METHODOLOGY AND APPROACH - dpsa.gov.za · Enumerator Areas (EAs) comprising informal settlements,...

METHODOLOGY AND APPROACH

Geographic Accessibility Analysis Methodology and Approach

Copyright © 2016 Page i


Copyright © 2016 Page ii

T A B L E O F C O N T E N T S

1 METHODOLOGY ................................................................................................... 1

1.1 METHODOLOGY AND APPROACH ....................................................................................... 1

1.2 COLLECTION AND PREPARATION OF DATA....................................................................... 1

1.2.1 SETTLEMENT TYPOLOGY ......................................................................................... 2

1.2.2 THUSONG SERVICE CENTRES AND THUSONG SERVICE CLUSTERS ............... 3

1.2.3 2011 STATS SA POPULATION DATA ........................................................................ 4

1.2.4 2011 STATS SA GEOGRAPHIC DATA ....................................................................... 5

1.2.5 CENSUS DWELLING FRAME POINTS ...................................................................... 8

1.2.6 TRAVEL NETWORK .................................................................................................. 10

1.3 ASSESSMENT OF THE CURRENT PROVISIONING AND LOCATION OF FACILITIES .... 11

1.4 FACILITY LOCATION ANALYSIS TO OPTIMISE THE PROVISIONING OF FACILITIES ... 13

1.5 THUSONG SERVICE CENTRE TYPOLOGY ........................................................................ 14

2 LIMITATIONS AND CONSIDERATIONS ............................................................. 16


Copyright © 2016 Page iii

LIST OF TABLES

Table 1: Settlement typology .................................................................................................. 2

Table 2: Buffer distance for urban and rural areas ................................................................. 4

Table 3: Census Dwelling Frame Points - feature descriptions and categories....................... 9

Table 4: Maximum travel distance per settlement typology .................................................. 12

Table 5: Thusong Service Centre typology ........................................................................... 15


Copyright © 2016 Page iv

LIST OF FIGURES

Figure 1: Methodology and approach ..................................................................................... 1

Figure 2: Settlement typology ................................................................................................. 3

Figure 3: Example of a Thusong Service Cluster ................................................................... 4

Figure 4: Census 2011; Interactive data in SuperCROSS ...................................................... 5

Figure 5: Small Area Layer (SAL) boundaries ........................................................................ 6

Figure 6: Centroid points derived from SAL boundaries ......................................................... 6

Figure 7: SAL areas greater than 300, 500, 800 and 1000 km2 .............................................. 7

Figure 8: Population distribution to EAs ................................................................................. 8

Figure 9: National road network ........................................................................................... 10

Figure 10: Feed-links or centroid connectivity lines .............................................................. 11

Figure 11: Travel network .................................................................................................... 11

Figure 12: Shortest route calculation .................................................................................... 12

Figure 13: Location analysis................................................................................................. 14


Copyright © 2016 Page 1

1 METHODOLOGY

1.1 METHODOLOGY AND APPROACH

1.1.1 The Geographic Accessibility Analysis takes into account factors such as population

size and density, the distance for people to travel to service points, the availability of

roads, as well as population thresholds and capacity parameters of service points. The

analysis uses a movement network to determine travel distance, hence actual travel

distances are used.

1.1.2 The methodology was implemented as follows:

1.1.3 The approach of this study was to optimise the provisioning and location of different

types of Thusong Service Centres to achieve at least 75% population coverage in

each province, except Gauteng and Mpumalanga which already have 97% and 81%

coverage respectively. In order to reduce the cost involved in establishing additional

infrastructure, optimum locations were identified to achieve the maximum possible

population coverage with the least number of additional facilities.

Figure 1: Methodology and approach

1.2 COLLECTION AND PREPARATION OF DATA

This step included the collection of various datasets, as well as preparation of the data

for use in a Geographic Information System (GIS). These included the 2011 Stats SA

Census data, facility data of departments and South African road network data. Facility

data includes the Thusong Service Centres, Department of Home Affairs, Department

of Labour, South African Social Security Agency and the South African Police Service.

A mapping exercise was performed together with provincial managers of the Thusong

Service Centre Programme to verify current locations of Thusong Service Centres in 8

of the 9 provinces (Western Cape was excluded since a previous study was already

done in the Province). A database of existing Thusong Service Centres and service

points of departments were developed.

Step 1

•Collection and preparation of data

Step 2

•Assessment of the current provisioning and location of facilities

Step 3

•Facility location analysis to optimise the provisioning of facilities



1.2.1 SETTLEMENT TYPOLOGY

1.2.1.1 Settlement typologies were defined to differentiate between urban and rural spaces.

As no single official classification currently exists in South Africa to distinguish between

urban and rural spaces, for the purpose of this study the South African Geospatial

Data Dictionary and its application, compiled by South African National Standards

(SANS) served as a standard and guide to differentiate urban and rural areas.

1.2.1.2 Urban type classification is based on dominant settlement type and land use. Cities,

towns, townships, suburbs, etc., are typical urban settlements. Enumerator Areas

(EAs) comprising informal settlements, hostels, institutions, industrial and recreational

areas, and small holdings within or adjacent to any formal urban settlement are

classified as urban.

1.2.1.3 Rural type is any area that is not classified as urban.

1.2.1.4 The CSIR guidelines for the provision of social facilities in South African settlements

were used as a general guideline to subdivide the urban and rural type into 5

settlement typologies based on population totals. Table 1 depicts the settlement

typology that is used in this study.

URBAN

Type Classification Description

Urban Metro Areas classified as urban and situated in Metropolitan areas

Urban Major Urban Town Urban areas with population above 25 000

Urban Urban Town Urban areas with population below 25 000

RURAL

Type Classification Description

Rural Rural Town Rural areas with population above 25 000

Rural Rural Rural areas with population below 25 000

Table 1: Settlement typology



1.2.1.5 Figure 2 shows a map of the 5 settlement typologies.

Figure 2: Settlement typology

1.2.2 THUSONG SERVICE CENTRES AND THUSONG SERVICE CLUSTERS

1.2.2.1 The study includes both Thusong Service Centres and Thusong Service Clusters in

the accessibility calculations. Thusong Service Centres are permanent structures

where services of departments are housed under 1 roof and where services are

provided on a daily basis (Monday – Friday). Thusong Service Clusters are

permanent facilities consisting of 3 or more key departments which are located in close

proximity to each other. These serve as de facto Thusong Service Centres.

1.2.2.2 A Thusong Service Cluster is defined by creating a geographic distance buffer around

service points of the Departments of Home Affairs, Department of Labour, South

African Social Security Agency and the South African Police Service. Clusters are

identified when the buffers of at least 3 departments intersect with one another. (Figure

3).



Figure 3: Example of a Thusong Service Cluster

1.2.2.3 The buffer size is determined by the particular settlement typology type in which

facilities are located. Table 2 lists the buffer distances which were used.

Settlement Typology Buffer Radius

Metro 1 km

Major Urban Town 1 km

Urban Town 1 km

Rural Town 2 km

Rural 5 km

Table 2: Buffer distance for urban and rural areas

1.2.3 2011 STATS SA POPULATION DATA

1.2.3.1 The Census 2011 Community Profiles in SuperCROSS was used in this study to

retrieve population data. SuperCROSS allows filtering of population figures which can

be geographically mapped to a Small Area Layer (SAL) boundary (polygon).



Figure 4: Census 2011; Interactive data in SuperCROSS

1.2.4 2011 STATS SA GEOGRAPHIC DATA

1.2.4.1 The Census 2011 SAL dataset was used to map population distribution in the country.

The dataset contains the smallest geographical areas for the entire country that is

relationally linked to the Census 2011 Demographical Profile Alphanumeric Database.

The SAL dataset consists of 84 907 polygons with an average area size of 13.5 km2.

(Figure 5).



Figure 5: Small Area Layer (SAL) boundaries

1.2.4.2 A centroid point (midpoint of a polygon) was created for each SAL polygon to calculate

travel distances between populated areas and Service Centre(s). (Figure 6).

Figure 6: Centroid points derived from SAL boundaries

1.2.4.3 Although the Census 2011 SAL layer covers the entire country, spatial gaps do

however exist between towns and villages. These gaps occur mostly in the following

provinces; Eastern Cape, KwaZulu-Natal, Limpopo and North West.

1.2.4.4 The Census 2011 SAL polygons were derived by Stats SA through combining multiple

adjacent EAs into larger SAL polygons. These polygons vary in area size. Densely



populated urban areas have several SAL polygons, compared to more sparsely

populated rural areas that usually have less SAL polygons but are larger in area size.

The use of centroid points as the midpoint for the SAL polygons work well in urban

areas, but remains a challenge in rural areas due to the large differences in polygon

size.

1.2.4.5 In an attempt to reduce the overall area size of SAL polygons in rural areas, the

original EAs were used in areas where SAL polygons exceed 500 km2. The following

maps display the geographical spread of SAL polygons across the country where the

area size per SAL is greater than 300, 500, 800 and 1000 km2. (Figure 7).

Figure 7: SAL areas greater than 300, 500, 800 and 1000 km2

1.2.4.6 Since Census 2011 population data is only available on SAL level, a spatial

aggregation function was used to transpose SAL population figures to selected EAs.

The number of Census Dwelling Frame points (consisting of more than 13 million

dwelling points across the country) per EA served as a weighting factor to divide SAL

population proportionally to selected EAs.

1.2.4.7 The sub-dividing of SAL polygons into smaller units allows for better travel distance

calculations to Thusong Service Centres and Thusong Service Clusters. Figure 8



provides a geographic view of the method used to transpose SAL population into

numerous EAs.

Figure 8: Population distribution to EAs

1.2.5 CENSUS DWELLING FRAME POINTS

1.2.5.1 The dataset was captured from the EA Summary Books used during Census 2011.

The EA Summary Books contain a register of all places to be visited during the

Census. (Source: Stats SA_Census2011 Listing Database.docx.) Table 3 shows the

various available categories and descriptions. Only the highlighted categories and

feature descriptions were used for the study.

Feature Number

DB Abbreviation

Feature description Category Count

<IS NULL> 157 128

1 DU Dwelling Unit Residential 10 815

385

2 VD Vacant Dwelling Residential 272 694

3 NDUC Dwelling under construction Residential 113 629

4 DEM Demolished structure Residential 109 153

5 UNOCC Unoccupied Dwelling Residential 471 545

6 OTH Other Other 446 316

7 VS Vacant stand/plot Other 217 544



Feature Number

DB Abbreviation

Feature description Category Count

8 VL Vacant Land Other 111 576

9 PARK Park Recreation 3 226

10 GAR Garage Business 22 856

11 OFF Offices Business 28 981

12 BUS Business Business 80 102

13 PO Post Office Business 3 467

14 SHOP Shop Business 47 198

15 SCH School Community Services

28 262

16 FACT Factory Business 18 361

17 BANK Bank Business 655

18 PLSN Police station (Police cells) Community Services

828

19 FLSN Filling station Business 1 511

20 BS Bottle store Business 2 245

21 CEM Cemetery Community Services

3 899

22 CH_MSQ Church or other place of worship Community Services

45 750

23 CL Day clinic Community Services

1 191

24 OVAL Sports oval, stadium Recreation 5 810

25 HALL Community Hall Community Services

4 916

26 OSUC Other Structure Under Construction

Other 1 204

27 SDI/HOTEL Non-residential Hotel/ Motels/B&Bs/Lodges/Guesthouses

Business 8 467

28 SDI/HOSP Hospital/ Frail Care Centre/Nursing Homes

Community Services

1 139

29 SDI/STRESS Students’ Residence Residential 726

30 SDI/HOME Home for the aged (other than Frail Care Centre)

Residential 4 656

31 SDI/CCARE Child Care Institution/ Orphanage Residential 1 663

32 SDI/WHOST Workers’ Hostel Residential 2 543

33 SDI/SHOST Boarding School Hostel Residential 1 048

34 SDI/INIT Initiation School Community Services

204

35 SDI/CON Convent/ Monastery/ Religious Retreat/ Reformatories

Community Services

168

36 SDI/BARR Defence Force Barracks/ Camp/ Ship in Harbour

Residential 254

37 SDI/PRSN Prison/ Correctional Institution/ Police Cells

Community Services

363

38 SDI/HALL Community/ Church Hall (in cases of refuge for disaster)

Community Services

372

39 SDI/SHEL Refugee Camp/ Shelter for the Homeless

Residential 356

40 MARK Market Business 1 059

13 038 450

Table 3: Census Dwelling Frame Points - feature descriptions and categories


Copyright © 2016

Page

10

1.2.6 TRAVEL NETWORK

1.2.6.1 A national travel network was used to calculate the shortest travel distance between

SAL centroid points and Service Centres. The network was also used in the process to

identify optimal number and location of facilities. The network consists of a road

network, feed-links (or centroid connectivity lines) that connect centroid points to the

road network and a triangular irregular network (TIN) to connect rural settlements in

areas that lack sufficient road infrastructure. The travel network requires spatial

typology, ensuring that all the line segments are connected to a single network. Figure

9 displays the national road network that was used in the study. Various software

packages were used to build the national routable travel network which includes

ArcMap with Network Analyst, Maptitude and TransCAD.

Figure 9: National road network

1.2.6.2 Feed-links (or centroid connectivity lines) connect each SAL centroid point and Service

Centre to the closest line segment of the road network. SAL centroid points can

connect to any road type, except highways. (Figure 10). Connectivity lines that cross

natural barriers will be removed. Natural barriers include rivers, dams, nature reserves

and mountainous areas.


Copyright © 2016

Page

11

Figure 10: Feed-links or centroid connectivity lines

1.2.6.3 A TIN connects SAL centroid points to each other where nearby road infrastructure is

not available. This method creates straight lines between all SAL centroid points in

non-overlapping triangles. A TIN is based on a Delaunay triangulation. TIN lines were

removed if they cross natural barriers such as rivers, dams, nature reserves and steep

mountainous areas. Figure 11 displays the complete travel network.

Figure 11: Travel network

1.3 ASSESSMENT OF THE CURRENT PROVISIONING AND LOCATION OF FACILITIES

1.3.1 Accessibility modelling was done to determine the current population coverage of

Thusong Service Centres and Thusong Service Clusters. Catchment area analysis


Copyright © 2016

Page

12

was done to quantify the demand on each Centre and to identify areas that are under-,

over- or poorly-served.

1.3.2 Population Coverage is measured by calculating the distance people travel along the

travel network to their nearest Service Centre within the maximum allowed travel

distances. The maximum travel distance in urban areas is 15 km and 25 km in more

sparsely populated rural areas. It is assumed for this study that people are more likely

to travel to their nearest Thusong Service Centre. Table 4 depicts the maximum travel

distances per settlement typology that were applied in the analysis.

Settlement Typology Maximum Travel Distance

Metro 15 km

Major Urban Town 15 km

Urban Town 15 km

Rural Town 25 km

Rural 25 km

Table 4: Maximum travel distance per settlement typology

1.3.3 Figure 12 displays the shortest route from one SAL centroid location to the nearest

Thusong Service Cluster. The total travel distance is 24.78 km.

Figure 12: Shortest route calculation


Copyright © 2016

Page

13

1.4 FACILITY LOCATION ANALYSIS TO OPTIMISE THE PROVISIONING OF FACILITIES

1.4.1 Location analysis was done to identify the optimum number and location of facilities

and to identify other possible options for meeting the unserved population demand.

This analysis accepted the current locations of Thusong Service Centres and Thusong

Service Clusters, while determining the minimum number of additional locations that

would be required to increase population coverage to set percentage.

1.4.2 The analysis is a desktop exercise that utilises the data analysis capabilities of

Geographic Information Systems (GIS) to identify optimal locations. For this reason,

the proposed locations identified in the study refer to generalised locations and not

exact locations. Since the population distribution, number of Services Centres and

road infrastructure differ in each province a facility location analysis was done for

separately for each province.

1.4.3 Optimal locations were identified in unserved areas through various GIS processes

including density mapping, Location-Allocation procedures by using maximum

capacitated coverage and target market share functionalities in ArcMap through the

utilisation of Network Analyst. These methods rely on a demand layer (SAL centroids

with population values) and a potential candidate layer. The candidate layer was

derived from densely populated SAL locations. This procedure selects candidate

locations by maximising attendance (SAL centroids) within an impedance cut-off travel

distance (Figure 13). A target market share of 75% was used in combination with a

maximum travel distance of 10, 15 and 20 km to generate a list of possible locations.

In densely populated metropolitan areas, a maximum capacitated coverage of 300 000

was used to identify possible locations that will reduce population pressure on existing

Centres. Preference was given (where possible) to areas where existing service points

of departments are in close proximity.


Copyright © 2016

Page

14

Figure 13: Location analysis

1.4.4 Proposed location was added to the current mix of Thusong Service Centres and

Thusong Service Clusters to calculate the percentage population coverage per

province, based on the maximum defined travel distances in each settlement typology

type. The catchment population of each Service Centre (Thusong Service Centres,

Clusters and proposed locations) was also determined to identify population demand

per Centre.

1.4.5 The proposed population thresholds for Thusong Service Centres, and Thusong

Service Clusters were applied to classify current facilities and proposed locations for

Service Centres as Large, Small or Satellite/Mobile Centres.

1.5 THUSONG SERVICE CENTRE TYPOLOGY

1.5.1 The Thusong Service Centre typology which was determined for the purpose of this

accessibility study is provided in Table 5. The typology was informed by the Thusong

Service Centre business plan and the findings of the accessibility analysis.

Thusong Urban Mall concept

Permanent facility located in a regional shopping mall.

Thusong Service Centres in such malls may provide a

wider range of services, including services of a

specialised nature. It is assumed that people gravitate

over larger distances towards central locations to

access a wide range of products, economic services

and other benefits. Thusong services are generally

provided during shopping hours.


Copyright © 2016

Page

15

Thusong Service Centre

Permanent structure where services of departments are

housed under 1 roof and where services are provided

on a daily basis (Monday – Friday).

Thusong Service Cluster

Permanent facilities of 3 or more departments which are

located in close proximity. These serve as de facto

Thusong Service Centres. In densely populated areas

the facilities of departments could be located in a radius

of 1 – 2 km, while facilities in more sparsely populated

areas could be clustered together within a 5 km radius

from each other.

Satellite Service Centre

Thusong Service Satellite Centre - permanent

infrastructure where all the Thusong Anchor

Departments have established periodic service delivery

points, i.e. operate according to fixed service delivery

schedules.

Thusong Mobile Service

A Thusong Mobile is a Mobile Service where all the

Thusong anchor departments and other relevant

government departments deliver integrated services in

outlying and more sparsely populated areas which have

intermittent small settlement clusters. In these areas it is

not financially viable to operate either a Thusong

Service Centre or a periodic satellite service. It should

be noted that road links (or lack thereof) and road

conditions have a major impact with respect to access

times and/or the delivery of Mobile Services.

Thusong Outreach projects

Thusong Outreach projects serve extreme sparsely

populated areas with no settlement clusters. Such a

service could possibly utilise Community Development

Workers (CDWs) to access far outlying areas where the

population density does not warrant a Thusong Mobile

or Thusong Service Satellite Centre.

Table 5: Thusong Service Centre typology


Copyright © 2016

Page

16

2 LIMITATIONS AND CONSIDERATIONS

2.1 The geographic accuracy of base data is crucial for any spatial analysis project.

Accurate origin locations (population data), destination locations (service points) and a

national street centreline layer are required to measure travel distances and to

calculate coverage statistics.

2.2 This accessibility study relies on the accurate location of service points since Thusong

Service Clusters are identified where 3 or more service points are in close proximity to

each other. Inaccurate service point locations could potentially influence the number of

Thusong Service Clusters and in turn this could influence the coverage figures, as well

as the calculations to identify the optimum number and location of facilities for meeting

the unserved population demand.

METHODOLOGY AND APPROACH - dpsa.gov.za · Enumerator Areas (EAs) comprising informal settlements,...

Documents

Transcript of METHODOLOGY AND APPROACH - dpsa.gov.za · Enumerator Areas (EAs) comprising informal settlements,...