Aspect is the compass direction (azimuth) of the maximum rate of change in elevation (downhill) (Conolly and Lake 2006, 190–1; Huggett and Cheesman 2002, 21–2). In the northern hemisphere, south-facing slopes (i.e. hills with a south aspect) receive considerably more solar exposure than north-facing slopes. According to Nicke (2001, 18), ridgeways were preferred routes in the Bergisches Land, second-best were locations with a south aspect, and the third option was east-facing slopes, whereas west and north aspects were avoided. Similarly, Lay (1992, 8) notes that hillside paths suitable for wet weather and for wheeled vehicles tended to follow the sunnier southern side of the ridge. In the Black Forest, Germany, people used to take routes with a south aspect in winter, whereas roads with north aspect were preferred in summer (Tim Kerig, pers. comm.).
Polla (2009) discusses aspect in connection with LCP analysis and points out that northern exposure is supposed to increase costs of progress. But she does not include this cost component in her analysis because her focus is on east-west routes. Barbe (2007) notes that in his study area in Germany, slopes with a north-west aspect are especially exposed to weather conditions and hardly receive any sunshine, so that wet areas remain wet for a longer period than on other aspects. For this reason, he assigns a cost weight of 3 to all slopes with an aspect between 247.5° and 22.5° whereas all other aspects receive a weight of 1.
Aspect maps can be created by push-button functionality of most GIS software (Conolly and Lake 2006, 191–4; Wheatley and Gillings 2002, 120–1). The accuracy of DEM-based aspect grid rasters depends on the factors discussed above for slope computation. Different algorithms are available for aspect calculation, as for the slope maps. Special care has to be taken for flat areas where aspect is not relevant.
Pecere (2007) assumes that travelling costs increase with elevation, though elevation accounts for only 3% of her cost model. Costs of travelling at elevations of 200 to 500m are assigned a value of 2, whereas 1 is the cost value of lower areas of 0 to 200m above sea level. Yousef et al. (1972) measured energy expenditure in man and burro (a type of small donkey) in the desert at altitudes of 800m and 3800m asl on various grades and found that the economy of climbing is unchanged at altitude.
The European Commission (1995–2010) assumes that there is no negative effect on walking speed below an altitude of 2000m. For altitudes above this value, the following function is used to model the factor of speed reduction:
f = 0.15*e0.0007h
where h is the elevation in m asl. Some example calculations show that there might be a misprint in the publication, as the factor for 2000 m is about 0.61 – but a factor below 1 does not seem appropriate. Replacing 0.15 by 0.25 in the formula given above creates more reasonable results. With this correction, the factor is 1.01 for 2000m, 2.04 for 3000m and 12.6 for 5600m.
Minetti et al. (2006) refer to a study which found that the maximum metabolic power decreases at high altitude, i.e. at 5600m asl, the maximum metabolic power of a Caucasian is only 61% of that measured at sea-level. So the values given by the (corrected) formula used by the European Commission seem to exaggerate the effects of high altitudes. In the experiments of the Minetti research group, both Nepalese porters and Caucasians experienced a drop of 24% of the maximum metabolic power at an altitude of 5050m asl. Temperature decreases with altitude, and according to Langmuir (2004, 102), the daily energy requirements increase in cold weather conditions. This is an argument for including the altitude into the cost calculation for landscapes where large variations in elevation are present.