Influence of Rapid Transit on Accessibility Pattern and Economic Linkage at Urban Agglomeration Scale in China

3.1.1 Accessibility Model at Regional Scale

Temporal distance accessibility

By calculating the shortest weighted distance from each grid to a destination with the shortest path algorithm based on raster data, we may obtain the time-based accessibility. The ArcGIS ‘Cost Distance’, a spatial analysis tool, is efficient in measuring the least accumulative cost for each cell to the nearest source based on raster datasets [21]. Using a cost raster graph, along with data from the administrative stations of research units, a layer of the shortest time distance from each raster to a destination is obtained through the cost distance weighted function, following which the raster costs of other administrative stations are extracted and averaged to express the time distance-based accessibility of the destination.

Building a cost raster graph is important for the model of time distance-based accessibility. The study area was divided into regular raster grid cells with a spatial resolution of 50 m × 50 m, and each cell was assigned a related time cost as an attribute. Based on the speed determined by previous studies [22, 23], combined with actual transportation data from South Jiangsu, the velocities of each cell are shown in the following table.

Economic distance accessibility

Economic distance-based accessibility refers to the time cost required for the route along which the minimum cost is incurred. Through inquiry of the railway timetable and long-distance bus timetable, the fare for each public transport line in South Jiangsu and the corresponding road mileage are obtained. The fare index is defined as the average fee per unit of road mileage and is calculated as follows:

where I refers to the fare index, P_i refers to the fare of the i^th sample in the event of travel by means of transportation, S_i refers to the corresponding distance, and n refers to the number of samples. To estimate the fare, only public transportation by passenger is considered, excluding travel by private car, freight, bike, and foot. The price difference arising from vehicle conditions is ignored, and the fare of a hard seat or second-class seat is considered in the event of travel by rail.

Through construction of a vector-based traffic network dataset, values are assigned to the fares of different means of transport (as shown in Table 2). The GIS-based closest facility analysis method is used to seek n-1 paths along which the minimum cost is incurred from the administrative station of a research unit to other administrative stations. By converting the minimum cost to the required cost time, the economic distance-based accessibility of an urban node may be expressed as the average obtained.

3.1.2 Accessibility Model of Regional Center City

Selecting Nanjing as the research object, this paper further discusses the effect of high-speed rails on the accessibility of a metropolis. Nanjing is a central city and an important hub of communication in South Jiangsu. Considering that multiple high-speed lines stop in Nanjing and that the station of origin for both the Beijing-Shanghai HSR and the Shanghai-Nanjing HSR isNanjing, it is logical that Nanjing is selected as the regional HSR hub city for this study.

Based on the general trip rule, the time cost for different segments along the trip is calculated, through the separate activities along the trip. HSR travel can be regarded as a three-phase travel chain based on a door-to-door approach [24]. The trip cost is the sum of the costs of the four parts between origin and destination, which can be formulated as below:

where T is the total time of the trip, TO and TD are the time consumed in the internal transportation network of the starting city and destination city, respectively, TOD refers to the time consumed internally in the HSR system between stations, and Tt refers to the transfer time at both railway stations, which is set to 1 h [25].

This study adopts an accessibility evaluation model called the minimum distance partition based on GIS. To divide the sub-regions,we establish the principle that people tend to choose the nearest HSR site. Through sub-regional division, the accessibility of each sub-region and internal sub-region is calculated separately. Then, the results are added together to acquire the accessibility evaluation

results from any node to the central city. In the model, the accessibility of the sub-region is based on HSR schedules where the shortest time between two points is calculated. The accessibility of internal sub-regions is then calculated through the creation of a regional mesh, with the running speed set for each grid lattice cell, yielding the results for the cost distance algorithms. The main steps can be seen in Figure 2 and is described as follows:

Figure 2

Technique routine of the accessibility model of regional center city

1. Rasterize the road features to generate cost grid charts (as defined in Table 1) with and without HSR, and use the cumulative cost distance algorithm to generate n subdomains;

2. Use the Cost Distance Toolbox in ArcGIS to calculate the cost grid charts of the intra-site’s accessibility, t₁, for each separate subdomain;

3. Calculate cost grid charts of accessibility, t₂, for the entire area of Nanjing without HSR and extract the cost value, T₁, from the Nanjing station to the Nanjing south railway station;

4. Query the train schedules to obtain the shortest time, T₂, for each site to the Nanjing station or the Nanjing south railway station and add it to T₁, with the results being assigned to the subdomain to generate cost grid charts t₃;

5. Add cost grid charts, t₁ and t₃, to the space congruence function of the GIS, which yields the cost grid charts of accessibility, t₄, for the entire area of Nanjing due to HSR;

6. Use the Cell Statistic function in ArcGIS and the minimum algorithm between t₂ and t₄ to obtain the final cost grid chart, t₅.

4.1.1 Accessibility Pattern at Regional Scale

Spatial distance, time distance, and economic distance are fully considered in the region-oriented accessibility evaluation model. Through averaging the three distances of each urban node on a weighted basis, the accessibility of each of the 31 research units in South Jiangsu is obtained. Spatial interpolation is carried out with the tension spline function interpolation method, and it is divided into several grades at the same interval (10 min). The result is shown in Figure 3.

Figure 3

The contour of the accessibility value at regional scale

There are significant spatial differences in the effect of high-speed rail on the accessibility of South Jiangsu. The accessibility is high in the northern and central areas and low in the southern and marginal areas. Overall, from least-accessible Suzhou, Wuxi, and Changzhou to areas outside these three cities, the accessibility gradually increases, presenting an irregular annular distribution. With the launch of high-speed rail, accessibility in the northern and central areas have greatly improved, where over half of the low-accessibility research nodes are located (less than 120 min). After the implementation of high-speed rail, those along the rail lines have become more accessible than the geometric centers, and the areas around the Shanghai-Nanjing railway line and Nanjing-Hangzhou railway line have greatly improved in accessibility. Thanks to high-speed rails, highly accessible corridors are created, yielding the corridor effect.

4.1.2 Accessibility Pattern of Regional Center City

Isochrone maps display areas of similar travel time from a selected starting location, and it reflects the degree of compactness of the spatial association between the central city and the neighboring regions. The uneven pattern of change in accessibility is measured by the range, intensity, and direction of the isotime circle. In this study, based on a 30-minute time cost, the entire area was divided into 7 isochronous rings—30 minutes, 60 minutes, 90 minutes, 120 minutes, 150 minutes, 180 minutes, and more than 180 minutes. From Figure 4, we know that with the operation of HSR, isochronous rings lapse outward with the HSR line, as the 90-minute isochronous ring along the Nanjing-Hangzhou high-speed line extends to Liyang and Yixing, and the 120-minute isochronous ring along the Shanghai-Nanjing and Beijing-Shanghai high-speed lines extends to Kunshan. In other words, the operation of HSR clearly expands the ranges of the 90- and 120-minute isochronous rings. Along HSR, isotime circles clearly move outwards and their spatial scale extends axially. The analysis suggests that the effect of the HSR is much more significant in intermediate and long distance regions.

Figure 4

Isochronous rings and change rate of accessibility value at regional center city scale

The higher the variation rate of the accessibility value, the greater the distance from Nanjing is and the less the distance to the HSR station is. While the compression efficiency of travel time in each city is remarkably different, the compression rates are higher along the HSR sites (e.g., change rate in Liyang, Yixing, East Wuxi and South Kunshan station exceed 30%), which are the major beneficiaries of time convergence. Areas with low change rates—less than 5%—are distributed, for the most part, in the peripheral regions of Nanjing. Due to its proximity, the HSR has less impact on the compression effect in those regions. In addition, the change rates of accessibility are low toward the outer edges of the study area and in areas away from an HSR station, such as Wujiang and Changshou. Thus, HSR services are intended to be medium- and long-distance connections.

5.1.1 The unfairness of HSR effect and its potential influence

The topic of accessibility is considered in the context of the transformation from qualitative research to orientation analysis of geography. Accessibility has been measured according to a place-based perspective [29]. Despite increasingly deepening research on accessibility and other measuring indexes, characterization of the effects, practical significance, and intelligibility of indexes are quite different. This study attempted to obtain the routes along which a minimum cost was required regardless of time and distance, through the estimation of fares, and measured the evolvement of the accessibility of South Jiangsu, as a function of space, time, and cost.

According to the result analysis, HSR has significantly improved the external accessibility of cities but such improvement is not evenly distributed among various cities. It has demonstrated the existence of an uneven space-time convergence on the regional and central-city scales. On one hand, a low-to-high accessibility pattern is exhibited from north to south and from center to edge in South Jiangsu, and the cities where high-speed railway stations exist enjoy more benefits from the change in accessibility. On the other hand, an inequality in the accessibility distribution has been found in the regional central city. Statistically, isochronous rings of 90 minutes and 120 minutes have significantly expanded in scope due to HSR. The 120- to 150-minute rings suffered a substantial reduction in area, suggesting that the influence on the medium or distant areas were particularly more violent and outstanding.

It is not difficult to debate the effects of HSR from the above discussion. Are there siphon effects making neighboring cities lose profits? Are there diffusion effects making neighboring cities receive greater profits? The accessibility effect could result in regional economic cooperation eradicating the traditional geographic boundaries and, thus, promote greater competition among cities [30, 31]. This reminds us that some small cities and less developed areas in this region have not profited significantly since HSR came into service, but rather were left further behind, which requires much attention from local governments.

5.1.2 More time is consumed outside the railway system in the total travel, which calls for policy decisions

After researching the accessibility of railway stations in Holland, Rietveld [32] concluded that the market potential for railway service largely depends on the quality of the entire journey from origin to destination. That is to say, the quality of neither end of the travel chain has been improved in nature because of the operation of HSR. Does this summary equally apply to HSR in China? To determine this, we separated the time cost of internal city transport (TO and TD) from the accessibility value, and calculated its proportion (see Figure 6) It is evident that the proportion of the internal city’s transport time cost is greater than 0.5 at more than half of the research nodes. These cities are usually nearer to central cities. Furthermore, there are larger

Figure 6

The proportion taken by internal transportation time in the whole travel time

differences between the time cost for HSR (TOD) and the time cost for internal city transport (TO and TD). We can take Lishui and Jurong as examples. It takes 15 minutes on the HSR, but more than 45 minutes to reach and leave the HSR stations.

We can further conclude from above discussion that the time saved by HSR will be offset by the distance between downtown and an HSR station that is far from the city center. The accessibility of regions that are well connected in transportation with HSR stations has largely improved, while the accessibility of regions that are poorly connected in transportation with HSR stations has improved with reduced effects. Therefore, internal transportation of neighboring cities must be improved to achieve seamless connections between HSR stations and other transportation stations, reducing transfer time. Marginal regions can improve their accessibility by building other fast transportation channels, connecting with HSR stations and strengthening the transportation connections with HSR. Deriving a comprehensive post-HSR planning strategy for a more integrated transportation system would be useful for local authorities [33].