9342 Social Applications of GIS

1
Accessibility Modeling
Chris Medlin - GISCA
2
Why Model Accessibility?
Why bother modelling Accessibility / Remoteness?

• Maximise efficiency of service delivery
• Identify gaps in service delivery
• Assists with understanding costs of service
  delivery
• As an aid to future planning
• Others

3
Using GIS to Model Accessibility
Understanding accessibility is important across a
range of disciplines

• Public Transport Planning
• Urban Planning
• Service Delivery and Planning
• Business Planning
• Education Planning
• Recreation Planning
• Health Planning
• More

4
Accessibility Modeling at GISCA
GISCA has been involved in the development of the
following Accessibility Models...

• Accessibility / Remoteness Index of Australia
  (ARIA ARIAtm)
• Australian Bureau of Statistics Remoteness
  Areas
• Access to General Practitioners across Australia
  (GPARIA)
• Access to Pharmacies across Australia (PHARIA)
• Access to Centrelink Services across Australia
  (Centrelink ARIA)
• Metropolitan Accessibility Index (Metro ARIAtm)

5
Urban/Rural vs Accessibility
Are the concepts of Urban vs Rural the same as
Accessible vs Remote?

• Rural areas can have good access to services
• EXAMPLES
• Urban areas can be remote
• EXAMPLES

6
Urban / Rural Measures in the Australian Standard
Geographical Classification (ASGC)
The Australian Bureau of Statistics has a number
of Urban / Rural Measures

• Section of State structure
• Urban Centres/Localities structure

7
Urban Centres / Localities (UCLs)
Groups CDs together to form defined areas
according to population size criteria

• Each UCL represents an aggregation of one or more
  contiguous
  
  CDs.
• Aggregation into an Urban Centre based on
• Population density gt200 persons/km2 (where pop
  gt20,000)
• Proximity to other CDs (eg. surrounded by urban
  CDs)
• Distance to other Urban Centres (ie. gt or lt 3 km
  gap)
• Subjective inspection or air photos (where pop lt
  20,000)
• Urban Localities
• Non-farm population of 200 999
• Minimum 40 occupied dwellings

8
Section of State (SOS)
Uses population counts from the census to class
Census Districts as Urban or Rural

• Each SOS represents an aggregation of
  non-contiguous
  
  geographical areas of a
  particular type.
• SOS is based on UC/Ls
• Groups UC/Ls into the following categories
• MAJOR URBAN Pop 100,000
• OTHER URBAN Pop 1,000 to 99,999
• BOUNDED LOCALITY Pop 200 to 999
• RURAL BALANCE Remainder

9
MelbourneSection of State
10
Accessibility in the ASGC
ABS structures adequately define Urban or
Rural.
But provided no indication of the level of access
to services experienced by these locations.
11
Sometime prior to 1999
Graham Hugo, David Wilkinson and Errol Bamford
put forward the following recommendation to the
Commonwealth Government We believe that to
classify places in Australia at least two
different fundamental concepts require
acknowledgement. The first is that the urbanness
or rurality of a location is a consequence of the
characteristics of the settlement itself,
primarily population density. We believe that the
current ASGC system effectively measures and
classifies urbanness/rurality. However, we
recommend that the ASGC be modified to allow
greater differentiation of cities and towns of
varying sizes. The second concept is that of
accessibility or remoteness. We argue that the
dimension of remoteness/accessibility needs to be
included in the ASGC as a separate, additional
classification structure.
12
Accessibility in the ASGC
ABS structures adequately define Urban or
Rural.
But provided no indication of the level of access
to services experienced by these locations.
This was the seed for the development of the
Accessibility / Remoteness Index of Australia or
ARIA.
The ASGC now has a remoteness classification
based on ARIA referred to as Remoteness Areas.
13
Before ARIA
BUT before ARIA there was RRMA

• the Rural, Remote and Metropolitan Areas
  Classification
• RRMA has its origins in earlier work on an Index
  of Remoteness developed by the Department of
  Primary Industry and Energy.
• This was further refined by DPIE and the
  Department of Human Services and Health who
  produced the 1991 Census edition of the Rural,
  Remote and Metropolitan Areas Classification.

14
RRMA

• Statistical local areas (SLAs) were used as the
  basic building blocks for RRMA.
• Assumed that remoteness is a function of
  straight-line distance between goods and
  services, and hence from the nearest urban
  centres of various minimum sizes
• SLAs in each of the States and Territories are
  divided into three groups
• Metropolitan areas
• Rural zones and
• Remote zones.
• These were further subdivided into seven
  categories. These categories are based on an
  Index of Remoteness that builds upon the work of
  Faulkner and French.

15
RRMA - Remoteness Index
16
RRMA Criticisms

• SLAs are often large and heterogeneous and are an
  inappropriate unit for the calculation of a
  remoteness index.
• All capital cities are placed in a single
  category yet there are vastly different levels of
  service provision between say Sydney and Darwin
• There are some glaring anomalies in the
  classification obtained by the methodology.
  Population centres such as Mildura and Nhulunbuy
  are classified in the same categories of
  remoteness despite having vastly different levels
  of service provision, population and access to
  larger order centres.
• A simple straight-line distance measure does not
  capture all of the dimensions of accessibility.

17
ARIA the Accessibility / Remoteness Index of
Australia

• ARIA is based on the idea that the Geographical
  remoteness of a location can be measured in terms
  of how far one has to travel to service centres
  (towns) of various sizes.
• ARIA is a purely geographical measure of
  remoteness
• ARIA has a high level of precision and spatial
  resolution.
• ARIA measures remoteness by measuring the road
  distance between where people live and the places
  those people travel to in order to obtain goods
  and services.

18
ARIA
A further advantage of ARIA is that the index is
a continuum. Accessibility and Remoteness can be
seen as two ends of a continuum which describe
the ease or difficulty with which one can access
a range of services, some of which are available
in smaller and others in larger centres
Accessible Remote Easy access a wide range of
Locationally disadvantaged - goods and
services very little access to goods and
services
19
BrieflyARIA Methodology

• For each populated locality (11,879 Australia
  wide) the distance to the nearest service centre
  is recorded for each of the five categories.
• For each category (A thru E), the average (mean)
  distance to the nearest service centre is
  calculated (based on distances recorded for
  11,879 locations).
• Then for each of the 11,879 populated localities
  a ratio of the measured distance to average
  distance is calculated for each class of service
  centre.
• Ratios are then thresholded at 3 (ie. 3 times the
  average distance for that category). All ratio
  values range between 0 and 3.
• Addition of 5 ratios to give a Remoteness Value
  (0 -15) for 11,879 populated localities.
• Ratio values then interpolated to a 1km grid for
  all of Australia.

20
Index Creation Overview
21
ARIA Methodology
ARIA begins by classifying Urban Centres whose
population is greater than 4999 into four
categories Class A 250,000 and above Class B
48,000 to 249,999 Class C 18,000 to
47,999 Class D 5,000 to 17,999
ARIA adds a fifth category Class E 1,000 to
4,999
22
ARIA Methodology
11,879 Populated Localities
23
ARIA Methodology
The Minimum Road Distance from each populated
locality(11,338) to the nearest service centre
for each of the four classes was extracted
Assumption that if closer to a larger centre
the distance to the large centre predominates
24
Calculating ARIA

• 800km to A avg 413km
• 200km to B avg 239km
• 110km to C avg 139km
• 20km to E avg 88km
• ARIA Calculation
• Category A score 800/413 1.93
• Category B score 200/239 0.84
• Category C score 110/139 0.79
• Category D score 50/88 0.59
• Total ARIA score
• 1.93 0.84 0.79 0.59 4.58

25

• Remoteness values for all localities within an
  ABS defined urban centre were standardized

26
ARIA - Populated Localities Showing Remoteness
Value
Dark Red -Remote Pale Red - Accessible
27
Interpolation to 1km Grid
28
ARIA Classified into 5 Classes
Highly Accessible 0 0.2 Accessible 0.2
2.4 Moderately Accessible 2.4 5.92 Remote
5.92 10.53 Very Remote 10.53 15.00
29
Aggregating up to other Spatial Units

• Average ARIA values were then calculated for
  standard units of the the Australian Standard
  Geographical Classification
• Census Districts
• Statistical Local Areas
• Postal Areas
• Etc

30
ARIA Value by SLA
31
Other ApplicationsMetro ARIA

• Metro ARIA applies ARIA methodologies to a
  metropolitan setting.
• Instead of measuring the road distance to Service
  Centres (towns) of various sizes, the distance by
  road to a number of different services is
  measured.
• Instead of calculating distance measures and
  ratios for localities, distance measures and
  ratios are calculated for each and every land
  parcel (1.6 million in Melbourne).

32
Metro ARIA - Component Services

• Those services include
• Public Transport
• Bus Stops
• Train Stations
• Tram Stops
• Shopping
• Supermarket
• Major shopping precinct
• CBD
• Health
• Acute Care Hospitals
• All Hospitals
• General Practitioners (data not available for
  Melbourne)
• Education
• Primary Schools
• Secondary Schools
• TAFE Colleges
• Universities
• Finance

33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
Index Construction
Ratio of Measured to Average distance is
calculated for each service for every land
parcel. Final index is derived by weighted
addition of ratios as follows
42
Metro ARIA - Melbourne
43
Metro ARIA - Melbourne
44
(No Transcript)
45
(No Transcript)
46
(No Transcript)
47
Conclusion

• Sufficiently detailed and precise accessibility
  indices
• Are useful for service delivery planning
• Can assist with analysis of service delivery
  outcomes
• Can be used as a variable to help explain
  observed phenomena eg. health studies
• Provide objective geographic measures of the ease
  with which services can be accessed

48
References
Faulkner, HW, French S. Geographic Remoteness
Conceptual and Measurement Problems, Bureau of
Transport Economics, Canberra, Reference Paper
No. 54, 1993 Griffith DA. Development of a
Spatial Model to Quantify Access to Services in
Rural and Remote Areas of Australia (PhD
Dissertation) 1996, Northern Territory
University, Darwin