Is transport infrastructure serving population and urban growth?

July 7th, 2014, Published in Articles: PositionIT, Featured: PositionIT

 

This study looks at whether or not transport infrastructure is serving population and urban growth in the Johannesburg region. It takes a multifaceted approach and includes spatial analysis, information derived from the Quality of Life (QoL) surveys by the Gauteng City Region Observatory (GCRO), as well as literature reviews.

Fig. 1:  Revised PPDAC process.

Fig. 1: Revised PPDAC process.

The population of Gauteng has grown considerably and with it has been a corresponding large-scale urban growth. The province has experienced an above average increase in traffic due to the development of housing, offices, retail and industrial properties over the past ten years. Unfortunately, provision of road infrastructure has not kept up with the increased traffic demand, resulting in a road and freeway network that is over capacity [1]. Concentrated population growth has resulted in strong spatial polarisation, urban sprawl and tracts of under-utilised land between main urban centres [2].

This article considers whether or not existing transport infrastructure and new transport initiatives are well placed in order to successfully serve the needs of Johannesburg’s growing urban population. It draws on multiple sources of data and information including spatial analysis through the use of geographical information system (GIS) software and in particular, the Esri ArcGIS package, results from the Gauteng City Region Observatory (GCRO) Quality of Life Survey (QoL), and literature reviews from a variety of sources, and combines them in order to draw conclusions. A structured approach and methodology is utilised to examine all the data. The assumption is made that the sample size of the QoL survey is representative of the entire Johannesburg region.

The results were achieved by successfully modelling urban growth, identifying the correlation between urban growth and existing and new transport infrastructure, including corresponding patterns, and providing a conclusion as to whether transport infrastructure is correctly placed for current and future growth. New transport initiatives such as the Bus Rapid Transit (BRT) system, the Gautrain and Gauteng Freeway Improvement Project (GFIP) are also examined.

Background

The planning and coordination of transport infrastructure and urban development is made difficult due to the spatial structure of South African cities that is a legacy of Apartheid. Adding to this challenge are the unique socio-economic factors such as high unemployment, the large number of informal dwellings, the state of the public transportation system, high levels of crime, and the lack of reliable information from the taxi industry.

Despite the significance of the connection between land use systems and transport systems, South African cities have typically continued to be planned and managed by different sets of professionals, working within or for different government departments, operating within different statutory planning frameworks, and producing different plans and improvement programmes implemented through different budgets [3].

Existing solutions

As mentioned, the unique socio-economic factors facing South Africa have made it difficult to adopt transportation models and strategies used in other cities and regions around the world. Central and local government have a number of strategies and proposals at their disposal. An escalation in the number of infrastructure programmes being implemented is testament to the government’s realisation that a lack of a polycentric metropolitan transport system has limited inter-firm linkages, agglomeration economies and intra-regional trade. Numerous projects have been launched to connect the region through additional bypasses, rail links and road improvements [2].

However, it must be noted that worldwide a large number of urban mass transportation theories, approaches, models, policies, plans and strategies have been experimented, modified and implemented with mixed results [4]. In order to successfully determine whether or not transport infrastructure is sufficient, reliable and accurate information is required. This proves to be difficult with the unorganised and disjointed taxi industry in South Africa that still carries approximately 60% of commuters [5].

The conventional Four Step Model (FSM) has dominated the history of travel demand modelling in South Africa [6]. The FSM has however drawn criticism over the past decades and in particular its application in developing countries and is deemed unsuitable due to its shortcomings [7]. A number of solutions and approaches have been taken to address the situation in South Africa, including a prototype model of the UPTrans model for the Gauteng Global City Region. The UPTrans study modelled travel behaviour of Gauteng residents utilising the Multi-Agent Transport Simulation (MATSim), and a replacement for the traditional FSM was developed using the Transportation Analysis and Simulation System (TRANSIMS) [8, 9].

These existing solutions do not specifically investigate the link between transport infrastructure and population and urban growth. An initiative aimed at overcoming the many challenges in this sphere was the creation of the GCRO, that is responsible for building a knowledge base through the collection of data and benchmarking the city-region, providing policy analysis and support through the application of applied research.

Overview

A multifaceted approach is taken in this study due to the fact that any attempt at looking at a singular aspect in isolation runs the risk of drawing an incorrect conclusion, as important information may not have been considered. It must be noted that this approach is presented to provide an overview of the methodology used in attempting to answer the question of whether or not transport infrastructure is addressing the challenges of urban growth and population growth.

Methodology

Fig. 3: Walking time to closest public transport access point in Johannesburg. ??

Fig. 2: Walking time to the closest public transport access point in Johannesburg.

I many instances the process of spatial analysis follows a number of well-defined (often iterative) stages: problem formulation; planning; data gathering; exploratory analysis; hypothesis formulation; modelling and testing; consultation and review; and ultimately final reporting and/or implementation of the findings [10]. The methodology utilised is the sequential five steps statistical method: plan, problem, data, analysis, conclusion (PPDAC) [11].

The original PPDAC process as proposed by Mackay and Oldford suggested a relatively linear progression from the problem definition through to the conclusion. Although spatial analysis as a whole may be considered to follow a very similar process to that described by Mackey and old ford [11], a revised approach of the PPDAC process is extended in order to cater for spatial analysis as is shown in Fig. 1 [10]. PPDAC is a flexible and dynamic methodology, not a rigid set of procedures or forms, and thus may be applied at several stages of a project [10].

Spatial analysis

A data reduction technique is required for the large amount of data that is considered for this study. Numerous methods exist for GIS data reduction. Prominent methods include GIS database visualisation, using algorithms to compress and reduce data, querying, buffering and windowing. All the methods listed above are utilised through the geoprocessing tools available in ArcGIS and form part of the PPDAC process at various stages [12].

Data gathering

The first phase of the process dictated that the focus areas of the project are clearly defined. These areas include: population, urban growth in the form of building counts and transport network infrastructure. Numerous datasets that contain information related to these areas were obtained. 1.

Data reduction
The first phase of the data reduction consisted of GIS database visualisation. Data integrity and quality was an extremely important aspect to take into account. Only the most recent and complete datasets that are available at the time are utilised in the study. The datasets utilised are displayed in Table 1.

Data processing and manipulation

Once the relevant datasets are identified, the process moves on to the extraction of relevant and useful data. ArcGIS has the functionality to perform structured Query Language (SQL) expressions and queries. Queries are based on selecting only attributes pertaining to urban growth and the area of focus. For example, the datasets utilised to model urban growth are the Eskom Spot Building Counts for 2008 and 2010 [13, 14]. These datasets contained the most recent and reliable building counts from the data analysed. The datasets contained the locations of formal dwelling units and non-dwelling structures, as well as informal areas. The 2008 and 2010 datasets are made up of different shapes with the 2008 data contained as polygons and the 2010 data as points. The geoprocessing tools contained within ArcGIS were utilised to combine the two datasets. A spatial join is conducted whereby any points from the 2010 set that fall within a polygon from the 2008 set are combined.

The data from 2010 supplements the 2008 data and the results takes into account any dwellings that may no longer exist in 2010. The data is manipulated further in order to achieve the desired output.

Name Year Company
GTI demographics 2009 2010 GeoTerraImage
GTI demographics 2001 2001 GeoTerraImage
Growth indicator (2001 – 2009) 2009 GeoTerraImage
Passenger rail stations 2004 Gauteng Dept of Economic Development: Development Planning
Taxi ranks 2004 Gauteng Dept of Economic Development: Development Planning
SPTN Bus stops 2008 Gauteng Dept of Roads and Transport
SPTN taxi ranks 2008 Gauteng Dept of Roads and Transport
SPTN railway stations 2008 Gauteng Dept of Roads and Transport
Eskom Dwelling Inventory / Spot Building Count 2008 CSIR Satellite Applications Centre
Eskom Dwelling Inventory / Spot Building Count 2010 CSIR Satellite Applications Centre
Demprokey-X 2011 Lightstone

Table 1: Datasets.

Thematic maps

Through the extraction of the relevant data, thematic maps are created that describe useful information. Multiple layers are placed onto the maps in order to display the relevant information. This includes a base layer of the Johannesburg region subdivided into either enumeration areas or sub places, which are varying levels of the divisions of municipalities, and is dependent on the level of detail to be shown. Density plots of urban growth are created with transport networks overlaid. This allows high growth areas to be identified. In terms of this model, urban growth is purely defined as the change in density of buildings and dwellings.

Quality of Life survey

The GCRO conducted two “Quality of Life” surveys; one in 2009 and the other in 2011. The intention of these surveys was to analyse the quality of life of citizens, identify key areas and groups needing intervention and support, and provide a holistic assessment of life in the Gauteng City-Region. A multi-stage Probability Proportional to Size (PPS) sampling approach is used for both surveys with error bars of 1,3% and 0,7% respectively [15, 16]. These surveys provided useful information in terms of transport patterns and habits, and the public perception of certain transport infrastructure. Both surveys were weighted according to ward boundaries. Weighting is used to correctly proportion the sample across the sample frame, which is the ward boundary, and weighting by adult population is used in this case [17]. The transportation sections of the survey are analysed. The IBM
SPSS Statistics software package is utilised to extract the relevant data from the surveys.

Literature review

A number of documents and literature sources provide useful information in cases where spatial analysis and the QoL surveys were not useful, or the required data was not available. In conforming with the PPDAC process, focus areas were identified and included the following documents: CoJ: Framework For Non-Motorised Transport, CoJ: 2011/2012 Integrated Development Plan, The Growth and Development Strategy for the CoJ, Gauteng Land Transport Framework and Gauteng 25-Year Integrated Transport Master Plan [18 – 22].

Information on the new transport initiatives that include the BRT, Gautrain and GFIP, was primarily obtained from literature reviews. Information integrity and quality is extremely important when looking at literature reviews and where possible verification of information is done. Information obtained from sources such as the City of Johannesburg (CoJ), Department of Roads and Transport: Gauteng and National, Gauteng Provincial Government, South African National Roads Agency Limited (SANRAL), GCRO, Engineering News and reports from stakeholders directly involved in transport initiatives are considered credible and reliable sources for this study.

Key results and findings

Spatial analysis

The spatial analysis of urban growth indicates high growth areas primarily in the north of the city with moderate growth occurring in the central and southern regions. As mentioned earlier, urban growth is defined as buildings/dwellings density. This was done in order to show population growth separately and then compared to urban growth.

Some of the thematic maps indicating areas of urban and population growth with transport networks overlaid were consulted. These also indicate that there is sufficient existing transport infrastructure in high growth areas and the Gautrain and BRT routes service these areas. Thus existing infrastructure is well placed to serve the needs of the populace and urban areas. Sufficiency in this context is based on access to a public transport access point. The analysis shows that 86% of the populace falls within a 1 km radius of a public transport access point.

Public transport access points include bus stops, taxi ranks, train stations, Gautrain stations and BRT stations [23]. An interesting observation is that the Gautrain bus system covers large sections of northern Johannesburg. The system covers a radius of approximately 15 km around each Gautrain station. It must be noted that the radial buffer is used as a simple means of indicating accessibility and whether or not infrastructure is in the correct position. Further study is required in terms of sidewalk availability, cost of service, frequency of service etc. that have a major impact on the utilisation of public transport.

Quality of Life Survey

The results from the QoL survey were also analysed. A key result obtained from the GCRO QoL surveys is shown in Fig. 2. In addition, according to SANRAL, an independent study conducted by navigation services provider TomTom, the GFIP, where implemented, has improved traffic flow by up to 50% [24].

Gautrain ridership has seen an average growth of 5% since May with approximately 37 000 to 40 000 people using the train each day. Capacity has had to increase by 30% during morning and afternoon peak times. The Gautrain bus system has proven to be extremely efficient with a bus arriving every 12 minutes during peak hours and a punctuality rate of above 90%. This bus system carries an average of 13 000 to 14 000 people a day which is a 100% growth from that of 2011 [25, 26].

The BRT currently transports approximately 40 000 people on a daily basis. The implementation of the system has been delayed due to serious opposition from the taxi industry. The long-term plan is to place more than 85% of Johannesburg’s population within 500 m of a BRT trunk or feeder corridor [25, 26]. According to the CoJ, 11% of BRT passengers were former private car users, 63% were taxi users, 17% train users, and 8% Metrobus users [25, 26].

Discussion of results

Interpretation of the results of spatial analysis can at times be subjective, as it is presented visually. This is why the inclusion of the QoL Survey and literature reviews are important in drawing conclusions from this research paper. Geoprocessing tools within ArcGIS are utilised to verify the results presented in the survey. A number of steps are required for this, including merging all known public transport access points, including bus stops,
taxi ranks and railway stations. Dissolved buffers were then created around each of the points at radiuses of 500 m, 1 km and 2 km respectively. The 500 m and 1 km buffers are meant to model a 10 minute walk as given in, and the 2 km is meant to model a walking time of up to 30 minutes [27].

A trade-off of this research approach is that there is not as much attention to detail of any one particular aspect, but it does provide a more accurate analysis of the entire situation. In particular, the Metrobus Transit System of the CoJ was not thoroughly investigated. It has been noted that the system does cover large parts of Johannesburg [27]. Another drawback to this study is that due to the large volumes of data considered, scoping the study was difficult. The data reduction process was laborious and time consuming.

The BRT has faced much resistance from the taxi industry, which is a primary cause for the delay in its full implementation. The BRT has the potential to provide an effective mass public transportation system that will reduce the general reliance on private vehicles.

The Gautrain covers areas that the existing rail network does not and extends through many of the new urban growth areas in Johannesburg. Although the Gautrain bus system is efficient, it has been noted that many of these buses are under-utilised during their daily operation periods. There is a possibility that if integrated with other existing systems, this bus service could be more effectively utilised. The under utilisation of the bus system could possibly be attributed to the high costs of using the bus system when not using the Gautrain. Also, the Gautrain stations, with exception of Park Station, are located in the northern suburbs of Johannesburg, which are generally considered higher income areas where people would have access to private vehicles.

The rollout of the GFIP has had some delays due to the recent e-tolling saga whereby opposition from civil society has halted the planned tolling of highways. Studies have indicated that traffic flow has improved due to the highway improvements themselves.

The analysis performed in this paper indicates that there is sufficient transport infrastructure in terms of access points in Johannesburg to cater for the needs of a growing population and urban expansion. In particular, public transport access appears to be within reasonable reach for the majority of the general population, but it is not effectively utilised. There needs to be an improvement in existing transport infrastructure with a focus on the existing Metrorail network in particular. This could be due to a number of reasons such as those put forward in the Quality of Life survey. It is however evident that further analysis is required to conclusively answer this question [8, 9].

There also needs to be a focus on shifting the mindset of the public away from the use of private transport. The shift in thinking is evident as the Growth Management Strategy of the CoJ states, “Establishment of an efficient movement system that moves away from private transport dominance to efficient and effective public transport systems that are fully integrated within the urban fabric. Integrated public transport modes will improve accessibility and mobility, at the very least, along the main transport routes within the city” [29].

Recommendations for future work

A number of improvements could be made to the work presented in this article. A study of this nature can by no means be considered to be conclusive. The research problem has simply too many factors that need to be taken into account. Conclusions can only be made utilising the available data at the time.

Using spatial analysis for non-spatial data

The visualisation of non-spatial information (statistics, databases, online books, web pages) is becoming more commonplace in a variety of fields. Non-spatial data often has no corresponding geocoded representation and often the only useful way of presenting this data is through charts and graphs. There is an approach that involves the utilisation of SPSS Statistics and ArcGIS to convert data into spatial form and manipulate it in order to display it on a GIS platform [30].

Analysis of additional information

Reliable information on the taxi industry must be obtained without which a comprehensive analysis cannot be achieved. The CoJ does have limited information on taxis but this was not available to the authors. The recent installation and operation of the SANRAL e-toll gantries should provide valuable information in terms of transportation patterns. The information contained in the 2011 census could prove to be very reliable and useful as it one of the larger and more comprehensive surveys to be conducted in recent times.

Utilisation of ArcGIS tools

ArcGIS contains a number of built-in tools such as the spatial analyst tool, which allows the analysis of spatial relationships, the building of spatial models, and the performing of complex cell-based (raster) operations. ArcGIS also contains a spatial statistics toolbox that allows for the analysis of the distribution of geographic features. Custom scripts can be written within the ArcGIS package to perform any operation that is required. ArcGIS caters for the Python and Visual Basic (VB) languages.

Automated methods of data reduction

The data reduction method presented in this article largely consisted of manually performing the process. If larger sets of data are to be examined, there are a number of processes and procedures that could be used. Principle Component Analysis (PCA) and Factor Analysis could be utilised with software packages such as SPSS and Statistical Analysis Software (SAS) as presented in [31]. PCA is a useful tool in reducing the complexity of data and can be utilised effectively for GIS analysis [32]. Another data reduction method of large databases is the utilisation of algorithms for Knowledge Discovery of Databases (KDD) through a spatial database system (SDBS) [33]. Although the principle of KDD has been in existence for a long time, many modern algorithms exist for this form of spatial data mining.

Conclusion

This article shows that through the utilisation of the described method, there is sufficient transport infrastructure in terms of public transport access points in place to serve the needs of population and urban growth. The existing infrastructure is not effectively utilised and there is a need to improve the reliability, efficiency and safety of using the existing systems. It is however noted that the full effectiveness of transport infrastructure and initiatives cannot be successfully measured without reliable data from the taxi industry. The unique socio-economic conditions, existing spatial patterns and political considerations, especially with regards to the taxi industry, must be factored into account. Further study in terms of the reasons for under utilisation is required to conclusively answer the question.

New transport initiatives such as the BRT, Gautrain and GFIP have had mixed results though they have all shown varying degrees of success. It is, however, too early to determine the effectiveness of these initiatives. Once fully implemented, they have the potential to have a major impact in assisting with the mobility needs of a growing city.

Co-ordinated development strategies in the form of the Growth Management Strategy, Growth Development Strategy 2040, and the Integrated Transport Master Plan indicate that there is definitely a focus on the integration of transport systems and infrastructure. The approach proposed in this article whereby multiple sources of information and data need to be taken into consideration has highlighted the importance of not solely relying on spatial analysis for studies ofthis nature.

References

[1] South African National Roads Agency: Gauteng Freeway Improvement Project. 1 July. www.nra.co.za/live/content.php?Item_ID=260 [Accessed  20 October 2012], (2009).
[2] R Behrens: Centre For Transport Studies, University of Cape Town. Integrated Land Use – Transport Planning. http://cfts.uct.ac.za/study_4a.html [Accessed 21 October 2012].
[3] OECD Territorial Reviews: The Gauteng City-Region, South Africa. OECD Publishing, 2011.
[4] J Chakwizira: The Question Of Road Traffic Congestion And Decongestion in the Greater Johannesburg Area: Some Perspectives. CSIR Built Environment, 2007.
[5] J Dugard and MA Sekhonyane: Violent Legacy: The Taxi Industry and Government at Loggerheads. SA Crime Quarterly. No 10.
[6] DR Diedericks and JW Joubert: Urban Transport XII: Urban Transport and the Environment in the 21st Century. Southampton: Wit Press, 2006.
[7] M Mokonyama and C Venter: “UPtrans: An Incremental Transport Model with Feedback for Quick-Response Strategy Evaluation”, CSIR Built Environment, 2009.
[8] J Valabh: “Is Transport Infrastructure Serving Population and Urban Growth”, Preliminary Report, University of Witwatersrand, 2012.
[9] J Valabh: “Is Transport Infrastructure Serving Population and Urban Growth”, Final Report, University of Witwatersrand, 2012.
[10] MJ De Smith, MF Goodchild and PA Longley: Geospatial Analysis: A Comprehensive Guide to Principles, Techniques and Software Tools. 3rd Edition. Leicester: Troubador, 2009.
[11] RJ Mackay and RW Oldford: Scientific Method, Statistical Method and the Speed of Light. Statistical Sciences. Vol. 15, No. 3: 254 – 278, 2000.
[12] S Deshpande and MA McAdams: “Database Reduction Techniques For Large GIS Databases”, Texas A&M University-Kingsville, 2000.
[13] CSIR Satellite Applications Centre: Eskom Dataset. Eskom Dwelling Inventory/Spot Building Count, 2008 version.
[14] CSIR Satellite Applications Centre: Eskom Dataset. Eskom Dwelling Inventory/Spot Building Count, 2010 Version.
[15] Gauteng City-Region Observatory. May 2010. The Gauteng City-Region through the eyes of its residents. www.gcro.ac.za/project/quality-life-survey [viewed on 25 October 2012].
[16] Gauteng City-Region Observatory: Quality of Life in the Gauteng City-Region: 2011 findings. www.gcro.ac.za/project/quality-life-survey [viewed on 25 October 2012].
[17] Gauteng City-Region Observatory: Quality of Life Survey 2 – Final Weight Calculations, 2011. www.gcro.ac.za/project/quality-life-survey [viewed on 25 October 2012].
[18] City of Joburg: “Framework For Non-Motorised Transport”, 2009.
[19] City of Joburg: “Integrated Development Plan 2011/2012”, 2011.
[20] City of Joburg: “Growth and Development Strategy 2040”, 2011.
[21] Gauteng Department of Roads and Transport: “Gauteng Land Transport Framework 2009 – 2014”, 2011.
[22] Gauteng Department of Roads and Transport: “Gauteng 25-Year Integrated Transport Master Plan”, 2012.
[23] A Memee: “Is Transport Infrastructure Serving Population and Urban Growth?”, Research Paper, University of the Witwatersrand, 2012.
[24] Engineering News: Sanral says study shows GFIP upgrades improved traffic flow. 4 July, 2012. www.engineeringnews.co.za/article/sanral-says-study-shows-gfip-upgrades-improved-traffic-flow-2012-07-04 [Accessed 20 October 2012].
[25] I Venter: Mind The Gap. Engineering News. Vol. 32, No. 25, July 2012.
[26] I Venter: Engineering News. Bombela ponders capacity, extended hours as Gautrain enters year three of operation. 12 October 2012, http://m.engineeringnews.co.za/article/bombela-ponders-capacity-problems-extended-hours-as-gautrain-enters-year-three-of-operation-2012-10-12 [Accessed 10 October].
[27] City of Joburg: MetroBus: Bus Zones And Major Routes, 2006. www.joburg.org.za/index.php?option=com_content&task=view&id=59&Itemid=71&limitstart=5. [Accessed 21 October 2012].
[29] City of Joburg: “Spatial Development Framework (2007/08)”, Growth Management Strategy, 2008.
[30] LJ Old: Using Spatial Analysis For Non-Spatial Data. Esri User Conference Proceedings.California, 2000.
[31] R Agarwal and AR Rao: “Data Reduction Techniques”, Indian Agricultural Statistics Research Institute.
[32] A Demirci and M McAdams: “The use of principle component analysis in data reduction for GIS Analysis of water quality data”, Geography Department, Faith University, 2006.
[33] M Ester, HP Kriegel and J Sandar: Spatial Data Mining: A Database Approach. Proc. of the Fifth Int. Symposium on Large Spatial Databases. Berlin, 1997.
[34] Lightstone: Demprokey-X, March 2011.

Contact J Valabh, Whiterock Tech, Tel 072 312-8000, j.valabh@wrtech.co.za

Related Articles

  • Trial tests harmonised navigation products
  • Cloud-based subscription makes high-def aerial imagery accessible
  • New compact and versatile 4K projectors
  • Landsat infrared instrument ready after stringent testing
  • SA hosts international standard bodies in ISO assembly