Centre for Analysis of Regional Systems
cenars.upol.cz
Inner structure of functional
regions: relationships between
proto-centres
Martin Tomáš
Pavel Klapka
Marián Halás
Martin Erlebach
Concept of a Functional Region
Based on the heterogeneity of geographic space (horizontal flows of
people, goods, energy, information…)
A functional region should aim to maximize:
1) internal cohesiveness
2) external separation
Correctly delineated functional regions can serve as a powerful
planning tool:
- assessment of regional disparities
- labour market policies
- distribution of subsidies
- public transport planning, etc…
Inner Structure of a Functional Region:
basic theory
J. Von Thünen, W. Christaller, A. Lösch, W. Isard:
- conceptualised simple inner structure of a functional region
(centre-hinterland-periphery)
P. Haggett:
- identifies a crucial role of direction, orientation, pattern of
interaction flows as well as an existence of several centres at
different hierarchical levels
Based on these inner characterics, various types of functional
regions can be identified:
functional urban regions
daily urban systems
travel-to-work areas
local labour market areas
Centres of a Functional Region
The centre is the key trait in the inner structure of a functional
region
Existence of at least 1 centre (instances of a region with no core
are possible but very scarce)
Existence of several cores implies four types of relationships
based on incident interaction:
1) indifference
2) cooperation
3) complementarity
4) competition
Centres of a Functional Region:
method
Data basis: daily travel-to-work flows (2001 census)
Output: 160 functional regions delineated using the CURDS
measure
Same source provided also so called „proto-centres“ as a result of
the first step of the multi-stage regionalisation algorithm
Definition of Proto-centres
In order to qualify as a proto-centre, a BSU has to fulfil two limiting
conditions:
1) the labour function of a proto-centre
2) residence-based self-containment of a municipality
Definition of Proto-centres
Both conditions are very modest – the municipalities that fulfil them
cannot be denoted as centres but rather proto-centres
667 proto-centres have met both restricting conditions
The analysis comprises all municipalities fulfilling these conditions
for two reasons:
1) this set has been tested in the first step of the regionalisation
algorithm
2) larger number of proto-centres enables us to capture better the
inner structure of a region according to distribution and intensity of
commuting flows
Hierarchy of Proto-centres
Hierarchy of proto-centres has been determined by the number of
jobs which is the sum of all in-commuting flows into municipality i
plus employed residents in i.
Four hierarchical levels have been identified
Fig. 1: Hierarchy of proto-centres according to the
number of jobs
Source: own calculationSource: own calculationSource: own calculation
Source: own calculation
Relationships between Proto-centres
To assess relationships between proto-centres of functional regions several
steps have been taken:
1) the CURDS measure has been calculated for all pairs within each
functional region
2) maximum and minimum values for the CURDS measure have identified
the strongest and the weakest flow
3) a filter has been used to rule out flows not meeting the relevance criterion
- statistical evaluation of the set of the CURDS measure
- the critical threshold has been set to 0.1
- the number of flows was reduced from 1,942 to 1,132 flows
In order to compare individual intensities, the flows have been relativized
according to the strongest flows recorded in the Czech Republic (which was
considered as 100 %)
Fig. 2: Flow intensities between proto-centres in
functional regions
Source: own calculation
Relationships between Proto-centres:
evaluation
Final identification of the relationship type between proto-centres has been
based on:
1) the values of the CURDS measure between a pair of proto-centres
2) their hierarchical level
The CURDS measure has been decomposed into two parts, one for
direction ij, and one for the opposite direction ji:
Variables X and Y provide relativized data for both directions of interactions
between two proto-centres and are used to sort the relationships into types
Relationships between Proto-centres:
evaluation
If both values of variables X, Y are lower than 0.1 it means that the
relationship between two proto-centres is indifferent
The cooperation is determined on symmetric relationship between proto-
centres
In order to identify this relationship values X and Y have to be numerically
close:
- as an absolute comparison is not possible, the numerical distance
between proto-centres has been expressed by an average proportional
deviation from mean values of X and Y:
where P is the average deviation of X value
from mean values for X and Y
Relationships between Proto-centres:
evaluation
The cooperative relationship is determined by the level of 0.25
In the next step the hierarchical relationship between two proto-centres has
been assessed and cooperative relationships between proto-centres of the
same and different levels identified
The same has been done for complementary relationships
Fig. 3: Typology of relationships between
proto-centres in functional regions
Source: own calculation
Relationships between Proto-centres:
results
The strongest relativized interaction between proto-centres was recorded
for the pair Ústí nad Labem and Trmice
The weakest interaction was recorded for the pair Dobříš and Říčany
Limiting values, as have been discussed above, have produced out of 1,942
pairs:
- 1,018 cases of indifference
- 220 cases of cooperation
- 704 cases of complementarity
Out of 220 cooperative relationships 172 (78 %) occurred at the same
hierarchical level and 48 (22 %) at different hierarchical levels.
The cooperation is more frequent if the hierarchical level of proto-centres is
equal
Relationships between Proto-centres:
results
Out of 704 complementary relationship 356 (51 %) occurred at the same
hierarchical level and 348 (49 %) at different hierarchical levels
The former case regarded particularly the relationships between protocentres
at lower hierarchical levels
Centre for Analysis of Regional Systems
cenars.upol.cz
Thanks for your attention