Movement Spaces and Movement Traces - Computational Movement Analysis - page 18

t x -3 t x -2 t x -1

(a) (b)

w 1

scale 1

w 2

scale 2

w 3

scale 3

Fig. 2.4 Cross-scale derivation of movement descriptors. a Systematic variation of the interval
operator width w between black sampling points when computing, for example, speed. b Dropping
speed values for cow #404020 with coarser sampling rates (in ms − 1 ). Adapted from Laube and
Purves ( P13 . 2011 ) (Republished from Laube, P. and Purves, R., How fast is a cow? Cross-scale
Analysis of Movement Data, Transactions in GIS , 15(3), pp. 401-418, 2011, John Wiley & Sons
Ltd, DOI:10.1111/j.1467-9671.2011.01256.x.)

and sinuosity were derived. Methodologically the study illustrates the adaptation of
methods to CMA that proved useful in other geo-disciplines. The methods design
draws on analogies to the classic Fisher et al. ( 2004 ) multi-scale piece “Where is
Helvellyn?” that showed how the computation of slope or the labeling of landforms
may vary with the sampling point spacing, that is the analysis scale.
Given tracking data of cows with a fine sampling rate of a fix every 0.25 s, the
study derived movement descriptors for six temporal scales w = [5s, 10s, 1min,
5min,10min,30min](Fig. 2.4 a). The study indeed found that the results for various
movement descriptors vary with the used analysis scale. For example, it confirmed
earlier findings, for example mentioned in Laube et al. ( P3 . 2007 ), that speed values
drop with coarser sampling, as the straight-line connectors between wide spaced
fixes systematically underestimate the actual path travelled (Fig. 2.4 b). However, it
also became obvious that such cross-scale effects should not be discussed without a
careful consideration of the uncertainty of the original GPS data (see Sect. 2.3.3 ) .
Finally, phenomenon scale refers to the region over which geographic processes
(here movement) occur (Montello 2001 ). The intuitive rule requires that the analysis
scale matches the actual phenomenon scale. However, just as other geographic phe-
nomena, it is often not a priori evident what that phenomenon scale is, or movement
processes can even express characteristics at and across different scales. Think of a
penguin leaving fine grained movement traces on a ice sheet floating on a continental
ocean current. Methods for up-scaling and down-scaling knowledge remain an open
challenge in CMA.

2.3.3 Uncertainty and Data Quality

Uncertainty is another classic theme in GIScience. Uncertainty is an unavoidable
property of the world, information about the world and our cognition about the