Next Article in Journal
A Social–Aware Recommender System Based on User’s Personal Smart Devices
Next Article in Special Issue
Site Selection for Pre-Hospital Emergency Stations Based on the Actual Spatiotemporal Demand: A Case Study of Nanjing City, China
Previous Article in Journal
Optimized Spatiotemporal Data Scheduling Based on Maximum Flow for Multilevel Visualization Tasks
Previous Article in Special Issue
Damage Signature Generation of Revetment Surface along Urban Rivers Using UAV-Based Mapping
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
IJGI
Volume 9
Issue 9
10.3390/ijgi9090517
ijgi-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

An Economic Development Evaluation Based on the OpenStreetMap Road Network Density: The Case Study of 85 Cities in China

by
Bo Liu
1,
Yu Shi
1,
Da-Jun Li
1,*,
Yan-Dong Wang
2,
Gabriela Fernandez
3 and
Ming-Hsiang Tsou
3
1
Faculty of Geomatics, East China University of Technology, 418# Guanglan Road, Nanchang 330013, China
2
State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, 129# Luoyu Road, Wuhan 430079, China
3
Department of Geography, Center for Human Dynamics in the Mobile Age (HDMA), San Diego State University, 5500 Campanile Drive, San Diego, CA 92182-4493, USA
*
Author to whom correspondence should be addressed.
ISPRS Int. J. Geo-Inf. 2020, 9(9), 517; https://doi.org/10.3390/ijgi9090517
Submission received: 14 July 2020 / Revised: 15 August 2020 / Accepted: 17 August 2020 / Published: 28 August 2020
(This article belongs to the Special Issue Geo-Information Technology and Its Applications)
Browse Figures
Versions Notes

Abstract

:
The evaluation of urban economies has been one key concern identified by scholars. In the past, most research methods on urban development assessments have been based on statistical data, and the analysis results have been presented in the form of statistical tables. Moreover, the development of urban road networks reflects the status of urban development and spatial metrics, which are obtained from the urban road network which can be used to evaluate the growth of the urban economy. The OpenStreetMap (OSM) is collected through crowdsourcing, and the OSM road network has the characteristics of a simplified and efficient approach to collect data, update data, free available data, etc. Therefore, in this paper, the OSM road network density is used as a spatial metric which is taken as the main study subject, to evaluate the economic development of Chinese cities. In our experiment, results show that there is a significant regression correlation between the OSM road network density and municipal gross domestic product (GDP). For the 85 selected Chinese cities, a total of 71 cities with residuals between −0.1 and 0.1 account for 83.53%, and a total of 79 cities with residuals between −0.2 and 0.2 account for 92.94%. Therefore, it is apparent that the OSM road network density can be used as a spatial metric to evaluate the municipal GDP, and as a result, can be used by local governments and scholars to estimate, evaluate, and forecast the urban economic development of China.

1. Introduction

In recent years, the rapid development of remote sensing, volunteered geographic information (VGI) and other technologies, spatial data acquisition have become easier to employ. As a result, there is more and more research on land-cover, land-use, and urban development using spatial metrics [1,2,3,4,5,6,7,8]. A few different approaches used to represent spatial concepts have resulted in the development of various spatial metrics [1,2,3,4]. More commonly applied metrics used by scholars have included patch size, dominance, number of patches, and density, edge length and density, nearest neighbor distance, fractal dimension, contagion, etc. [4]. Herold et al. [5] used spatial metrics and texture measures to describe the spatial characteristics of land-cover objects within each land-use region as derived from interpreted aerial photographs. These spatial metrics include percentage of landscape, patch density, mean patch size, area standard deviation, edge density, largest patch index, Euclidean mean nearest neighbor distance, Euclidean nearest neighbor distance standard deviation, area weighted mean patch fractal dimension, fractal dimension standard deviation, etc. Herold et al. [6] argued that remote sensing and spatial metrics lead to an improved understanding and representation of urban dynamics while helping to develop alternative conceptions of urban spatial structure and change. In order to cover as many metrics as possible, Reis et al. [7] presented an extended, updated, and more thorough portfolio of spatial metrics to measure the urban growth and urban shrinkage patterns. In their paper, spatial metrics were defined as the quantitative measures used to assess the spatial characteristics of urban settlements. Reis et al. [8] assembled spatial metrics into three groups: landscape metrics; geo-spatial metrics, and spatial statistics. They indicated that landscape metrics such as fractal dimension, shape, mean perimeter-area ratio mean shape, etc., have been traditionally used to quantify several aspects of landscape configuration and composition. Geospatial metrics has mostly been used to measure urban spatial patterns. However, there are metrics that are similar between geospatial metrics and landscape metrics. Moreover, an important difference between the metrics from landscape metrics is that the latter include a set of metrics that evolved in a “top-down” approach. Spatial statistics are metrics based on statistical tools, and these spatial metrics are often used in combination with regression and spatial econometric models. In this paper, the main purpose of this research is to discuss whether the OpenStreetMap (OSM) road network density can be used to evaluate the level of urban economic development in cities. Firstly, using a regression model to establish the relationship between the OSM road network density and municipal gross domestic product (GDP) from 2014 to 2017. Then, the OSM road network density and municipal GDP in 2018 were used to verify the regression model, to discuss whether the OSM road network density can be used as a spatial metric to evaluate the level of urban economic development in cities. Because the OSM road network has the advantage to be an easier approach to collect data, in a more efficient way, update data, and provide free available data, etc. If we can use the OSM road network density to evaluate the level of urban economic development in cities, then this research can help policy makers in China monitor and evaluate their cities towards more transparent and efficient cities. Moreover, the OSM road network density is based on statistics and mainly uses a regression model, therefore, we take the OSM road network density as a spatial metric to evaluate the level of urban development and transparency in cites.
In general, there is a significant correlation between urbanization and the level of economic development, and it seems that each country or region conforms to this rule to a certain degree [8,9,10], and urbanization has been a defining global phenomenon and a key driving force for social and economic development during the past century [11,12]. Despite China being the world’s largest developed country, its urbanization has progressed at an unprecedented rate [13], as urbanization has shifted due to the rise of industries and population in and around cities, facilitating the development of economies of scale [14]. Cai et al. [15] indicated that the improvement of urban infrastructure attracted corporate investments, created new jobs, and led to an influx of labor. Therefore, the proportion of China’s population living in cities increased from 17.9% in 1978 to 58.5% in 2017 [16]. Heshmati et al. [17] calculated a multi-dimensional composite index of urban infrastructure analyzing 31 provinces and six regions in China during 2005 to 2014, and indicated that the economy, employment, human development, utilities, and technology components of urban infrastructure had positive and significant effects on China’s urbanization, and suggested that the government should guide investments to more efficient transportation systems that improve the development of a city.
As described in the above literature, the efficient transportation systems have a positive impact on urbanization, and the higher the urbanization level of a city, the greater the density of its road network, the higher economic level of a city. In order to discuss the relationship between the road network density and urban economic development, based on geospatial metrics and spatial statistics, many researchers have studied the relationship between spatial metrics and urban economic development. Yu et al. [18] explored the role of the motorway in the evolution of spatial economic agglomerations, they indicated that an improvement in the motorway network lead to a higher degree of geographic concentration of economic activities. Jiao et al. [19] studied the relation between road accessibility and economic growth in China from 1990 to 2010. They studied a total of 337 cities in China and explored the bivariate analysis framework of accessibility, economic growth, and increased rates. The analysis showed that there is a significant positive relationship between the accessibility and economic growth of a city, and that the economic increased rates are largely influenced by the change of accessibility. Worku [20] studied the trends, stock of achievements, and impact of the road network on the economic growth of Ethiopia, Africa. He indicated that the impact of the road network was seen to be less strong on the agricultural GDP growth, but had pronounced impact of the industrial and service sector of GDP. Ivanova and Masarova [21] used the series and correlation method to analyze the effects of road infrastructure development on the economic growth and competitiveness of Slovak’s economy, they indicated that the road infrastructure was a prerequisite for economic growth. Beyzatlar et al. [22] investigated the Granger causality relationship between income and transportation of EU-15 countries. They indicated that there was an endogenous relationship between income and transportation. Gao et al. [23] studied the relationship between the comprehensive transportation freight index and GDP in China. They indicated that the volume of freight traffic and freight turnover in China are positively correlated with GDP. Fan and Chan-Kang [24] studied the impact of road investment on the overall economic growth, rural and urban growth, and rural and urban poverty reduction. They indicated that road investments yielded the highest economic returns in the eastern and central regions of China, while the contributions to poverty reduction were the greatest in western China. It can be seen from the above research that the road network can be correlated with GDP, and most researchers have used regression analysis to investigate the relationship between spatial metrics and urban economic development in cities.
In the last decade, OSM has achieved tremendous development with the popularity and development of the Internet. OSM is a user-generated street map, most of the data sources are provided by the public or volunteers [25], and until now, there are more than six million registered members in the world, and more than six billion nodes in the OSM database [26]. Today, the OSM has become available to be applied in many ways, such as 3D city modeling, road updating, etc. [27,28,29,30,31,32,33,34,35,36,37]. For example, based on the free geographic data provided by the OSM project and the public domain height information provided by the Shuttle Radar Topography Mission, Over et al. [27] studied the prospects of using an interactive 3D city model in Germany and pointed out that the point of interest (POI) in the OSM data provides new opportunities for 3D city modeling. Fonte et al. [28,29] used the OSM data and GlobeLand30 image data to process, resulting in a more accurate and detailed land use coverage map, and more details can be shown by this method than any other methods. Mobasheri et al. [34] explored the feasibility of using the OSM data as geo-navigation data in several German cities by using factors such as the number of features and integrity, and the results showed that the sidewalk data in the OSM can be used to route navigation. Zhang et al. [37] explored the relationship between road density and road type diversity based on data obtained from China’s OSM road network in May 2014, taking 340 prefecture-level cities in China as its study area, it is concluded that the OSM road diversity reflects the demand and value of road-related geographic information and it also reflects the interests of users towards employing the OSM geographical information; Goetz [38] used the points of interest data in the OSM to focus on the detailed display in 3D city modeling. Wang and Zipf [39] used an algorithm to extract the building information in the OSM data for modeling, and the building interior details can be displayed by using the proposed method. In the study of path navigation, Bergman and Oksanen [40] took the OSM data and mobile sport tracking data as research objects, the hidden Markov model (HMM) based as a research method, pointed out that the OSM data has feasibility in bicycle path navigation. In terms of geographic mapping, Rosina et al. [41] took Slovenia and Austria as research objects, and added the OSM data to the Copernicus imperviousness layer to improve the population distribution map, drawing methods of the two countries. The experimental result showed that the total error is reduced after adding the OSM data for auxiliary processing, and the addition of the OSM data has certain improvement effects; Zhao et al. [42] studied the evolution of the OSM road network in Beijing from four aspects. Through the experimental analysis, they believed that the development of the OSM road network in Beijing was significantly related to the number of volunteers, and the growth of the OSM road network was very similar to the development process of the real road network. Dingil et al. [43] used the OSM to estimate and analyze the passenger transport energy per person per year of 57 cities, distributed over 33 countries, the results indicating that high private car mode share is a main cause for the high transport energy usage of such cities.
From the above application research on OSM, it is apparent that using OSM data to do research has become more widespread. In this paper, we aim to focus on developing an economic development evaluation for 85 Chinese cities by using OSM spatial metric road network density. The main purpose is to explore the application of the OSM to evaluate the urban economic development of Chinese cities, and take the OSM road network density as a spatial metric to evaluate the level of urban economic development which is measured by the municipal GDP. In our experiment, we selected 85 cities to verify the proposed method, the results show that the correlation between the OSM road network density and municipal GDP are significant. As a result, it is feasible to predict the level of urban economic development by using the OSM road network density.
This paper is organized as follows: the data sources and basic methods are introduced in Section 2 and Section 3, respectively. Experimental results and analysis are reported in Section 4. Conclusions are drawn in Section 5.

2. Study Areas and Data Source

2.1. Study Areas

Since the reform and reopening in 1978, China’s urbanization level has continuously improved. As the pioneer area of China’s economy development, coastal cities and provincial cities have a higher urbanization rate, which has exceeded 60% in 2017 [43]. This is because eastern coastal and provincial cities have comparative advantages in resource adsorption, innovation, transportation, and so on, making them leaders in China’s overall economic and social development [43]. Since eastern coastal and provincial cities are the pioneers of China’s economic and urbanization development, the open-up policy, new technical innovation, and adjustment of input structure methods has improved the urbanization efficiency of China [44,45].
In this paper, we selected a total of 85 cities in China. Among the 85 cities, there were a total of 62 eastern cities, 12 central cities, and 11 western cities. In addition, among the 85 cities, the study included a total of 27 provincial cities and 4 municipalities.

2.2. Data Collection

There are three main types of data: OSM road network, municipal GDP, and urban area in each selected city. More details about data sources and data formats are as follows.

2.2.1. OSM Road Network

Currently, OSM is one of the most successful and popular VGI projects, and has achieved tremendous development. Until now, there are more than six million registered members in the world, and more than six billion nodes in the OSM database [26], and a lot of research on OSM. Because OSM data are collected through crowdsourcing, the quality of OSM has been often discussed, and usually evaluated based on its quality with authority data. Haklay [46] was the first researcher who analyzed and investigated the data quality of the OSM road network for England, UK. Since then, many researchers have analyzed the data quality of OSM for Germany [47,48], France [49], and China [50]. Luo et al. [50] selected three large, medium, and small cities in China, and compared the length integrity of the OSM road network with the Baidu road network, and Google road network. They indicated that the length integrity of the OSM road network is basically consistent with Baidu’s road network and the Google road network, and the length integrity of the OSM road network is better than Baidu’s road network, and the Google road network in some regions. Hecht et al. [51] measured the completeness of the OSM data on buildings by comparing them with official survey data. It is clear from their research that the OSM building data in urban areas, particularly near the town center, achieve a much higher level of completeness. Singh Sehra et al. [52] introduced a comprehensive review of the assessment of OSM data. Some researchers indicated that OSM has a higher location accuracy, completeness, etc., data quality characteristics in urban areas [48,53]. Because, most urban areas with a higher population density inherit larger numbers of contributors, who influence the quantity and quality of the collaboratively crowdsourced OSM objects [45,46,47,48,49,50,51,52,53]. Therefore, in this paper, we use the urban road network of the OSM to calculate the road network density for 85 Chinese cities.
The OSM road network is downloaded from the OSM website (http://download.geofabrik.de/), which is the ESRI Shapefile. In addition, the projected coordinate system used is the Universal Transverse Mercator (UTM) coordinate system. In this paper, data were collected and downloaded from 2014 to 2018. The OSM road network has about 20 road classes such as cycle way, footway, motorway, residential, primary, and secondary, etc.

2.2.2. Municipal Gross Domestic Product

Over the past 40 years, China’s economic reform has been successful, becoming one of the most important economic power engines around the world, and the second largest economy measured by GDP. China’s rapid economic growth has largely depended on abundant use of natural resources, low-cost investment, and labor with support of a high saving rate, and government policies have also played an important role in promoting infrastructure construction [54]. In recent years, China’s economic growth rate has fallen from the double-digit rate from 5% to 7%. China’s economy has entered “The New Normal Economy”. There are three main characteristics of The New Normal of China’s Economy: (a) a shift from high growth rates to medium-high growth rates; (b) an on-going process of optimizing and upgrading the economic structure, and narrowing the urban-rural gap, with higher personal income as a share of GDP, and an increasing number of people benefiting from economic development; and (c) a transition from growth driven by input and investment to one driven by innovation. These characteristics and measures can promote the steady growth of the Chinese economy, enhance development potential, and further unleash market vitality [55]. No matter if China’s economy is in the double-digit growth rate or the current “new normal economy” medium-high growth rate, the GDP is an index used to describe the economic development level of a city, and the efficient transportation system, population, technology, etc., have a positive impact on GDP. Among the 85 selected cities, a total of 5 first-grade cities, 31 second-grade cities, and 49 third-grade cities were identified (https://www.yicai.com/news/5293378.html). Overall, the GDP and urbanization rate of the first-grade cities are higher than the second-grade cities, and the GDP and urbanization rate of the second-grade cities are higher than the third-grade cities.
In our research, municipal GDP data are collected from the National Bureau of Statistics (http://www.stats.gov.cn/), data on GDP are from 2014 to 2018, and municipal GDP unit is in trillion CNY.

2.2.3. Exploring the Urban Area of Each Selected City

The study scope analyzed a total of 85 main urban areas. The study applied the unit of area square kilometers. Take Shanghai for example, the scope of the main urban area is extracted from the Shanghai Bureau of Planning and Natural Resources Department (http://ghzyj.sh.gov.cn/). Figure 1 shows the location and the municipal GDP in 2018 of the selected 85 cities.

3. Methodology

Regression analysis is one of the most commonly statistical analysis methods used to describe the correlation between independent variables and dependent variables [56]. In our paper, we used a regression model to study the correlation between the OSM road network density and the municipal GDP, and employed the OSM road network density as a spatial metric to evaluate the level of urban economic development of cities. In this section, we do not describe the details of the regression model, but only introduce the calculation method of the OSM road network density.

Calculating the OSM Road Network Density of a City

The road network density is an important index for the evaluation of regional road traffic [57,58,59]. The OSM road network density comprises of geographical information that reflect a real-world road network, and an index for assessing the quality of OSM geographic data. In this study, we calculated the OSM road network density of 85 Chinese cities [60,61]:
D i = L i / A i   i ∈ [ 1 , 2 , 3 , … , 85 ] ,
where one main urban area is i , the OSM road network density in the main urban area i is D i , the OSM road network length for the main urban area i is L i , and the area of the main urban area i is A i .
Figure 2 shows the OSM road network in three different grades based on city. The blue polylines are the OSM road network in 2018. The red polylines are the OSM road networks in 2014. Figure 1 shows the cities of Beijing, Shanghai, and Guangzhou as first-grade cities. Compared to the cities of Nanjing, Wuhan, and Chengdu with second-grade cities, and the cities of Guiyang, Haikou, and Lanzhou with third-grade cities, these cities have different levels of economic development and OSM road network density.

4. Results and Analysis

4.1. Fit Analysis

In this paper, we used Equation (1) to calculate the OSM road network density from 2014 to 2018, then applied the OSM road network density data from 2014 to 2017 as the independent variable, and the municipal GDP from 2014 to 2017 as the dependent variable. Moreover, the regression models of the 85 Chinese cities were obtained using the unary linear regression model, and then applied the OSM road network density and the municipal GDP data in 2018 to validate the regression model.
The OSM road network density, the municipal GDP, the regression models, and the coefficient of determination (R2) of the 85 Chinese cities from 2014 to 2017 are shown in Table A1 (Appendix A), the statistics of R2 are shown in Figure 3. Among all the cities, the maximum value of R2 is Guangzhou, and its R2 is 0.999; the minimum value of R2 is Hohhot, and its R2 is 0.0005. The distribution of coefficient of determination is shown in Figure 4.
In this paper, we also used the OSM road network density and population data from 2014 to 2017 as the independent variables, the municipal GDP from 2014 to 2017 as the dependent variable, the regression models of 85 Chinese cities were obtained using a binary linear regression model, and then used the OSM road network density and the municipal GDP data in 2018 to validate the same regression model.
The statistics of R2 are shown in Figure 5, and the distribution of coefficient of determination is shown in Figure 6. Among all the selected cities, the maximum value of R2 is Taizhou, its R2 is 0.9999. The minimum value of R2 is Baotou, its R2 is 0.3353. In this regression model, the R2 of Guangzhou is 0.9996, and the R2 of Hohhot is 0.4842. The distribution of coefficient of determination is shown in Figure 6. In addition, we can see that the R2, which is calculated by the OSM road network density and population, is higher than the R2, which is calculated by the OSM road network density and is consistent with reference [24,62]. Fan and Chan-Kang [24], Savaş [62] indicated that there is a high correlation in cities with road investments, population, and economic growth.
Figure 3 shows a total of 72 cities with significant correlation with coefficient of determination R2 above 0.7, accounting for 84.71%. These 72 cities experienced rapid economic development during the four years from 2014 to 2017, and with their municipal GDP increasing steadily year by year; 5 cities with coefficient of determination R2 below 0.5, they are Shenyang, Urumqi, Dalian, Baotou, and Hohhot, respectively, and their R2 are 0.499, 0.483, 0.342, 0.303, and 0.0005, respectively. The municipal GDP statistical data and the OSM road network density for these five cities are shown in Table 1 and Table 2. It can be seen that the municipal GDP for these five cities does not increase year by year, but the density of the OSM road network increases year by year.
Hohhot shows the minimum coefficient of determination, and the coefficient of determination is 0.0005. We found that the GDP of Hohhot decreased from 317.359 billion CNY in 2016 to 274.372 billion CNY in 2017. However, the density of the OSM road network in the main urban area gradually increased from 2014 to 2017. More specifically, in 2016, with an increase of 0.204 compared with 2015, which is much higher than other years. Baotou and Hohhot, which are located in the Inner Mongolia Autonomous Region of China, have similar patterns. In 2017, the GDP of Baotou was 275.303 billion CNY, a decrease of 111.46 billion CNY from 386.763 billion CNY in 2016. The density of the OSM road network in the main urban area increased from 0.629 in 2015 to 0.852 in 2016, an increase of 0.22, which is much higher than changes in other years.

4.2. Validation of the Model

In order to validate the regression model obtained by the OSM road network density in the above section, we used the real municipal GDP of 85 Chinese cities in 2018 to validate the regression model. The statistical results of the absolute residuals and relative residuals when using the OSM road network density are shown in Table A2, Figure 7, and Figure 8, respectively. The distribution of absolute residual and relative residual results when using the OSM road network density is shown in Figure 9 and Figure 10, respectively.
The calculation method of absolute residual ( R a b s o l u t e ) and relative residual ( R r e l a t i v e ) is as follows:
R a b s o l u t e = P r e d i c t i v e i − R e a l i   i ∈ [ 1 , 2 , 3 , … , 85 ] ,
R r e l a t i v e = ( P r e d i c t i v e i − R e a l i ) R e a l i × 100   i ∈ [ 1 , 2 , 3 , … , 85 ] ,
where one main urban area is i , the predictive GDP of the main urban area i in 2018 is P r e d i c t i v e i , the real GDP of the main urban area i in 2018 is R e a l i .
At the same time, we validated the regression model obtained by using the OSM road network density and population, we also used the real municipal GDP of 85 cities in 2018 to validate the regression model, and the statistical results of the absolute residuals are shown in Figure 11 and Figure 12, respectively. The distribution of residual results is shown in Figure 13 and Figure 14, respectively.
We calculated the difference between the residual, which is obtained by using the OSM road network density, and population. The statistical results of the differences are shown in Figure 15, and the distribution of the differences are shown in Figure 16. The results show 69 cities with residual differences between −0.1 and 0.1, accounting for 81.18%; a total of 76 cities with residual differences between −0.2 and 0.2, accounting for 89.41%. Overall, the forecasting results by using the OSM road network density are found to have similar characteristics in the forecasting results when using the OSM road network density and population. This shows that the OSM road network density can be used as a spatial metric to evaluate the level of urban economic development in cities.
Considering that the main purpose of this paper is to study whether the OSM road network density can be used as a spatial metric to evaluate the level of urban economic development, the following discussion will focus on forecasting results obtained by using the OSM road network density.
As shown in Figure 7 and Table A2, there are a total of 50 Chinese cities that have negative absolute residuals, and 35 cities with positive absolute residuals. The largest positive absolute residual is Chongqing, with a value of the absolute residual of 0.444; the smallest negative absolute residual is Shenzhen, with a value of the absolute residual of −0.2817. The absolute residuals of 44 cities are between −0.1 and 0.0, accounting for 51.76%, and the absolute residuals of 27 cities are between 0.0 and 0.1, accounting for 31.76%. The absolute residuals of 71 cities are between −0.1 and 0.1, accounting for 85.53%. The absolute residuals of 4 cities are between −0.2 and −0.1, accounting for 4.71%. The absolute residuals of 4 cities are between 0.1 and 0.2, accounting for 4.71%. The absolute residuals of 79 cities are between −0.2 and 0.2, accounting for 92.94%.
As shown in Figure 8 and Table A2, there are a total of 50 Chinese cities that have negative relative residuals, and 35 cities with positive relative residuals. The smallest negative relative residual is Anshan, with a relative residual of −34.6088, and the largest positive relative residual is Shijiazhuang, with a value of the relative residual of 47.8711. The relative residuals of 41 cities are between −10 and 0.0, accounting for 48.24%, and the relative residuals of 20 cities are between 0.0 and 10, accounting for 23.53%. The relative residuals of 61 cities are between −10.0 and 10.0, accounting for 71.76%. The relative residuals of 77 cities are between −20.0 and 20.0, accounting for 90.59%.
Results indicate that the regression models obtained from the above section have a high prediction accuracy, and the OSM road network density can be used as a spatial metric to forecast the municipal GDP. In order to directly discuss the difference between the predicted GDP in 2018 and GDP in 2018, the following discussion is based on the absolute residuals in Figure 7 and Table A2.
The largest positive value of residual is Chongqing, the value of residual is 0.444. Chongqing is located in the southwest of China, and is the only municipality in the southwest of China, its total area is 82,400 square kilometers, and its main urban area is 7220 square kilometers. The municipal GDP and OSM road network density from 2014 to 2018 are shown in Table A1 and Table A2. From 2014 to 2017, the municipal GDP of Chongqing increased year by year, the average annual municipal GDP growth of Chongqing is 172.084 billion CNY. However, the municipal GDP growth in 2018 is 93.824 billion CNY, far lower than the average annual municipal GDP growth in the previous 3 years. The OSM road network density growth in 2018 is 0.38, and the average OSM road network density growth is 0.13, different from the municipal GDP growth change in 2018, the OSM road network density growth in 2018 is much higher than the average OSM road network density growth during the previous 3 years, and the inconsistent trends in municipal GDP growth and OSM road network growth have resulted in excessive forecast residual.
Tianjin has the second largest residual, the value of the residual is 0.3149. Tianjin is a municipality located in the north of China, total area is 11,900 square kilometers, and the main urban area is 1007 square kilometers. From 2014 to 2018, the total municipal GDP of Tianjin increased year by year, but the municipal GDP growth varied greatly. The municipal GDP growth in 2018 is 260.45 billion CNY, compared to the average GDP growth of 94.075 billion CNY. As a result, the municipal GDP growth in 2018 is much lower than the average GDP growth from the previous three years. However, the OSM road network density of Tianjin is increasing quickly, more specifically in 2018, the OSM road network density growth is 0.89, and the average OSM road network density growth is 0.26, the OSM road network density growth in 2018 is much higher than the average OSM road network density growth during the previous 3 years, and these factors finally caused the excessive forecast residuals.
On the contrary, Shenzhen has the smallest negative residual, the value of the residual is −0.2817. Shenzhen is an important special economic development zone in the south of China, its total area is 1997 square kilometers, and the main urban area is 927 square kilometers. Shenzhen’s economy is developed with a steady growth from 1.600182 billion CNY in 2014 to 2.4222 billion CNY in 2018. However, during 2014 to 2017, the annual growth of road density in Shenzhen showed a downward trend, but an annual growth in 2018. Thus, the difference between the variation tendency of the municipal GDP and the variation tendency of the OSM road network density makes the forecast value of the municipal GDP smaller than the real data, which leads to Shenzhen’s residual to have the smallest negative residual.
Overall, between the 85 cities, a total of 8 cities were found with absolute residual larger than 0.1, and 6 cities with absolute residual lower than −0.1, these 14 cities include Beijing, Shanghai, Shenzhen, Tianjin, Chongqing, Hangzhou, Dalian, Ningbo, Ji’nan, Suzhou, Shenyang, Wuxi, Foshan, and Shijiazhuang. These 14 cities all have characteristic of developed cities. In recent years, the pace of urban construction of these developed cities has increased at a rapid pace, having large populations, which can contribute a lot to the municipal GDP growth. From 2014 to 2018, the changes of municipal GDP are not consistent with the OSM road network density. More specifically, the municipal GDP annual growth for the above 14 cities in 2018 is much lower than that in 2017, but at the same time, the OSM road network density is increasing, which caused a big absolute residual value.
However, when considering the other 71 cities, the absolute residuals are found to be between −0.1 and 0.1. From 2014 to 2018, the municipal GDP growth of these cities did not change drastically, the change trend of the municipal GDP and OSM road network density has similarities, resulting in a smaller absolute residual. Among these 71 cities, most cities are second and third grade cities, these cities were found to have smaller populations. Therefore, to some extent, the analysis of the absolute residual in the paper also confirms that the current economic development strategy of China with a high-speed economic stage to a medium-high-speed economic stage has a relatively predicted impact in developed cities, when compared to the second and third grade cities, where the impact is not significant.

5. Conclusions

This study focused on using the urban OSM road network density as a spatial metric to evaluate the urban economic development of 85 Chinese cities. The following conclusions are identified in this paper:
(1) The OSM road network density can be used as a spatial metric to estimate and predict the level of urban economic development of cities which is commonly measured by the municipal GDP.
(2) It is feasible to analyze the level of urban economic development of cities by using the OSM road network density. There is a significant correlation between the OSM road network density and the municipal GDP. Results demonstrated a total of 71 Chinese cities with absolute residuals between −0.1 and 0.1, accounting for 83.53%; and 79 cities with absolute residuals between −0.2 and 0.2, accounting for 92.94%. There are 61 cities with relative residuals between −10 and 10, accounting for 71.76%; and there are 77 cities with relative residuals between −20 and 20, accounting for 90.59%.
(3) In our experiment, the R2 is calculated with the OSM road network density and population is higher than the R2. which is calculated by the OSM road network density, but for the residuals, the absolute and relative residuals by using the OSM road network density are found to have similar characteristics as the absolute and relative residuals by using the OSM road network density and population. Among the 85 cities, there are 73 cities with absolute residuals between −0.1 and 0.1, accounting for 85.88%; and 78 cities with absolute residuals between −0.2 and 0.2, accounting for 91.76%.There are a total of 64 cities with relative residuals between −10 and 10, accounting for 75.29%; and there are 77 cities with relative residuals between −20 and 20, accounting for 90.59%.
(4) The development strategy of the Chinese economy is changing from high-speed development to medium-high speed development, these findings are more apparent in economically developed cities, compared to ordinary cities. Moreover, economically developed cities pay more attention to the influence of high technology.
The research in this paper has identified the imperfections found in the method when using the OSM road network density to evaluate and predict the level of urban economic development in Chinese cities. It is evident that the OSM road network density can be used as a spatial metric for predictive analysis and policy making to gain transparency. In the future, we will use the proposed spatial metric to do some related research.

Author Contributions

Methodology, Bo Liu, Yu Shi, Da-Jun Li and Ming-Hsiang Tsou; Software, Yu Shi; Visualization, Yu Shi and Bo Liu; Writing—original draft, Bo Liu, Yu Shi and Yan-Dong Wang; Writing—review and editing, Yan-Dong Wang, Gabriela Fernandez and Ming-Hsiang Tsou. All authors have read and agreed to the published version of the manuscript.

Funding

This work was funded by the National Key Research and Development Program of China (No.2016YFB0501403), The National Natural Science Foundation of China (Nos. 41201395, 416014160), The Key Laboratory of Earth Observation and Geospatial Information Science of NASG (201811), and the China Scholarship Council Foundation of China (No. 201808360267).

Acknowledgments

Thank you to the Center for Human Dynamics in the Mobile Age, Department of Geography, San Diego State University, San Diego, California, USA.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

Table
Table A1. The statistical results of the municipal GDP, coefficient of determination R2, and regression equation of 85 Chinese cities from 2014 to 2017.
Table A1. The statistical results of the municipal GDP, coefficient of determination R2, and regression equation of 85 Chinese cities from 2014 to 2017.
No.CitiesLocation of the CitiesMunicipal GDP (trillion CNY)OSM RND Represent the OSM Road Network Density Regression EquationCoefficient of Determination (R2)
20142015201620172014201520162017
1BeijingEastern2.133082.296862.489932.800044.9582245.1415465.3578835.499061y = 1.174x − 3.7210.9515
2ShanghaiEastern2.3560942.4964992.7466153.0133863.1552283.5668784.2138394.775253y = 0.4047x + 1.06350.9954
3GuangzhouEastern1.6706871.8100411.9547442.1503152.6769913.5342834.428085.49531y = 0.1697x + 1.2120.999
4ShenzhenEastern1.6001821.7502861.949262.2490066.8279658.9590369.3945079.633126y = 0.1792x + 0.32740.6705
5TianjinEastern1.5726931.6538191.7885391.8549195.9015926.2483476.4411556.67073y = 0.3837x − 0.7060.956
6ChongqingCentral1.426221.5717271.7740591.9424730.6624380.814810.9001021.039407y = 1.4165x + 0.46870.9746
7HangzhouEastern0.9206161.0050211.1313721.2603362.7828853.0225153.9859844.35391y = 0.1934x + 0.39560.9616
8NanjingEastern0.8820750.9720771.0503021.171512.4973842.9306333.5501944.445725y = 0.1445x + 0.53390.9909
9QingdaoEastern0.869210.9300071.0011291.1037281.7600781.8877592.1113932.364463y = 0.3787x + 0.20680.9962
10DalianEastern0.7655580.7731640.681020.736391.750481.8378181.9174771.981108y = −0.2446x + 1.19690.3415
11NingboEastern0.7610280.8003610.8686490.984211.5607261.9690642.0564872.125499y = 0.31x + 0.2560.6433
12XiamenEastern0.3273580.3466030.3784270.435176.8837437.5039137.6902567.821306y = 0.0947x − 0.33580.6945
13Ji’nanEastern0.577060.6100230.6536120.7201961.302191.3537391.4277731.692866y = 0.3468x + 0.13940.9479
14SuzhouEastern1.3760891.4504071.547511.7319513.4400613.8594974.5613666.548412y = 0.1107x + 1.01680.9822
15WuhanCentral1.0069481.090561.1912611.3410342.7504182.9513963.206143.695856y = 0.3509x + 0.05170.9939
16ChengduWestern1.0056831.0801161.2170231.3889393.3587564.2364984.5941475.561516y = 0.1801x + 0.37380.9487
17ChangshaCentral0.7824810.8510130.9455361.0210131.5335761.7563672.0521843.297249y = 0.1215x + 0.63750.8346
18Xi’anWestern0.5492640.580120.6257180.7469853.1963163.8654944.1359664.471107y = 0.1421x + 0.06880.7828
19ShenyangEastern0.7098710.7280.5460010.5864972.0141672.1504152.2030392.292415y = −0.5466x + 1.82590.4997
20ZhengzhouEastern0.6676990.7311520.8025310.913021.7404562.1151132.6767312.880727y = 0.1937x + 0.32270.9216
21DongguanEastern0.5881180.6275060.6827670.7582091.0305981.5013662.1975822.854381y = 0.092x + 0.48980.992
22FuzhouEastern0.5169160.5618080.6197640.710340.9022372.0820212.2571732.395907y = 0.0982x + 0.41470.6464
23WuxiEastern0.8205310.8518260.9210021.051182.1531783.1628213.3012543.867276y = 0.1253x + 0.51990.7635
24HarbinEastern0.5340070.5751210.6101610.6355050.4831180.519340.5847940.615797y = 0.7215x + 0.19130.9782
25FoshanEastern0.7603280.8003920.8630.954964.6588895.2971525.5170815.916113y = 0.1522x + 0.03090.8897
26ChangchunEastern0.5342430.5530030.5917940.6530030.9375741.2290541.4321111.486014y = 0.1836x + 0.34970.7557
27ShijiazhuangEastern0.5170270.544060.5927730.6460882.6509262.7560922.8288022.846626y = 0.5834x − 1.04130.8326
28TaiyuanCentral0.2531090.2735340.295560.3382180.9259420.969811.2308861.398845y = 0.1582x + 0.11110.9397
29YantaiEastern0.6002080.6446080.692570.7338950.5717820.6246880.6922350.923017y = 0.3473x + 0.42370.8592
30HefeiCentral0.5157970.5660270.6274380.7213451.078471.9162912.424562.662393y = 0.1166x + 0.3720.8532
31KunmingWestern0.3712990.3968010.4300080.4857640.7640010.9181561.014961.177625y = 0.2808x + 0.14890.9709
32WenzhouEastern0.4302810.4619840.504540.5453170.9879311.4470271.5845842.000244y = 0.1169x + 0.30960.9462
33NanningWestern0.314830.3410090.3703390.4118830.3049560.4052850.4060030.589919y = 0.3344x + 0.21690.9107
34NanchangCentral0.3667960.4000010.4354990.5003194.4456485.0801185.2074995.528507y = 0.1164x − 0.16420.8569
35TangshanEastern0.622530.610.630620.710610.6928590.8589970.9055331.472886y = 0.1277x + 0.5180.9037
36ZiboEastern0.4029770.413020.4412010.4781320.9535030.9889581.81032.013233y = 0.058x + 0.35030.8951
37ChangzhouEastern0.4901870.527320.5773860.6622281.3409961.8016081.8485983.019328y = 0.1003x + 0.36340.9294
38QuanzhouEastern0.5733360.6137740.6646630.7548010.6344380.9203261.3848511.342822y = 0.1867x + 0.45170.7318
39GuiyangEastern0.2497270.2891160.315770.3537961.40451.5610811.730262.215922y = 0.1197x + 0.09520.9203
40JiaxingEastern0.335280.3517060.376010.4355241.1272911.2660251.5987741.725241y = 0.1447x + 0.16780.8476
41NantongEastern0.5652790.614840.676820.773460.5872341.2433942.1896512.424369y = 0.0989x + 0.49820.8832
42JinhuaEastern0.3206640.340650.3635010.3870220.3144450.6634070.7058740.910499y = 0.1096x + 0.28190.8947
43ZhuhaiEastern0.1857320.2024980.2226370.2564732.2573832.8928063.1783393.573405y = 0.0525x + 0.06050.9154
44HuizhouEastern0.300070.3140030.3412170.3830580.545940.6818920.7330180.937024y = 0.2204x + 0.17490.9566
45XuzhouEastern0.4963910.5319880.5808520.6605950.8396670.8335491.6041023.131913y = 0.0638x + 0.46530.9418
46HaikouEastern0.109170.1161960.1257670.1390581.2155751.390471.3915861.593147y = 0.0796x + 0.01130.9019
47UrumqiWestern0.2461470.2631640.2458980.2743821.4715621.4886151.5485171.676137y = 0.1043x + 0.09620.4828
48ShaoxingEastern0.4265830.4466650.4710.5108040.7581670.8031631.0002261.245672y = 0.1613x + 0.31030.9789
49ZhongshanEastern0.282330.3010030.3202780.3450311.6368481.7598372.1447332.393204y = 0.0758x + 0.16170.9693
50Taizhou (Zhejiang)Eastern0.3387510.3558130.3842810.4388220.5311150.5762760.8323871.119472y = 0.1603x + 0.25680.9833
51LanzhouWestern0.2000940.2095990.2264230.2523540.6238520.6449020.7230111.678987y = 0.0412x + 0.18430.8392
52WeifangEastern0.4786740.5170530.552270.5858630.8082871.0670391.2514962.839208y = 0.0438x + 0.46810.7604
53BaodingEastern0.303520.3000340.322730.358091.084391.3070911.3213591.482171y = 0.1299x + 0.15240.637
54ZhenjiangEastern0.3252380.3502480.3833840.4010360.8777171.2111881.3858881.712882y = 0.0949x + 0.24190.955
55YangzhouEastern0.3697890.4016840.4449380.5064921.4402311.8011232.2981182.526604y = 0.117x + 0.19470.9365
56HohhotWestern0.2894050.3090520.3173590.2743720.5673680.6558920.8598380.869946y = −0.0029x + 0.29970.0005
57LangfangEastern0.2175960.240190.270630.288060.4376210.7267910.7912531.032201y = 0.1226x + 0.16260.9129
58LuoyangCentral0.3284570.3508750.378290.4290190.9083341.1103871.9129282.084188y = 0.069x + 0.26790.856
59WeihaiEastern0.2790340.3001570.321220.348011.1581461.2496031.2970571.523613y = 0.1835x + 0.07230.935
60YanchengEastern0.3835620.421250.4576080.5082690.6573490.6929850.8941441.233145y = 0.194x + 0.2740.9275
61LinyiEastern0.356980.376320.4026750.4345391.4807542.0841112.3178472.54728y = 0.0686x + 0.24810.8741
62JiangmenEastern0.2082760.2240020.2418780.2690251.34651.5588391.6600161.74101y = 0.1434x + 0.00970.8806
63Taizhou (Jiangsu)Eastern0.3370890.3655530.4101780.4744530.8224941.2579071.5964252.136174y = 0.107x + 0.24140.9821
64ZhangzhouEastern0.2506360.2767450.3125340.3528530.9169131.1517411.2000811.38604y = 0.2206x + 0.04140.9201
65HandanEastern0.3080010.314540.333710.366630.6752820.8130870.9145550.949443y = 0.1836x + 0.17690.7366
66JiningWestern0.3800060.4013120.4301820.4650570.4719340.5268960.7760550.91366y = 0.1739x + 0.30220.966
67WuhuEastern0.230790.2457320.2699440.3065520.6727680.9652521.2276871.957186y = 0.0598x + 0.19120.9873
68YinchuanCentral0.1395670.1480730.1617280.1803170.4976750.596030.6933120.716513y = 0.164x + 0.05480.8525
69LiuzhouEastern0.2208510.2298620.2476940.2755640.4552060.4856490.6045880.599445y = 0.2719x + 0.09770.7542
70MianyangWestern0.1579890.1700330.1830420.2074750.4459440.5540540.7073220.734055y = 0.1434x + 0.09210.838
71ZhanjiangEastern0.225870.2380020.2584780.2824031.4057691.6596092.2614922.575143y = 0.0455x + 0.16130.9729
72AnshanEastern0.23490.23260.144080.160210.7272660.7680870.8165420.828872y = −0.9284x + 0.92190.8297
73DaqingEastern0.4070.298350.2610.268050.4999831.1911771.22281.222723y = −0.1858x + 0.50070.9601
74YichangCentral0.3132210.338480.3709360.3857171.9967362.1268152.1727782.900025y = 0.0641x + 0.20480.6429
75BaotouEastern0.3636310.3781930.3867630.2753030.6011520.6289970.8516390.900335y = −0.1854x + 0.48920.3029
76JilinEastern0.273020.245520.2531350.230280.6017620.6958630.7813480.880384y = −0.1324x + 0.34840.7854
77Huai’anEastern0.2455390.2745090.30480.3387430.5495580.6569790.6957440.776926y = 0.4165x + 0.01190.9658
78CangzhouEastern0.3133380.324060.353340.381690.7462550.8099760.8884091.36473y = 0.1017x + 0.24620.8629
79XiangyangCentral0.312930.3382120.3694510.406490.3479270.4247440.5636340.594033y = 0.3347x + 0.19520.9255
80YueyangCentral0.2669390.2886280.3100870.3258030.787180.8292110.8702211.291411y = 0.0898x + 0.21310.67
81TaianEastern0.3002190.315840.331680.3585280.9356671.0417361.7692951.918874y = 0.0462x + 0.26120.8588
82DongyingEastern0.3430490.3450640.347960.3801780.4558150.4969660.7372270.774951y = 0.0783x + 0.30580.5307
83NanyangCentral0.2676880.2875020.3118770.337770.325720.3870650.397980.513531y = 0.3688x + 0.15150.9074
84XiningWestern0.1065780.1131620.1248170.128492.1488552.3364192.5530522.433314y = 0.0526x − 0.00620.7804
85LhasaWestern0.0347450.0389460.0424950.0479160.4972340.7883840.8576790.901169y = 0.0273x + 0.02030.7902
Table
Table A2. The statistical results for predictive GDP of 85 Chinese cities in 2018.
Table A2. The statistical results for predictive GDP of 85 Chinese cities in 2018.
No.CitiesLocation of the CitiesGDP in 2018 (trillion CNY)Predictive GDP in 2018 by Using OSM Road Network Density (trillion CNY)Predictive GDP in 2018 by Using OSM Road Network Density and Population (trillion CNY)Absolute Residuals by Using OSM Road Network DensityRelative Residuals by Using OSM Road Network DensityAbsolute Residuals by Using OSM Road Network Density and PopulationRelative Residuals by Using OSM Road Network Density and Population
1BeijingEastern3.03203.19383.44220.16185.33640.410213.5290
2ShanghaiEastern3.26803.06373.0811−0.2043−6.2515−0.1869−5.7191
3GuangzhouEastern2.28592.23862.2264−0.0473−2.0692−0.0595−2.6029
4ShenzhenEastern2.42222.14052.4473−0.2817−11.62990.02511.0362
5TianjinEastern1.88102.19582.21280.314916.73580.331817.6396
6ChongqingCentral2.03632.48032.06820.444021.80430.03191.5666
7HangzhouEastern1.35091.54461.50630.193714.33860.155411.5034
8NanjingEastern1.28201.20241.0471−0.0797−6.2090−0.2349−18.3229
9QingdaoEastern1.20021.23301.25370.03282.73290.05354.4576
10DalianEastern0.76690.59970.7002−0.1672−21.8021−0.0667−8.6974
11NingboEastern1.07460.95311.1785−0.1215−11.30650.10399.6687
12XiamenEastern0.47910.44740.4887−0.0317−6.61660.00962.0038
13Ji’nanEastern0.78571.02870.88260.243030.92780.096912.3330
14SuzhouEastern1.85651.95901.92160.10255.52110.06513.5066
15WuhanCentral1.48471.42281.4250−0.0619−4.1692−0.0597−4.0210
16ChengduWestern1.52541.59291.53920.06754.42510.01380.9047
17ChangshaCentral1.10031.06821.2354−0.0322−2.91740.135112.2785
18Xi’anWestern0.83500.75051.2377−0.0845−10.11980.402748.2275
19ShenyangEastern0.62920.48712.5457−0.1421−22.58421.9165304.5931
20ZhengzhouEastern1.01431.02721.01070.01281.2718−0.0036−0.3549
21DongguanEastern0.82790.82230.8294−0.0056−0.67640.00150.1812
22FuzhouEastern0.78570.69610.7764−0.0895−11.4038−0.0093−1.1837
23WuxiEastern1.14391.02921.1492−0.1146−10.02710.00530.4633
24HarbinEastern0.63010.67590.67570.04587.26870.04567.2369
25FoshanEastern0.99361.16021.13900.166616.76730.145414.6337
26ChangchunEastern0.71760.65330.6331−0.0643−8.9604−0.0845−11.7754
27ShijiazhuangEastern0.60830.89950.59670.291347.8711−0.0116−1.9070
28TaiyuanCentral0.38840.35840.3837−0.0301−7.7240−0.0047−1.2101
29YantaiEastern0.78330.76580.7777−0.0174−2.2341−0.0056−0.7149
30HefeiCentral0.78230.74520.8277−0.0371−4.74240.04545.8034
31KunmingWestern0.52070.51860.4968−0.0021−0.4033−0.0239−4.5900
32WenzhouEastern0.60060.55190.5490−0.0487−8.1086−0.0516−8.5914
33NanningWestern0.41470.45480.45320.04019.66960.03859.2838
34NanchangCentral0.52750.51440.5443−0.0131−2.48340.01683.1848
35TangshanEastern0.69550.74230.74440.04686.72900.04897.0309
36ZiboEastern0.50680.48390.4693−0.0230−4.5185−0.0375−7.3994
37ChangzhouEastern0.70500.67800.7372−0.0270−3.82980.03224.5674
38QuanzhouEastern0.84680.76870.7780−0.0781−9.2230−0.0688−8.1247
39GuiyangEastern0.37980.43470.36710.054914.4550−0.0127−3.3439
40JiaxingEastern0.48720.45130.4875−0.0359−7.36860.00030.0616
41NantongEastern0.84270.78870.9330−0.0540−6.40800.090310.7156
42JinhuaEastern0.41000.39520.4049−0.0148−3.6098−0.0051−1.2439
43ZhuhaiEastern0.29150.26260.3010−0.0288−9.91420.00953.2590
44HuizhouEastern0.41030.39100.3898−0.0193−4.7039−0.0205−4.9963
45XuzhouEastern0.67550.69190.70170.01642.42780.02623.8786
46HaikouEastern0.15110.20830.14590.057337.8557−0.0052−3.4414
47UrumqiWestern0.31000.28920.3181−0.0207−6.70970.00812.6129
48ShaoxingEastern0.54170.51720.5427−0.0244−4.52280.0010.1846
49ZhongshanEastern0.36330.37380.41420.01052.89020.050914.0105
50Taizhou (Zhejiang)Eastern0.48750.49790.49140.01052.13330.00390.8000
51LanzhouWestern0.27330.25720.2673−0.0161−5.8910−0.006−2.1954
52WeifangEastern0.61570.62890.60770.01322.1439−0.008−1.2993
53BaodingEastern0.35900.42590.42530.066918.63510.066318.4680
54ZhenjiangEastern0.40500.40950.50130.00451.11110.096323.7778
55YangzhouEastern0.54660.50260.5821−0.0440−8.04980.03556.4947
56HohhotWestern0.29040.29690.29570.00662.23830.00531.8251
57LangfangEastern0.31080.31120.31740.00040.12870.00662.1236
58LuoyangCentral0.46410.43070.4405−0.0334−7.1967−0.0236−5.0851
59WeihaiEastern0.36410.35680.3589−0.0073−2.0049−0.0052−1.4282
60YanchengEastern0.54870.55440.32280.00571.0388−0.2259−41.1700
61LinyiEastern0.47180.45710.4416−0.0147−3.1157−0.0302−6.4010
62JiangmenEastern0.29000.29100.30010.00100.34480.01013.4828
63Taizhou (Jiangsu)Eastern0.51080.47900.2763−0.0317−6.2255−0.2345−45.9084
64ZhangzhouEastern0.39480.47220.40060.077419.60490.00581.4691
65HandanEastern0.34550.35770.36000.01223.53110.01454.1968
66JiningWestern0.49310.49050.4776−0.0026−0.5273−0.0155−3.1434
67WuhuEastern0.32790.32600.3294−0.0019−0.57940.00150.4575
68YinchuanCentral0.19010.22110.17200.030916.3072−0.0181−9.5213
69LiuzhouEastern0.30840.30250.2880−0.0059−1.9131−0.0204−6.6148
70MianyangWestern0.23040.23760.16450.00723.1250−0.0659−28.6024
71ZhanjiangEastern0.30080.29470.2966−0.0061−2.0279−0.0042−1.3963
72AnshanEastern0.17510.11450.1012−0.0606−34.6088−0.0739−42.2045
73DaqingEastern0.28010.26200.2582−0.0181−6.4620−0.0219−7.8186
74YichangCentral0.40640.47840.39040.072017.7165−0.016−3.9370
75BaotouEastern0.29520.22980.3095−0.0654−22.15450.01434.8442
76JilinEastern0.22100.21800.2167−0.0030−1.3575−0.0043−1.9457
77Huai’anEastern0.36010.42670.34990.066618.4949−0.0102−2.8325
78CangzhouEastern0.36760.42660.40650.059016.05010.038910.5822
79XiangyangCentral0.43100.40960.4381−0.0214−4.96520.00711.6473
80YueyangCentral0.34110.40070.30010.059617.4729−0.041−12.0199
81TaianEastern0.36520.35790.3465−0.0072−1.9989−0.0187−5.1205
82DongyingEastern0.41520.41050.3266−0.0047−1.1320−0.0886−21.3391
83NanyangCentral0.35670.35010.3379−0.0066−1.8503−0.0188−5.2705
84XiningWestern0.12860.13220.13600.00352.79940.00745.7543
85LhasaWestern0.05280.04860.0515−0.0042−7.9545−0.0013−2.4621

References

  1. Gustafson, E.J. Quantifying landscape spatial pattern: What is the state of the art? Ecosystems 1998, 1, 143–156. [Google Scholar] [CrossRef]
  2. Hargis, C.D.; Bissonette, J.A.; David, J.L. The behavior of landscape metrics commonly used in the study of habitat fragmentation. Landsc. Ecol. 1998, 13, 167–186. [Google Scholar] [CrossRef]
  3. O’Neill, R.V.; Krummel, J.R.; Gardner, R.H.; Sugihara, G.; Jackson, B.; DeAngelis, D.L.; Milne, B.T.; Turner, M.G.; Zygmunt, B.; Christensen, S.W.; et al. Indices of landscape pattern. Landsc. Ecol. 1988, 1, 153–162. [Google Scholar] [CrossRef]
  4. McGarigal, K.; Cushman, S.A.; Neel, M.C.; Ene, E. FRAGSTATS: Spatial Pattern Analysis Program for Categorical Maps. 2002. Available online: www.umass.edu/landeco/research/fragstats/fragstats.html (accessed on 15 May 2020).
  5. Herold, M.; Liu, X.H.; Clarke, K.C. Spatial Metrics and Image Texture for Mapping Urban Land use. Photogramm. Eng. Remote Sens. 2003, 69, 991–1001. [Google Scholar] [CrossRef] [Green Version]
  6. Herold, M.; Couclelis, H.; Clarke, K.C. The role of spatial metrics in the analysis and modeling of urban land use change. Comput. Environ. Urban Syst. 2005, 29, 369–399. [Google Scholar] [CrossRef]
  7. Reis, J.P.; Silva, E.A.; Pinbo, P. Measure space: A review of spatial metrics for urban growth and shrinkage. In The Routledge Handbook of Planning Research Methods; Silva, E.A., Healey, P., Harris, N., Van den Broeck, P., Eds.; Routledge: New York, NY, USA, 2014; pp. 279–292. [Google Scholar]
  8. Reis, J.P.; Silva, E.A.; Pinbo, P. Spatial metrics to study urban patterns in growing and shrinking cities. Urban Geogr. 2016, 37, 246–271. [Google Scholar] [CrossRef]
  9. Chen, M.; Huang, Y.; Tang, Z.; Lu, D.; Liu, H.; Ma, L. The provincial pattern of the relationship between urbanization and economic development in China. J. Geogr. Sci. 2014, 24, 33–45. [Google Scholar] [CrossRef]
  10. Henderson, V. The Urbanization Process and Economic Growth: The So-What Question. J. Econ. Growth 2003, 8, 47–71. [Google Scholar] [CrossRef]
  11. Njoh, A.J. Urbanization and development in sub-Saharan Africa. Cities 2003, 20, 167–174. [Google Scholar] [CrossRef]
  12. Liu, Y.S.; Fang, F.; Li, Y.H. Key issues of land use in China and implications for policy making. Land Use Policy 2014, 40, 6–12. [Google Scholar] [CrossRef]
  13. Xiang, W.N.; Stuber, R.M.B.; Meng, X.C. Meeting critical challengers and striving for urban sustainability in China. Landsc. Urban Plan. 2011, 100, 418–420. [Google Scholar] [CrossRef] [PubMed]
  14. Li, Y.Z. Urbanization and economic growth in China: An empirical research based on VAR model. Int. J. Econ. Financ. 2017, 9, 210–219. [Google Scholar]
  15. Cai, Z.Y.; Liu, Q.; Cao, S.X. Real estate supports rapid development of China’s urbanization. Land Use Policy 2020, 95, 104582. [Google Scholar] [CrossRef]
  16. National Bureau of Statistics (NBS). China Statistical Yearbook; China statistics Press: Beijing, China, 2018.
  17. Heshmati, A.; Rashidghalam, M. Measurement and Analysis of Urban Infrastructure and Its Effects on Urbanization in China. J. Infrastruct. Syst. 2020, 26, 04019030. [Google Scholar] [CrossRef]
  18. Yu, N.N.; de Roo, G.; de Jong, M.; Storm, S. Does the expansion of a motorway network lead to economic agglomeration Evidence from China. Transp. Policy 2016, 45, 218–227. [Google Scholar] [CrossRef]
  19. Jiao, J.; Wang, J.; Jin, F.; Du, C. Understanding Relationship between Accessibility and Economic Growth: A Case Study from China (1990–2010). Chin. Geogr. Sci. 2016, 26, 803–816. [Google Scholar] [CrossRef] [Green Version]
  20. Worku, I. Road Sector Development and Economic Growth in Ethiopia. Ethiop. J. Econ. 2010, 19, 101–146. [Google Scholar]
  21. Ivanova, E.; Masarova, J. Importance of road infrastructure in the economic development and competitiveness. Compet. Nations Global Econ. 2013, 18, 263–274. [Google Scholar] [CrossRef]
  22. Beyzatlar, M.A.; Karacal, M.; Yetkiner, H. Granger-causality between transportation and GDP: A panel data approach. Transp. Res. Part A Policy Pract. 2014, 63, 43–55. [Google Scholar] [CrossRef] [Green Version]
  23. Gao, Y.; Zhang, Y.P.; Li, H.J.; Peng, T.; Hao, S.Q. Study on the Relationship Between Comprehensive Transportation Freight Index and GDP in China. Procedia Eng. 2016, 137, 571–580. [Google Scholar] [CrossRef]
  24. Fan, S.G.; Chan-Kang, C. Regional road development, rural and urban poverty: Evidence from China. Transp. Policy 2008, 15, 305–314. [Google Scholar] [CrossRef]
  25. Haklay, M.; Weber, P. OpenStreetMap: User-Generated Street Maps. IEEE Pervasive Comput. 2008, 7, 12–18. [Google Scholar] [CrossRef] [Green Version]
  26. OpenStreetMap. Stats-OpenStreetMap Wiki [Online]. Available online: https://wiki.openstreetmap/wiki/.orgStats (accessed on 20 June 2020).
  27. Over, M.; Schilling, A.; Neubauer, S.; Zipf, A. Generating web-based 3D City Models from OpenStreetMap: The current situation in Germany. Comput. Environ. Urban Syst. 2010, 34, 496–507. [Google Scholar] [CrossRef]
  28. Fonte, C.C.; Martinho, N. Assessing the applicability of OpenStreetMap data to assist the validation of land use/land cover maps. Int. J. Geogr. Inf. Sci. 2017, 31, 2382–2440. [Google Scholar] [CrossRef]
  29. Fonte, C.; Minghini, M.; Patriarca, J.; Antoniou, V.; See, L.; Skopeliti, A. Generating Up-to-Date and Detailed Land Use and Land Cover Maps Using OpenStreetMap and GlobeLand30. ISPRS Int. J. Geo-Inf. 2017, 6, 125. [Google Scholar] [CrossRef]
  30. Bittner, C. Diversity in volunteered geographic information: Comparing OpenStreetMap and Wikimapia in Jerusalem. GeoJournal 2016, 82, 887–906. [Google Scholar] [CrossRef]
  31. Mobasheri, A. A Rule-Based Spatial Reasoning Approach for OpenStreetMap Data Quality Enrichment; Case Study of Routing and Navigation. Sensors 2017, 17, 2498. [Google Scholar] [CrossRef] [Green Version]
  32. Zhou, Q. Exploring the relationship between density and completeness of urban building data in OpenStreetMap for quality estimation. Int. J. Geogr. Inf. Sci. 2018, 32, 257–281. [Google Scholar] [CrossRef]
  33. Goetz, M.; Zipf, A. Towards defining a framework for the automatic derivation of 3D CityGML models from volunteered geographic information. Int. J. 3-D Inf. Model. 2012, 1, 1–16. [Google Scholar] [CrossRef]
  34. Hennig, S. OpenStreetMap used in protected area management. The example of the recreational infrastructure in Berchtesgaden National Park. J. Prot. Mt. Areas Res. Manag. 2017, 1, 30–41. [Google Scholar] [CrossRef] [Green Version]
  35. Mobasheri, A.; Sun, Y.; Loos, L.; Ali, A. Are Crowdsourced Datasets Suitable for Specialized Routing Services? Case Study of OpenStreetMap for Routing of People with Limited Mobility. Sustainability 2017, 9, 997. [Google Scholar] [CrossRef] [Green Version]
  36. Juhász, L.; Hochmair, H.H. How do volunteer mappers use crowdsourced Mapillary street level images to enrich OpenStreetMap? In Proceedings of the 20th AGILE Conference on Geo-Information Science, Wageningen, The Netherlands, 11 May 2017; pp. 18–21. [Google Scholar]
  37. Zhang, Y.J.; Li, X.M.; Wang, A.M.; Bao, T.L.; Tian, S.Z. Density and diversity of OpenStreetMap road networks in China. J. Urban Manag. 2015, 4, 135–146. [Google Scholar] [CrossRef] [Green Version]
  38. Goetz, M. Towards generating highly detailed 3D CityGML models from OpenStreetMap. Int. J. Geogr. Inf. Sci. 2013, 27, 845–865. [Google Scholar] [CrossRef]
  39. Wang, Z.Y.; Zipf, A. Using Openstreetmap Data to Generate Building Models with Their Inner Structures for 3D Maps. ISPRS Ann. Photogramm. Remote Sens. Spat. Inf. Sci. 2017. [Google Scholar] [CrossRef] [Green Version]
  40. Bergman, C.; Oksanen, J. Conflation of OpenStreetMap and Mobile Sports Tracking Data for Automatic Bicycle Routing. Trans. GIS 2016, 20, 848–868. [Google Scholar] [CrossRef] [Green Version]
  41. Rosina, K.; Hurbã¡Nek, P.; Cebecauer, M. Using OpenStreetMap to improve population grids in Europe. Am. Cartogr. 2016, 44, 139–151. [Google Scholar] [CrossRef]
  42. Zhao, P.X.; Jia, T.; Qin, K.; Shan, J.; Jiao, C.J. Statistical analysis on the evolution of OpenStreetMap road networks in Beijing. Physica A 2015, 420, 59–72. [Google Scholar] [CrossRef]
  43. Dingil, A.E.; Schweizer, J.; Rupi, F.; Stasiskiene, Z. Updated Models of Passenger Transport Related Energy Consumption of Urban Areas. Sustainability 2019, 11, 4060. [Google Scholar] [CrossRef] [Green Version]
  44. Shang, Y.S.; Liu, S.G.; Liu, C.Y.; Yin, P. Spatial-temporal characteristics of urbanization efficiency in coastal cities of China. In Proceedings of the 7th Annual International Conference on Geo-Spatial Knowledge and Intelligence, Guangzhou, China, 20–21 December 2019. [Google Scholar] [CrossRef]
  45. Zou, Y.F.; Deng, M.; Li, Y.J.; Rong, Y. Evolution characteristics and policy implications of new urbanization in provincial capital cities in western China. PLoS ONE 2020, 15, e0233555. [Google Scholar] [CrossRef]
  46. Haklay, M. How good is volunteered geographical information? A comparative study of OpenStreetMap and Ordnance Survey datasets. Environ. Plan. B Plan. Des. 2010, 37, 682–703. [Google Scholar] [CrossRef] [Green Version]
  47. Zielstra, D.; Zipf, A. A comparative study of proprietary geodata and volunteered geographic information for Germany. In Proceedings of the 13th AGILE International Conference on Geographic Information Science, Guimarães, Portugal, 10–14 May 2010. [Google Scholar]
  48. Neis, P.; Zielstra, D.; Zipf, A. The street network evolution of crowdsourced maps: OpenStreetMap in Germany 2007–2011. Future Internet 2012, 4, 1–21. [Google Scholar] [CrossRef] [Green Version]
  49. Girres, J.-F.; Touya, G. Quality assessment of the French OpenStreetMap dataset. Trans. GIS 2010, 14, 435–459. [Google Scholar] [CrossRef]
  50. Luo, L.C.; Liu, B.; Liu, X.C. Data Quality Assessment and Application Analysis for OpenStreetMap Road Network. Jiangxi Sci. 2017, 35, 151–157. [Google Scholar]
  51. Hecht, R.; Kunze, C.; Hahmann, S. Measuring Completeness of Building Footprints in OpenStreetMap over Space and Time. ISPRS Int. J. Geo-Inf. 2013, 2, 1066–1091. [Google Scholar] [CrossRef]
  52. Singh Sehra, A.; Singh, J.; Singh Rai, H. Assessment of OpenStreetMap Data—A Review. Int. J. Comput. Appl. 2013, 76, 17–20. [Google Scholar]
  53. Ludwig, I.; Voss, A.; Krause-Traudes, M. A comparison of the street networks of Navteq and OSM in Germany. In Advancing Geoinformation Science for a Changing World; Geertman, S., Reinhardt, W., Toppen, F., Eds.; Springer: Berlin, Germany, 2011; pp. 65–84. [Google Scholar]
  54. Zhang, J.; Chen, J. Introduction to China’s new normal economy. J. Chin. Econ. Bus. Stud. 2017, 15, 1–4. [Google Scholar] [CrossRef] [Green Version]
  55. Li, C.; Zhang, X.J. Renminbi Internationalization in the New Normal: Progress, Determinants and Policy Discussions. China World Econ. 2017, 25, 22–44. [Google Scholar] [CrossRef]
  56. Montgomery, D.C.; Peck, E.A. Introduction to Linear Regression Analysis; Wiley: Hoboken, NJ, USA, 1982. [Google Scholar]
  57. Hawbaker, T.J.; Radeloff, V.C.; Hammer, R.B.; Clayton, M.K. Road density and land scape pattern in relation to housing density, and ownership, land cover, and soils. Landsc. Ecol. 2005, 20, 609–625. [Google Scholar] [CrossRef]
  58. Shen, J.; Wu, R. Urban Road and Transportation; Wuhan University Press: Wuhan, China, 2006. (In Chinese) [Google Scholar]
  59. Zhang, Q.; Wang, J.; Peng, X.; Gong, P.; Shi, P. Urban built-up land change detection with road density and spectral information from multi-temporal Landsat TM data. Int. J. Remote Sens. 2002, 23, 3057–3078. [Google Scholar] [CrossRef]
  60. Feng, Z.; Liu, D.; Yang, Y. Evaluation of transportation ability of China: From county to province level. Geogr. Res. 2009, 28, 419–429. [Google Scholar]
  61. Jin, F.J.; Wang, C.J.; Li, X.W. Discrimination method and its application analysis of regional transport superiority. Acta Geogr. Sin. 2008, 63, 787–798. [Google Scholar]
  62. Savaş, B. The Relationship between Population and Economic Growth: Empirical Evidence from the Central Asian Economies. Orta Asya Kafkasya Araştırmaları 2008, 6, 135–153. [Google Scholar]
Ijgi 09 00517 g001 550
Figure 1. Study research area and gross domestic product (GDP) in 2018.
Figure 1. Study research area and gross domestic product (GDP) in 2018.
Ijgi 09 00517 g001
Ijgi 09 00517 g002 550
Figure 2. The OpenStreetMap (OSM) road network with three different grades in Chinese cities.
Figure 2. The OpenStreetMap (OSM) road network with three different grades in Chinese cities.
Ijgi 09 00517 g002
Ijgi 09 00517 g003 550
Figure 3. Statistics of R2 calculated by the OSM road network density.
Figure 3. Statistics of R2 calculated by the OSM road network density.
Ijgi 09 00517 g003
Ijgi 09 00517 g004 550
Figure 4. Distribution of R2 calculated by the OSM road network density.
Figure 4. Distribution of R2 calculated by the OSM road network density.
Ijgi 09 00517 g004
Ijgi 09 00517 g005 550
Figure 5. Statistics of R2 calculated by the OSM road network density and population.
Figure 5. Statistics of R2 calculated by the OSM road network density and population.
Ijgi 09 00517 g005
Ijgi 09 00517 g006 550
Figure 6. Distribution of R2 which is calculated by the OSM road network density and population.
Figure 6. Distribution of R2 which is calculated by the OSM road network density and population.
Ijgi 09 00517 g006
Ijgi 09 00517 g007 550
Figure 7. The statistical result of the absolute residuals by using the OSM road network density.
Figure 7. The statistical result of the absolute residuals by using the OSM road network density.
Ijgi 09 00517 g007
Ijgi 09 00517 g008 550
Figure 8. The statistical result of the relative residuals by using the OSM road network density.
Figure 8. The statistical result of the relative residuals by using the OSM road network density.
Ijgi 09 00517 g008
Ijgi 09 00517 g009 550
Figure 9. Distribution of the absolute residuals of 85 Chinese cities by using OSM road network density.
Figure 9. Distribution of the absolute residuals of 85 Chinese cities by using OSM road network density.
Ijgi 09 00517 g009
Ijgi 09 00517 g010 550
Figure 10. Distribution of the relative residuals of 85 Chinese cities by using OSM road network density.
Figure 10. Distribution of the relative residuals of 85 Chinese cities by using OSM road network density.
Ijgi 09 00517 g010
Ijgi 09 00517 g011 550
Figure 11. The statistical result of the absolute residuals by using OSM road network density and population.
Figure 11. The statistical result of the absolute residuals by using OSM road network density and population.
Ijgi 09 00517 g011
Ijgi 09 00517 g012 550
Figure 12. The statistical result of the relative residuals by using OSM road network density and population.
Figure 12. The statistical result of the relative residuals by using OSM road network density and population.
Ijgi 09 00517 g012
Ijgi 09 00517 g013 550
Figure 13. Distribution of 85 Chinese cities’ absolute residuals using OSM road network density and population.
Figure 13. Distribution of 85 Chinese cities’ absolute residuals using OSM road network density and population.
Ijgi 09 00517 g013
Ijgi 09 00517 g014 550
Figure 14. Distribution of 85 Chinese cities’ relative residuals using OSM road network density and population.
Figure 14. Distribution of 85 Chinese cities’ relative residuals using OSM road network density and population.
Ijgi 09 00517 g014
Ijgi 09 00517 g015 550
Figure 15. Statistical result of the difference between two residuals.
Figure 15. Statistical result of the difference between two residuals.
Ijgi 09 00517 g015
Ijgi 09 00517 g016 550
Figure 16. Distribution of absolute residual differences.
Figure 16. Distribution of absolute residual differences.
Ijgi 09 00517 g016
Table
Table 1. The Chinese municipal GDP (trillion CNY) statistical results of five cities.
Table 1. The Chinese municipal GDP (trillion CNY) statistical results of five cities.
CityGDP (2014)GDP (2015)GDP (2016)GDP (2017)GDP (2018)
Shenyang0.7098710.7280000.5460010.5864970.62924
Urumqi0.2461470.2631640.2458980.2743820.309962
Dalian0.7655580.7731640.681020.736390.76685
Baotou0.3636310.3781930.3867630.2753030.29518
Hohhot0.2894050.3090520.3173590.2743720.29035
Table
Table 2. The OSM road network density of the five cities.
Table 2. The OSM road network density of the five cities.
CityOSM RND (2014)OSM RND (2015)OSM RND (2016)OSM RND (2017)OSM RND (2018)
Shenyang2.0142.1502.2032.2922.449
Urumqi1.4721.4891.5491.6761.851
Dalian1.7501.8381.9171.9812.441
Baotou0.6010.6290.8520.9001.399
Hohhot0.5670.6560.8600.8700.962
Note: OSM RND represents the OSM road network density.

Share and Cite

MDPI and ACS Style

Liu, B.; Shi, Y.; Li, D.-J.; Wang, Y.-D.; Fernandez, G.; Tsou, M.-H. An Economic Development Evaluation Based on the OpenStreetMap Road Network Density: The Case Study of 85 Cities in China. ISPRS Int. J. Geo-Inf. 2020, 9, 517. https://doi.org/10.3390/ijgi9090517

AMA Style

Liu B, Shi Y, Li D-J, Wang Y-D, Fernandez G, Tsou M-H. An Economic Development Evaluation Based on the OpenStreetMap Road Network Density: The Case Study of 85 Cities in China. ISPRS International Journal of Geo-Information. 2020; 9(9):517. https://doi.org/10.3390/ijgi9090517

Chicago/Turabian Style

Liu, Bo, Yu Shi, Da-Jun Li, Yan-Dong Wang, Gabriela Fernandez, and Ming-Hsiang Tsou. 2020. "An Economic Development Evaluation Based on the OpenStreetMap Road Network Density: The Case Study of 85 Cities in China" ISPRS International Journal of Geo-Information 9, no. 9: 517. https://doi.org/10.3390/ijgi9090517

APA Style

Liu, B., Shi, Y., Li, D.-J., Wang, Y.-D., Fernandez, G., & Tsou, M.-H. (2020). An Economic Development Evaluation Based on the OpenStreetMap Road Network Density: The Case Study of 85 Cities in China. ISPRS International Journal of Geo-Information, 9(9), 517. https://doi.org/10.3390/ijgi9090517

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop