). This approach was further detailed at the level of the Cairngorms National Park, and the Loch Lomond and The Trossachs National Park in Scotland (Carver et al. 2012) at a resolution of 20 m and later expanded to cover the whole of Scotland in a study by the Scottish Natural Heritage (Scottish Natural Heritage 2012).
At lower spatial extents the indicators of wilderness and human footprint remain the same but higher quality data are usually available making the mapping and modelling process more reliable and accurate. For example, Woolmer et al. (2008) rescaled the human footprint methodology of the Sanderson et al (2002) for the area of approximately 300,000 km2 of the Northern Appalachian ecoregion. They used ten datasets compiled from several sources: population density, dwelling density, urban areas, roads, rail, land cover, large dams, watersheds, mine sites, utility corridors for the electrical power infrastructure. The general patterns of human footprint were maintained when comparing the map based on 90 m2 resolution data at ecoregional scale with the map derived from the global analysis of Sanderson et al (2002) conducted with 1 km2 resolution data. However, the Spearman rank correlation coefficients between the two sets of human footprint data steadily decreased with the scale, reaching 0.41 ( p < 0.001) at 0.1 % of the Northern Appalachian ecoregion. The difference in the human footprint scores is that the ecoregion calculation compared with the global calculation leads to a reduction in the area with low levels of human footprint (46 % ecoregion extent vs. 59 % global extent) and an increasing of the area with moderate or high levels of human footprint (34 % ecoregion extent vs. 21 % global extent), evening out more the distribution of human footprint scores. A key finding was also that three parameters models add the most information to the calculation of human footprint while the model incorporating human settlements, roads and land-use was the best approximating model from all combinations of the ten datasets considered.
In Europe, an increased wilderness momentum has led to efforts by different actors to protect wilderness and advance a progressive wilderness research agenda (Jones-Walters and Čivić 2010). A continental level map of wilderness continuum has been produced using population density, road and rail density, linear distance from the nearest road and railway line, naturalness of land cover and terrain ruggedness (Carver 2010). This analysis identified wilderness areas concentrated in the Scandinavian Peninsula and the mountainous regions of Europe, revealing a strong positive altitudinal and latitudinal relationship. The same pattern was maintained even if terrain ruggedness was eliminated from the calculation. Beside the Scandinavian mountains and arctic areas, the Pyrenees, The Eastern Mediterranean islands, the Alps, the British Isles, the south-eastern Europe and the Carpathians also had significant areas of wilderness (Carver 2010) but one has to temper this with the knowledge that the current spatial data often misses historical information on local land use management such as past deforestation, drainage and grazing by domestic livestock. Currently, the wilderness mapping is being updated through the project of the European Wilderness Registry, which will record the most important wild sites, thus facilitating priority setting for protection.