Approaching past

movement: Time to
move a little further?
Dimitrij Mlekuž

Tuesday, January 4, 2011

 Time geography
Space, time and human
conditions. In: Dynamic
allocation of urban space, A.
Karlqvist et al. (Eds.), 1975.

Pred A., (Ed.) Space and

time in geography: Essays
dedicated to Torsten
Hägestrand., 1981.
Torsten Hägerstrand

Tuesday, January 4, 2011

 “Aquarium”, Path, Bundle, Prism

ism
p r

time
bundle
time

path

potential
path
area

space
space

Tuesday, January 4, 2011

 “ While people often move in cyclical patterns in
the course of routine activities, returning to the
same location again and again ... the places ... are
themselves continuously being physically altered
and decaying, as well as continuously being re-
evaluated and re-interpreted
Thomas 1996:90

Tuesday, January 4, 2011

Tuesday, January 4, 2011
 “ Landscape is time materialized.
Or, better, landscape is time materializing:
landscapes, like time, never stand still
Bender 1996:90

Tuesday, January 4, 2011

 Time-space prism
A B
Time

f(A)
t PPA(t)=f(A,t)∩p(B,t)
future
surface past
surface
p(B)

PPA Space
Tuesday, January 4, 2011
 Time-space prism

!
Tuesday, January 4, 2011
 Potential Path Area

Tuesday, January 4, 2011

 PPA Intersection

Tuesday, January 4, 2011

 Time-space prism of return trip
Time A
t0+2d

t0+d

t0
A

PPA Space

Tuesday, January 4, 2011

 Cost surface Isochrone
Future/past potential
surface path area

Tuesday, January 4, 2011

 Potential path area field

ij
PPAij

Tuesday, January 4, 2011

 Potential path field

PPAij+1 PPF=∑PPAij
ij
ij+1
ij
PPAij

Tuesday, January 4, 2011

 Potential path field
5 min time budget

High

Low
Tuesday, January 4, 2011
 Potential path field
15 min time budget

High

Low
Tuesday, January 4, 2011
 Potential path field
30 min time budget

High

Low
Tuesday, January 4, 2011
 Prominence filter
in the comparison. Many criteria may be used to define what is a neighbourhood
or what locations are part of it, here a neighbourhood will be defined by reference
to an arbitrary Euclidean distance around any location. The area, comprising all
locations within this radius, will define the neighbourhood at any location. Once a
neighbourhood is defined, it is possible to produce a simple definition of prominence,
Prominence of a property p, at any location i, Prom(t ) is defined as the average difference
p
between the property at that location p(i ), and that found at each of the other locations
j , p( j ), within an arbitrary neighbourhood of i, N , such that given an arbitrary distance
n n i
where N is the neighbourhood of i, N=Card(N ) and nµN (Natural numbers).
i i
∑ p(i )−p( j )
n
i Prom(i ) = jn µNi (2)
p N
The values for the prominence at any location can have any positive or negative
value. The upper and lower bounds will change with each neighbourhood size. The
value and sign of the prominence reflects the morphological character of the location.
For example, when calculated using a DEM, i.e. altitude being the property that is
being compared, higher positive values tend to indicate a sharper hill-top, while
more moderate values point towards a more rounded hill-top; values close to zero
indicate flat locations and negative values channel- or pit-like locations. It is import-
Llobera 2003, 37
ant to note that the definition at this stage does not include any sort of normalization.
This is because the index may be normalized in various ways; the normalization of
the entire raster by the maximum prominence value produces a result that is informat-
ive, if the analysis is restricted to one image and to its entirety. However, if the

Tuesday, January 4, 2011

 30 min time budget
500 m radius

High

Low
Tuesday, January 4, 2011
 30 min time budget
250 m radius

High

Low
Tuesday, January 4, 2011
 30 min time budget
125 m radius

High

Low
Tuesday, January 4, 2011
 30 min time budget
75 m radius

High

Low
Tuesday, January 4, 2011
 5 min time budget

75 m 125 m 500 m
High

Low

Tuesday, January 4, 2011

 30 min time budget,
multiscale

Tuesday, January 4, 2011

 Prominence
behaviour
= 510 m) (a) Increasing (b) Decreasing

Prominence

Prominence
0– 60– 70– 80– 90– Radii Radii
0% 70% 80% 90% 100%

nce
(c) Stable (d) Oscillatory
BA Round Barrows

IA Square Barrows
Prominence

Prominence
nce within 510 m.

er percentages found
tion of linear ditches
gh this concentration Radii Radii
are barrows presents Figure 11. Types of topographic prominence behaviour.
ne peak at 10–20%
0%.
), clearly points out The ‘‘behaviour’’ of the topographic prominence at 2003, Figure 11
Llobera
topographic promi- any location changes from one radius to another.
Tuesday, January 4, 2011 Adopting a very simplistic approach, it is possible to
 30 min, prominence
behaviour

Increases with distance

Stable

Decreases with distance

Tuesday, January 4, 2011

 125m radius;
change over
isochrone time

Increases with time

Stable

Decreases with time

Tuesday, January 4, 2011

 “ Animals and people see as they move and
move as they see.
Gibson 1979

Tuesday, January 4, 2011

 Visualscapes
(visibility field)

High

Low
Foreground Midground Background

Tuesday, January 4, 2011

 30 min, midground

High

Low

Tuesday, January 4, 2011

 Visually prominent
Prominent
accessibility
Both
Tuesday, January 4, 2011
 Conclusions

“
region:

an area concatenated by peregrinations between

the places it connects
Casey 1996: 24

Tuesday, January 4, 2011

 “ Ways of life are not therefore determined in advance, as
routes to be followed, but have continually to be worked
out anew. And these ways, far from being inscribed upon
the surface of an inanimate world, are the very threads
from which the living world is woven.
Ingold 2000, 242

Tuesday, January 4, 2011