Comparison of Socioeconomic Factors Between Surrounding and Non-Surrounding Areas of the Qinghai–Tibet Railway Before and After Its Construction

sustainability

Article Comparison of Socioeconomic Factors between Surrounding and Non-Surrounding Areas of the Qinghai–Tibet Railway before and after Its Construction

Shicheng Li 1,2, Zhaofeng Wang 1,*, Yili Zhang 1,2,3,*, Yukun Wang 4 and Fenggui Liu 4 1 Key Laboratory of Land Surface Pattern and Simulation, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences (CAS), Beijing 100101, China; [email protected] 2 University of Chinese Academy of Sciences, Beijing 100049, China 3 CAS Center for Excellence in Tibetan Plateau Earth Sciences, Beijing 100101, China 4 College of Life and Geographical Science, Qinghai Normal University, Xining 810008, China; [email protected] (Y.W.); [email protected] (F.L.) * Correspondences: [email protected] (Z.W.); [email protected] (Y.Z.); Tel.: +86-10-6488-8736 (Z.W.); +86-10-6485-6505 (Y.Z.)

Academic Editor: Marc A. Rosen Received: 21 April 2016; Accepted: 4 August 2016; Published: 11 August 2016 Abstract: As the world’s highest railway, and the longest highland railway, the Qinghai–Tibet Railway (QTR) has been paid considerable attention by researchers. However, most attention has been paid to the ecological and environmental issues affecting it, and sustainable ecological, social, and economic development-related studies of the QTR are rare. In this study, by analyzing the passenger trafﬁc, freight trafﬁc, passenger-kilometers, and freight-kilometers of the QTR for the period 1982–2013 and the transport structure of the Tibetan Plateau (TP) for 1990–2013, the evolutionary process of the transport system in the TP following the construction of the QTR has been revealed. Subsequently, by comparing Gross Domestic Product (GDP), population, industrial structure, and urbanization level at the county and 1 km scales between surrounding and non-surrounding areas of the QTR, the differences in socioeconomic performance before and after its construction were detected. The results show that (1) in the TP, the highway-dominated transport system will break up and an integrated and sustainable transport system will form; (2) at the county scale, the annual growth rates of GDP of counties surrounding the QTR were greater than those of non-surrounding counties for the period 2000–2010. At the 1 km scale, following the opening of the completed line, the GDP of surrounding areas had a greater growth rate than before; (3) analysis at the county and 1 km scales indicated that population was not aggregated into the surrounding areas of the QTR in the period 2000–2010; (4) in terms of industrial structure, the proportion of primary industry decreased continuously, while the proportion of secondary and tertiary industries increased overall in the period 1984–2012. The QTR had no obvious impact on changes in the urbanization level of its surrounding areas.

Keywords: Qinghai–Tibet Railway; comparison study; transport system; GDP; population; industrial structure; urbanization level

1. Introduction The report published by the World Commission on Environment and Development (Brundtland Commission) proposed the concept of sustainable development [1]—a process that includes, amongst other factors, the development of sustainable transport to provide accessible, safe, environmentally friendly and affordable transport systems [2,3]. The Qinghai–Tibet Railway (QTR)—an engineering marvel on the “roof of the world” and a “green” railway in western China—is one such system [4–6].

Sustainability 2016, 8, 776; doi:10.3390/su8080776 www.mdpi.com/journal/sustainability Sustainability 2016, 8, 776 2 of 17

Railway transport can influence regional development by reducing transportation costs and travel time [7,8], improving accessibility [9,10], and increasing the overall competitiveness of the system [11–13], especially for freight transport [9]. Wang and Jin [14] analyzed the evolution of the railway network, the changes in accessibility, and the relationship between the railway network distribution and spatial economic growth in the past one hundred years in China, and found that the construction of railways led to time–space convergence [14], by which places separated by great distances are brought closer together in terms of the time taken to travel between them. Gutiérrez [8] adopted three indicators to analyze the accessibility impacts of a high-speed railway along the Madrid–Barcelona–French border, and found that the effects of the new line on daily accessibility, economic potential, and location indicators were quite different. Moreover, bullet trains can facilitate social cohesion [13] and mitigate the cost of megacity growth [12]. As the world’s highest and longest highland railway [4,15], the QTR was carefully managed to ensure long-term environmental health and sustainability on the plateau [5,6]. More than 30 years have passed since the Xining–Golmud section went into operation, and more than ten years have passed since the entire line was opened to traffic. The railway has received considerable attention from many researchers [15,16], but most of this attention has been paid to the ecological and environmental issues associated with it [17–19], and socioeconomic-related studies are rare. Although some members of the Chinese Academy of Science have examined the QTR’s influence on the economic growth of Tibet in 2004 [20], it is still difficult to find relevant studies, not to mention studies concerning the QTR’s relationship with the sustainable ecological, social, and economic development of the Tibetan Plateau (TP) or western China, apart from a few initial studies on this theme. One such study, by Wang and Fang [21], discussed the influence of the QTR on the economic structure and layout of the surrounding areas. Jin [22] analyzed the impacts of the construction of the QTR on the economy of Tibet. In these studies, only the Xining–Golmud section was included, and theoretical predictions unrelated to economic realities were made. In addition, some of the studies were qualitative rather than quantitative. For instance, taking the QTR as an example, Bai and Li [23] analyzed qualitatively how the railway stimulated the economy of ethnic minority areas. Using questionnaire surveys, Su and Wall [24] evaluated the QTR’s impacts on tourists’ travel decisions and experiences in Tibet. Based on statistical data for the period 1997–2009, Wang and Wu [25] found that the local Gross Domestic Product (GDP) increased under the influence of the QTR. In this study, in order to reveal the potential influence of the QTR on the economic performance of the TP, the development of the transport system of the TP with the construction of the QTR was first studied. Then various socioeconomic factors were selected and their performance compared before and after the construction of the QTR, focusing on their differences between surrounding and non-surrounding areas. The results can be used to guide railway construction planning in the TP so as to promote sustainable socioeconomic development. It is our hope that the QTR will be seen as a good example of a sustainable railway that can be followed by other regions and countries.

2. Brief Introduction to the Qinghai–Tibet Railway As the main region of the Third Pole, the TP has a very harsh natural environment [26,27]. Consequently, compared with the mid-east areas of China, transportation infrastructure here is lacking. Before 1955, freight could only be transported into Tibet by camel. For the period 1955–1978, a highway was available. In 1978, the construction of the QTR began with government support, and it is still considered a landmark project [5,15], despite raising serious concerns about its potential environmental impacts. As a result, the Chinese government has invested a large amount of money to protect the TP’s fragile ecosystem [5]. The QTR connects Xining, the capital of Qinghai Province, to Lhasa, the capital of the Tibet Autonomous Region. The first section of the QTR was opened to traffic in 1984, starting from Xining (XG) and ending at Golmud (GL) (Figure1), with a length of 814 km. The second section was inaugurated and opened to a regular trial service in July 2006, starting from GL and ending at Lhasa Sustainability 2016, 88,, 776 3 of 17 inaugurated and opened to a regular trial service in July 2006, starting from GL and ending at Lhasa (Figure1), with a length of 1142 km. To minimize the negative influence of human activities on the (Figure 1), with a length of 1142 km. To minimize the negative influence of human activities on the environment of this area, 38 of the 45 stations in this section are unstaffed [5]. The length of the entire environment of this area, 38 of the 45 stations in this section are unstaffed [5]. The length of the entire line is 1956 km, 550 km of which were laid on permafrost [28]. A network of tunnels was added to line is 1956 km, 550 km of which were laid on permafrost [28]. A network of tunnels was added to avoid disrupting the seasonal migration routes of animals [5]. avoid disrupting the seasonal migration routes of animals [5].

80° E 85° E 90° E 95° E 100° E

40° N Gansu Province

Xinjiang Uygur Delhi Gangcha !( !( !( Autonomous Region !( !( !( !( Golmud!( Ulan Xining Qinghai

35° N Province

Tibet Autonomous Region Amdo !( !(Nagqu Sichuan Damxung !( Province !(!( 30° N Legend Lhasa !( County/city Xining-Golmud Golmud-Lhasa County boundary 0 250 km Yu nnan Province Provincial boundary

FigureFigure 1 1.. LocationLocation map map of of the the Qinghai–Tibet Qinghai–Tibet Railway.

The important technical details of the QTR are as follows. The XG section is double track and The important technical details of the QTR are as follows. The XG section is double track and the GL section is single track. The design speed for the QTR is 160 km/h, and the commercial speed the GL section is single track. The design speed for the QTR is 160 km/h, and the commercial speed for the XG section is 140 km/h (since 20 March 2015) and 120 km/h for the GL section. Note that the for the XG section is 140 km/h (since 20 March 2015) and 120 km/h for the GL section. Note that the commercial speed for sections laid on permafrost is 100 km/h [29]. Various measures were taken to commercial speed for sections laid on permafrost is 100 km/h [29]. Various measures were taken to ensure the stability of the railway embankment in permafrost regions [29]. In terms of traction, at ensure the stability of the railway embankment in permafrost regions [29]. In terms of traction, at first, first, diesel-powered locomotives were used for the whole line, but electric locomotives have been diesel-powered locomotives were used for the whole line, but electric locomotives have been used for used for the XG section since June 2011. The maximum train lengths for passenger trains and freight the XG section since June 2011. The maximum train lengths for passenger trains and freight trains are trains are 650 m and 850 m, respectively. The passenger carriages used on the QTR line are specially 650 m and 850 m, respectively. The passenger carriages used on the QTR line are specially built and built and fully enclosed and have an oxygen supply for each passenger [30]. Signs in the carriages are fully enclosed and have an oxygen supply for each passenger [30]. Signs in the carriages are in Tibetan, in Tibetan, Chinese, and English. The line has a capacity of eight pairs of passenger trains. Since Chinese, and English. The line has a capacity of eight pairs of passenger trains. Since October 2006, October 2006, five pairs of passenger trains run on the GL section and one further pair on the XG five pairs of passenger trains run on the GL section and one further pair on the XG section [31]. section [31]. 3. Methods and Materials 3. Methods and Materials 3.1. Methodology 3.1. Methodology Figure2 presents an overview of the methodology employed in this study. First, in order to understandFigure 2 the presents development an overview of the transportof the methodolog system iny the employed TP with thein this construction study. First, of the in QTR,order the to understandchanges in passenger the development traffic, freightof the transport traffic, passenger-kilometers, system in the TP with and the freight-kilometers, construction of the of QTR, Qinghai the changesand Tibet in since passenger the 1980s traffic, were freight analyzed. traffic, Then, pass toenger-kilometers, reveal the position and of freight-kilometers, railway transport in of transport Qinghai andsystem Tibet of thesince TP, the transport-modes-based 1980s were analyzed. transportThen, to reve structureal the ofposition Qinghai of andrailway Tibet transport was analyzed in transport further. systemSubsequently, of the TP, transport-modes-based in order to reveal the initial transpor effectst structure of the QTRof Qinghai on the developmentand Tibet was of analyzed Qinghai further.and Tibet, some indicators describing the socioeconomic situation were selected. During the selection process,Subsequently, the indicators in order that to could reveal reflectthe initial the effects socioeconomic of the QTR situation on the development of a region of were Qinghai the firstand Tibet,consideration. some indicators Data availability describing and the credibilitysocioecono weremic situation also considered were selected. since Qinghai During andthe Tibetselection are process,remote regions the indicators of China andthat theircould inventory reflect the records socioeconomic lag behind thosesituation of the of mid-east a region regions were ofthe China. first consideration.Eventually, GDP, Data population, availability industrial and credibility structure, we andre urbanizationalso considered level since were Qinghai all selected. and Tibet are remote regions of China and their inventory records lag behind those of the mid-east regions of China. Eventually, GDP, population, industrial structure, and urbanization level were all selected. Sustainability 2016, 8, 776 4 of 17 Sustainability 2016, 8, 776 4 of 17

Development of transport system in the TP ① railway transport capacity, ② transport structure

Influence

Socioe conomic factors of the TP ① GDP; ② population; ③ industrial structure; ④ urbanization level

Surrounding Non-surrounding regions regions Comparison at multiscale

Preliminarily reveal

Potential influence of QTR on economic performance of the TP

Figure 2 2.. Overview of the study.

Comparisons of the above-mentioned four factorsfactors between surrounding and non-surrounding areas of the QTRQTR beforebefore andand afterafter itsits constructionconstruction werewere implemented.implemented. For For GDP and population, comparisons werewere carriedcarried out out at at the the county county and and 1 1 km km scales. scales. At At the the county county scale, scale, there there are are 74 74 counties counties in Qinghaiin Qinghai and and 40 in40 Tibet in Tibet (i.e., (i.e., 114 in114 total) in total) [32]. [32]. The numberThe number of counties/cities of counties/cities surrounding surrounding or near or to near the QTRto the is QTR 22, accounting is 22, accounting for 19.30% for of the19.30% total of number the total of counties number in of Qinghai counties and in Tibet Qinghai combined. and Tibet GDP andcombined. population GDP density,and population and their density, annual and growth their rates annual (AGR) growth of surrounding rates (AGR) and of non-surroundingsurrounding and countiesnon-surrounding for the period counties 2000–2010 for the were period calculated. 2000–2010 At thewere 1 km calculated. scale, ten QTR-centeredAt the 1 km scale, buffers ten ranging QTR- incentered distance buffers from ranging 0 to 100 in km distance from the from railway 0 to 100 at 10km km from intervals the railway were at chosen 10 km tointervals analyze were the rateschosen of changeto analyze of GDP the rates and populationof change of density. GDP and The populati GDP andon population density. The density, GDP asand well population as the AGRs density, of each as buffer,well as werethe AGRs also analyzed.of each buffer, were also analyzed.

3.2. Data Sources StatisticalStatistical data data and and 1 1 km spatial data were major data sources for this study. The former were obtained mainly from the National Bureau of Statistics of China (NBSC). Detailed passenger and freight trafﬁctraffic data, GDP at the county scale, and and industrial industrial structure structure data were were obtained obtained from Qinghai Statistical Yearbooks and Tibet Statistical Yearbooks [[33,34].33,34]. Population data at the county scale for 2000, 2005, and 2010 were obtained from population statistical material at the county scale for the People’s Republic of China [[35–37].35–37]. It should should be be noted noted that that it it was was necessary necessary to to revise revise the the values values of ofthe the above above four four indicators indicators for the for theTanggula Tanggula Township Township located located in inthe the southwest southwest of of Qi Qinghai.nghai. From From the the perspect perspectiveive of of administrative divisions, the Tanggula Township is an enclave belonging to Golmud, but its GDP, population, and urbanization level level are are very very different different from from those those of of Golmud. Golmud. Thus, Thus, it was it was felt felt inappropriate inappropriate to use to usethe theGolmud Golmud values values to cover to cover the the Tanggula Tanggula Township. Township. Th Thee values values of of the the above above four four indicators indicators for the Tanggula Township werewere replacedreplaced byby thethe averageaverage values of surrounding counties, including Amdo, Bangor, Nierong, Zaduo,Zaduo, andand ZhiduoZhiduo counties.counties. The 1 kmkm GDPGDP andand population population datasets datasets for for China China for for 2000, 2000, 2005, 2005, and and 2010 2010 were were obtained obtained from from the Globalthe Global Change Change Research Research Data Data Publishing Publishing and Repositoryand Repository System System [38,39 [38,39].]. The1 The km 1 GDP km GDP datasets datasets were derivedwere derived from afrom spatial a spatial correlation correlation model thatmodel was that designed was designed based on based correlative on correlative relationships relationships between landbetween use andland three use and different three industrydifferent types.industry Precision types. veriﬁcationPrecision verification showed that showed the errors that inthe the errors spatial in GDPthe spatial data wereGDP indata the were range in6%–17% the range [39 6%–17%]. The 1 [39]. km populationThe 1 km population datasets for datasets China forfor 2000,China 2005, for 2000, and 20102005, wereand 2010 derived were from derived a multivariate from a multivariate statistical modelstatistical that model was established that was established based on the based correlation on the correlation relationships between population and land use types. Precision verification indicated that the errors in the spatial population data were in the range 4.5%–13.6% [38]. Sustainability 2016, 8, 776 5 of 17

The vector data of the XG and GL sections of the QTR were obtained from the Resources and Environment Dataset of China (at a scale of 1:400 million) [32].

4. Results

4.1. Evolution of the Transport System in the TP with the Construction of the QTR SustainabilityFirst, the2016 ,passenger8, 776 traffic, freight traffic, passenger-kilometers, and freight-kilometers of5 ofthe 17 QTR for the period 1982–2013 are presented below. Then the passenger- and freight traffic-based transportrelationships structures between of population the TP for and the land period use types. 1990–2013 Precision are verification presented. indicated Finally, thatthe thepassenger- errors in kilometersthe spatial populationand freight-kilometers data were in based the range transport 4.5%–13.6% structures [38]. of the TP for the period 1990–2013 are analyzed.The vector data of the XG and GL sections of the QTR were obtained from the Resources and Environment Dataset of China (at a scale of 1:400 million) [32]. 4.1.1. Analysis of Transport Capacity of the QTR for 1982–2013 4. ResultsOn the whole, passenger traffic on the QTR increased for the period 1982–2013 (Figure 3a). A marked increase can be seen after 1984, when the XG section opened to traffic, and after 2006, when 4.1. Evolution of the Transport System in the TP with the Construction of the QTR the GL section opened. One thing to note in particular is the decrease in passenger volume in 2008 comparedFirst, thewith passenger previous traffic, years. freight This was traffic, due passenger-kilometers, to civil unrest (the andMarch freight-kilometers 14 riot), which of involved the QTR beatings,for the period vandalism, 1982–2013 looting are presentedof shops, below.and arson. Then Damage the passenger- was estimated and freight at more traffic-based than 244 transport million Yuanstructures (basic of unit the of TP currency for the periodin China) 1990–2013 [40]. are presented. Finally, the passenger-kilometers and freight-kilometersRegarding freight based traffic transport on structuresthe QTR in of thethe TPperiod for the 1982–2013, period 1990–2013 a more arerapid analyzed. and steadily increasing tendency can be seen. During the 1980s and 1990s, freight traffic and its rate of increase 4.1.1. Analysis of Transport Capacity of the QTR for 1982–2013 were both low. In the 21st century, freight traffic presents a nearly exponential growth trend, which potentiallyOn the can whole, be explained passenger by traffic the construction on the QTR of increased the GL section for the in period 2001 and 1982–2013 the completed (Figure 3linea). openingA marked to increasetraffic in can2006. be In seen addition, after 1984,more whentypes theof freight XG section were transported opened to traffic, to Tibet and via after the QTR. 2006, Forwhen example, the GL sectionpetroleum opened. products, One thing including to note gas in particular and diesel, is the which decrease previously in passenger could volume only be in transported2008 compared to Tibet with previousby highway years. and This pipeline, was due were to civiltransported unrest (the by Marchrailway 14 for riot), thewhich first time involved on 5 Maybeatings, 2014 vandalism,[41]. This not looting only cut of shops,transportation and arson. cost Damages but also was made estimated the tran atsportation more than of244 petroleum million productsYuan (basic much unit safer. of currency in China) [40].

(a) (b) 7 4 7 30

6 6 25 3 t) t-km)

7 5 9

passenger) 5 passenger-km) 20 6 9 4 4 2 15 3 3 10 2

1 Freight Traffic (10

2 1 5 Freight-kilometers (10 Passenger Traffic (10

1 0 Passenger-kilometers (10 0 0 1980 1985 1990 1995 2000 2005 2010 2015 1980 1985 1990 1995 2000 2005 2010 2015 Year Year

Figure 3 3.. TrafficTrafﬁc volume and passengerpassenger (freight)-kilometers of the QTR for 1982–2013. ( a) Passenger and freight traffics; trafﬁcs; (b) Passenger-kilometers andand freight-kilometers.freight-kilometers.

Passenger-kilometersRegarding freight traffic and on thefreight-kilometers QTR in the period for 1982–2013, 1982–2013 a more are rapid presented and steadily in Figure increasing 3b. Comparedtendency can with be seen.the highway, During the the 1980s railway and is 1990s, the ma freightjor long-distance traffic and its mode rate of of increase transport. were Therefore, both low. comparedIn the 21st with century, passenger freight trafficand freight presents traffic a nearly values exponential alone, the growthdata for trend, the passenger-kilometers which potentially can and be freight-kilometersexplained by the construction of the QTR give of the a better GL section insigh int into 2001 railway and the transport completed since line they opening are the to trafficproduct in of2006. transport In addition, traffic more and transport types of freight distance. were From transported Figure 3b, to two Tibet phases via the for QTR. both For passenger-kilometers example, petroleum andproducts, freight-kilometers including gas can and be diesel, identified: which previouslythe slow increasing could only phase be transported for 1982–2005 to Tibet and by the highway rapid, increasingand pipeline, phase were for transported2006–2013, which by railway is potentially for the firstrelated time to onthe5 opening May 2014 of the [41 ].GL This section not to only traffic cut transportation costs but also made the transportation of petroleum products much safer. Passenger-kilometers and freight-kilometers for 1982–2013 are presented in Figure3b. Compared with the highway, the railway is the major long-distance mode of transport. Therefore, compared with passenger and freight traffic values alone, the data for the passenger-kilometers and freight-kilometers of the QTR give a better insight into railway transport since they are the product of transport traffic and transport distance. From Figure3b, two phases for both passenger-kilometers and freight-kilometers Sustainability 2016, 8, 776 6 of 17

Sustainabilitycan be identified: 2016, 8, 776 the slow increasing phase for 1982–2005 and the rapid, increasing phase6 of for 17 2006–2013, which is potentially related to the opening of the GL section to traffic in 2006. In 1982, 9 inthe 2006. passenger-kilometers In 1982, the passenger-kilometers and freight-kilometers and freight-kilometers were 0.2 ˆ 10 9werepassenger-km 0.2 × 10 passenger-km and 0.6 ˆ 10 and9 t-km, 0.6 9 9 9 ×respectively, 10 t-km, respectively, which increased which increased to 1.9 ˆ 10 to9 1.9passenger-km × 10 passenger-km and 9.9 andˆ 10 9.99 t-km, × 10 t-km, respectively, respectively, in 2005, in 9 9 2005,increases increases of 1.7 ofˆ 101.79 ×passenger-km 10 passenger-km and 9.3 andˆ 109.39 ×t-km, 10 t-km, respectively. respectively. In 2013, In the2013, corresponding the corresponding values 9 9 9 valueswere 6.6 wereˆ 10 6.69 passenger-km × 10 passenger-km and 27.1 andˆ 27.1109 ×t-km, 10 t-km, which which increased increased by 4.7 byˆ 4.710 9×passenger-km 10 passenger-km and 9 and17.2 17.2ˆ 10 ×9 10t-km t-km respectively. respectively. The The annual annual increases increases in passenger-kilometersin passenger-kilometers and and freight-kilometers freight-kilometers for forthe the second second phase phase were were 9.1 9.1 and and 9.6 9.6 times times those those of the of the first first phase, phase, respectively. respectively.

4.1.2.4.1.2. Passenger Passenger and and Freight Freight Traffic-Based Traffic-Based Transport StructureStructure inin thethe TPTP forfor 1990–20131990–2013 FigureFigure 44 showsshows thethe passenger-passenger- andand freightfreight traffic-basedtraffic-based transporttransport structurestructure ofof thethe TP.TP. FromFrom FigureFigure 44a,a, it can be be seen seen that that highway highway passenger passenger tr trafficaffic maintained maintained a dominant a dominant position position in the in the TP TPfor 1990–2013,for 1990–2013, and andthe theaverage average percentage percentage of oftotal total traffic traffic for for that that period period was was 91.31%. 91.31%. For For railway railway passengerpassenger traffic, traffic, the the percentage percentage was was low low in ineach each year year and and the theaverage average value value was wasonly only7.22%. 7.22%. The averageThe average ratios ratios of highway, of highway, railway, railway, and air and passenger air passenger traffic trafficwere 91.31:7.22:1.48, were 91.31:7.22:1.48, respectively. respectively. Note thatNote the that significant the significant decrease decrease in the in percentage the percentage of railway of railway passenger passenger traffic traffic after after 2008 2008 was was probably probably relatedrelated to to the the outbreak outbreak of civil unrest mentioned above [[40].40].

(a) 100 90 80 70 60 50 40 30 20 Highway Railway Air 10 Percent of passenger traffic(%) 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Year (b) 100 90 80 70 60 50 40 30 20 Highway Railway Air Pipeline Percent of freight traffic(%) 10 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Year

FigureFigure 4 4.. TrafficTraffic volume-based transport structure inin thethe TPTP forfor 1990–2013.1990–2013. ((aa)) PassengerPassenger traffic-based;traffic- based;(b) freight (b) freight traffic-based. traffic-based.

Figure 4b shows the freight traffic-based transport structure of the TP. The average ratios of Figure4b shows the freight traffic-based transport structure of the TP. The average ratios of highway, railway, air, and pipeline freight traffic were 78.66:15.57:0.01:2.76, respectively, which highway, railway, air, and pipeline freight traffic were 78.66:15.57:0.01:2.76, respectively, which indicated that highway freight traffic remained in a dominant position. For railway freight traffic, the indicated that highway freight traffic remained in a dominant position. For railway freight traffic, percentage was low in each year and the average value was 15.57%, but it can be seen that the the percentage was low in each year and the average value was 15.57%, but it can be seen that the percentage increased in the period 2000–2013, which matches well with the time point of the GL percentage increased in the period 2000–2013, which matches well with the time point of the GL section’s construction and opening to traffic. In 2000, the ratios of highway, railway, air, and pipeline section’s construction and opening to traffic. In 2000, the ratios of highway, railway, air, and pipeline freight traffic were 80.34:15.76:0.02:3.88, respectively, and in 2013 they were 72.65:25.86:0.02:1.47, respectively. The proportion of railway freight traffic increased by 10.10%.

4.1.3. Passenger- and Freight-Kilometers Based Transport Structures in the TP for 1990–2013 Sustainability 2016, 8, 776 7 of 17

As mentioned above, the railway is more often used than the highway to transport passengers andSustainability freight2016 over, 8 ,longer 776 distances. Therefore, the passenger-kilometers and freight-kilometers based7 of 17 transport structures are also important aspects of the transport structure (Figure 5). From Figure 5a it can be seen that the percentage of highway passenger-kilometers decreased graduallyfreight trafﬁc between were 1990 80.34:15.76:0.02:3.88, and 2013 as a whole, respectively, from 61.56% and in in 2013 1990 they to were38.99% 72.65:25.86:0.02:1.47, in 2013, and the percentagerespectively. of Theair passenger-kilometers proportion of railway gradually freight trafﬁc incr increasedeased, from by 6.67% 10.10%. in 1990 to 22.96% in 2013. In terms4.1.3. Passenger-of railway passenger-kilometers, and Freight-Kilometers first Based there Transport was a decreasing Structures trend in the for TP 1990–1996 for 1990–2013 followed by an increasing trend for 1996–2007. In 2007, one year after the QTR was fully opened to traffic, the percentageAs mentioned of railway above, passenger-kilometers the railway is more reac oftenhed 42.22%, used than which the highwaywas the maximum to transport for passengers the whole studyand freight period. over After longer that, distances. a relatively Therefore, stable pe theriod passenger-kilometers for 2008–2013 followed. and freight-kilometers The average ratios based of highway,transport railway, structures and are air also passenger-kilometers important aspects of were the transport56.46:31.10:12.44, structure respectively, (Figure5). for 1990–2013.

(a) 100 90 80 70 60 50 40 30 20 Highway Railway Air 10

Percent of passenger-kilometers (%) 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 (b) Year 100 90 80 70 60 50 40 30 20 Highway Railway Air Pipeline 10 Percent of freight-kilometers (%) 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Year

FigureFigure 5. 5. PassengerPassenger (freight)-kilometers-based (freight)-kilometers-based tran transportsport structure structure in the in theTP for TP 1990–2013. for 1990–2013. (a) Passenger-kilometers-based;(a) Passenger-kilometers-based; (b) (freight-kilometers-based.b) freight-kilometers-based.

The freight-kilometers based transport structure of the TP is presented in Figure 5b. It can be From Figure5a it can be seen that the percentage of highway passenger-kilometers decreased seen that both the railway and the highway maintain their dominant positions. The average ratios of gradually between 1990 and 2013 as a whole, from 61.56% in 1990 to 38.99% in 2013, and the percentage highway, railway, air, and pipeline freight-kilometers were 45.00:48.99:0.15:5.86, respectively. of air passenger-kilometers gradually increased, from 6.67% in 1990 to 22.96% in 2013. In terms Railway freight-kilometers were higher than highway freight-kilometers, not only the average value of railway passenger-kilometers, ﬁrst there was a decreasing trend for 1990–1996 followed by an but also for 18 of the 24 time points between 1990 and 2013. increasing trend for 1996–2007. In 2007, one year after the QTR was fully opened to trafﬁc, the Taking the above analysis and the fact that the initial railway network of Tibet will be percentage of railway passenger-kilometers reached 42.22%, which was the maximum for the whole constructed and the length of the railway in Tibet will reach 1300 km by 2020 into account, we can study period. After that, a relatively stable period for 2008–2013 followed. The average ratios of predict that the highway-dominated transport system will break up and an integrated transport highway, railway, and air passenger-kilometers were 56.46:31.10:12.44, respectively, for 1990–2013. system will form in the TP. Since the QTR and the coming railway network are environmentally The freight-kilometers based transport structure of the TP is presented in Figure5b. It can be friendly, it can also be said that a more sustainable transport system will come to the TP. seen that both the railway and the highway maintain their dominant positions. The average ratios of highway, railway, air, and pipeline freight-kilometers were 45.00:48.99:0.15:5.86, respectively. Railway 4.2. Comparisons of GDP freight-kilometers were higher than highway freight-kilometers, not only the average value but also for 18GDP of the is usually 24 time used points to betweendescribe 1990the economic and 2013. performance of a country or region. Based on the availableTaking data, the the above GDP analysis at the county and the and fact 1 thatkm scales the initial for 2000, railway 2005, network and 2010 of Tibetwere willanalyzed be constructed to reveal and the length of the railway in Tibet will reach 1300 km by 2020 into account, we can predict that the highway-dominated transport system will break up and an integrated transport system will form in Sustainability 2016, 8, 776 8 of 17 the TP. Since the QTR and the coming railway network are environmentally friendly, it can also be said that a more sustainable transport system will come to the TP.

4.2. Comparisons of GDP

SustainabilityGDP is2016 usually, 8, 776 used to describe the economic performance of a country or region. Based8 of on 17 the available data, the GDP at the county and 1 km scales for 2000, 2005, and 2010 were analyzed to revealthe economic the economic performance performance of the ofsurrounding the surrounding and non-surrounding and non-surrounding areas areasof the of QTR the QTRbefore before and andfollowing following its opening. its opening. InIn 2000,2000, onlyonly thethe XGXG sectionsection waswas openopen toto traffic.traffic. The comparison of GDP between 2000 and 2005 reflectsreflects thethe effectseffects ofof the XG section on the econom economicic performance of its its surr surroundingounding areas. areas. Similarly, Similarly, thethe comparisoncomparison betweenbetween 2000/20052000/2005 and and 2010 2010 reflects reflects the the impacts impacts of the entire line, especially the GL section, onon thethe economyeconomy ofof itsits surroundingsurrounding areas.areas.

4.2.1. Comparisons at the County Scale Figure 66 shows the the GDP GDP density density values values of of the the TP TP for for2000, 2000, 2005, 2005, and and2010. 2010. In terms In terms of the ofspatial the spatialpatterns patterns of GDP of density, GDP density, the counties the counties with high with va highlues were values distributed were distributed mainly mainlyin the Yellow in the YellowRiver– River–HuangshuiHuangshui River RiverValley Valley in 2000. in 2000. The TheGDP GDP densit densityy of ofall all counties counties surrounding surrounding the the XG XG sectionsection increasedincreased overover the period 2000–2005,2000–2005, especially in the Yellow River–Huangshui River Valley. At the same time, the GL section was under construction, and the GDP density of its surrounding areas also presentedpresented a slightlyslightly increasingincreasing trend. In 2010, the GDP density of the counties along the entire lineline increasedincreased markedly,markedly, especially especially along along the the GL GL section, section, compared compared with with 2000 2000 and and 2005. 2005. However, However, as as one one of theof the least least populated populated regions regions of China, of China, the initial the portioninitial portion of the GLof sectionthe GL (includingsection (including Zhiduo, Qumalai,Zhiduo, Anduo,Qumalai, and Anduo, Nierong and counties) Nierong sawcounties) a low ratesaw ofa low increase rate of in increase GDP density. in GDP This density. is because This two is because nature reserves,two nature Kekexili reserves, (Hoh Kekexili Xil) and (Hoh Qinghai Xil) Sanjiangyuan,and Qinghai Sanjiangyuan, were established were along established this section along [5], andthis large-scalesection [5], exploitationand large-scale of natureexploitation reserves of nature has been reserves forbidden has been by theforbidden government by the as government a measure toas implementa measure to sustainable implement development. sustainable development.

0250 500 0250 500 0250 500 km km km GDP Value (104 Yuan/km 2 ) 00.1 0.2 0.3 0.4 0.5 0.6 0.8 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10 15 20 50 100 200 5001000 4500 Figure 6.6. Spatial distributions of GDP density of QinghaiQinghai and Tibet for 2000, 2005, and 2010 at thethe countycounty scale.scale.

For comparison, both both the the average average GDP GDP density density and and the theAGR AGR of GDP of GDP of surrounding of surrounding and non- and non-surroundingsurrounding counties counties were calculated were calculated for 2000, for 2005 2000,, and 2005,2010 (Table and2010 1). The (Table average1). GDP The density average of GDPsurrounding density counties of surrounding was greater counties than that was of greater non-surrounding than that of counties non-surrounding over three time counties points. over In 2 three2000, the time GDP points. values In for 2000, surrounding the GDP and values non-surrounding for surrounding counties and were non-surrounding 2.71 yuan/km countiesand 0.88 2 2 wereyuan/km 2.71, yuan/km respectively,2 and a 0.88ratio yuan/km of 3.1:1. 2By, respectively, 2010, they ahad ratio increased of 3.1:1. to By 20.61 2010, and they 5.69 had yuan/km increased, torespectively, 20.61 and 5.69 a ratio yuan/km of 3.6:1.2, respectively, a ratio of 3.6:1.

Table 1. Mean GDP density and its AGR of surrounding and non-surrounding counties of the QTR for 2000, 2005, and 2010.

Mean GDP Density AGR of Mean GDP Density (%) Region (yuan/km2) 2000 2005 2010 2000–2005 2005–2010 2000–2010 Surrounding counties 2.71 7.69 20.61 23.14 21.80 22.47 Non-surrounding 0.88 2.33 5.69 21.61 19.52 20.56 counties

The average GDP density showed an increasing tendency both for surrounding and non- surrounding counties for the period 2000–2010. However, the AGR values of surrounding counties Sustainability 2016, 8, 776 9 of 17

Table 1. Mean GDP density and its AGR of surrounding and non-surrounding counties of the QTR for 2000, 2005, and 2010.

Mean GDP Density (yuan/km2) AGR of Mean GDP Density (%) Region 2000 2005 2010 2000–2005 2005–2010 2000–2010 Surrounding counties 2.71 7.69 20.61 23.14 21.80 22.47 Non-surrounding 0.88 2.33 5.69 21.61 19.52 20.56 counties

SustainabilityThe 2016 average, 8, 776 GDP density showed an increasing tendency both for surrounding9 of and17 non-surrounding counties for the period 2000–2010. However, the AGR values of surrounding werecounties greater were than greater those of than non-surrounding those of non-surrounding counties for countiesthe periods for 2000–2005, the periods 2005–2010, 2000–2005, and 2005–2010, 2000– 2010,and which 2000–2010, was probably which was related probably to the related opening to the to openingtraffic of to the trafﬁc QTR of (Table the QTR 1). For (Table instance,1). For instance,during theduring first ten the months ﬁrst ten of months its operation, of its operation, trade between trade between Tibet and Tibet external and external regions regions increased increased by 75%, by 75%,to 2.6to billion 2.6 billion yuan yuan [4]. [4].

4.2.2.4.2.2. Comparisons Comparisons at atthe the 1 km 1 km Scale Scale ComparedCompared with with the the counties counties of ofthe the mid-east mid-east region regionss of ofChina, China, the the land land areas areas of ofmost most counties counties of of thethe TP TP are are very very large. large. The The average average area area of of counties counties in in China China is is ~0.34 ~0.34 ×ˆ 10104 km4 km2, while2, while the the average average area area ofof counties counties in in Qinghai Qinghai and and Tibet Tibet is is ~1.69 ~1.69 × ˆ10104 km4 km2. Thus,2. Thus, the the county-scale county-scale comparison comparison is isinadequate inadequate toto reflect reﬂect the the details details of of GDP GDP growth growth difference differencess between between surrounding surrounding and and non-surrounding non-surrounding areas. areas. Therefore,Therefore, an an analysis analysis at at the the 1 1km km scale scale was was carried carried out out further. further. InIn this this study, study, of ofthe the QTR, QTR, ten ten buffer buffer areas areas of the of the 1 km 1 km GDP GDP dataset dataset fromfrom 0 to 0 100 to 100 km km with with 10 10km km intervalsintervals (Figure (Figure 7)7 )were were extracted. extracted. It Itcan can be be seen seen that, that, except except for for Xining Xining and and Lhasa Lhasa and and their their surroundingsurrounding regions, regions, the the other other buffer buffer regions regions of ofthe the QTR QTR had had a lower a lower GDP GDP value value for for the the three three time time points.points. From From 2000 2000 to to2005, 2005, the the GDP GDP value value of ofXining Xining and and its its near near areas areas increased increased markedly, markedly, while while that that ofof the the other other regions regions changed changed little. little. Between Between 2005 2005 and and 2010, 2010, the the GDP GDP values values of ofXining Xining and and Lhasa Lhasa and and theirtheir surrounding surrounding areas areas increased increased even even more more markedly. markedly.

GDP (104Yuan/k m 2) 0 5 10 20 30 40 50 100 200 300 400 500 600 700 800 900 1000 1500 3000 6000 45000 FigureFigure 7. 7.GDPGDP distribution distribution in insurrounding surrounding areas areas of ofthe the QTR QTR for for 2000, 2000, 2005, 2005, and and 2010 2010 at atthe the 1 km 1 km scale. scale.

ForFor each each buffer buffer region region centered centered on on the the QTR, QTR, the the mean mean GDP GDP and and the the AGR AGR of ofGDP GDP for for the the periods periods 2000–2005,2000–2005, 2005–2010, 2005–2010, and and 2000–2010 2000–2010 were were calculated calculated (Table (Table 2).2). It Itcan can be be seen seen that that GDP GDP was was high high in in thethe 0–10 0–10 km km buffer buffer region region and and lower lower in in other other buffer buffer regions. regions. In In other other words, words, GDP GDP was was concentrated concentrated mainlymainly in inthe the 0–10 0–10 km km region region in in2000, 2000, 2005, 2005, and and 2010. 2010. Overall, Overall, the the closer closer a region a region was was to tothe the railway, railway, thethe greater greater the the GDP GDP value value in in each each time time period. period. In terms of the AGR of GDP, except for that of the 90–100 km buffer region for 2000–2005, all other Table 2. GDP density values and AGR of GDP of each buffer regions for 2000, 2005, 2010. values were positive. In general, for all three time periods, the closer to the railway, the greater the AGR.Buffer The mean AGRMean values GDP of all(104 ten Yuan/km buffer2) regions for 2000–2005 AGR and of 2005–2010 GDP (%) were 17.95% and 22.39%,Region respectively, 2000 which meant2005 that following 2010 the opening 2000–2005 of the entire2005–2010 line to trafﬁc,2000–2010 the GDP of the0–10 surrounding km areas 30.06 of the QTR 77.42 had a greater 160.96 growth rate 20.83 than before the 15.76 opening. 18.27 10–20 km 4.90 18.50 67.19 30.45 29.43 29.94 20–30 km 3.56 9.05 41.67 20.48 35.73 27.88 30–40 km 4.64 12.02 35.23 20.95 24.00 22.47 40–50 km 2.71 6.51 19.82 19.14 24.92 21.99 50–60 km 1.89 3.51 8.11 13.18 18.20 15.67 60–70 km 3.22 6.17 12.61 13.90 15.36 14.63 70–80 km 2.13 7.71 20.20 29.38 21.25 25.25 80–90 km 1.58 2.89 6.57 12.89 17.84 15.34 90–100 km 3.11 2.86 7.53 −1.73 21.40 9.22

In terms of the AGR of GDP, except for that of the 90–100 km buffer region for 2000–2005, all other values were positive. In general, for all three time periods, the closer to the railway, the greater the AGR. The mean AGR values of all ten buffer regions for 2000–2005 and 2005–2010 were 17.95% Sustainability 2016, 8, 776 10 of 17

Table 2. GDP density values and AGR of GDP of each buffer regions for 2000, 2005, 2010.

Mean GDP (104 Yuan/km2) AGR of GDP (%) Buffer Region 2000 2005 2010 2000–2005 2005–2010 2000–2010 0–10 km 30.06 77.42 160.96 20.83 15.76 18.27 10–20 km 4.90 18.50 67.19 30.45 29.43 29.94 20–30 km 3.56 9.05 41.67 20.48 35.73 27.88 30–40 km 4.64 12.02 35.23 20.95 24.00 22.47 40–50 km 2.71 6.51 19.82 19.14 24.92 21.99 50–60 km 1.89 3.51 8.11 13.18 18.20 15.67 Sustainability60–70 2016 km, 8, 776 3.22 6.17 12.61 13.90 15.36 14.63 10 of 17 70–80 km 2.13 7.71 20.20 29.38 21.25 25.25 and 22.39%,80–90 respectively, km 1.58which meant 2.89 that following 6.57 the opening 12.89 of the entire 17.84 line to traffic, 15.34 the GDP of the surrounding90–100 km areas 3.11of the QTR had 2.86 a greater growth 7.53 rate ´than1.73 before the 21.40 opening. 9.22

4.3. Comparisons of Population The railway is capable of transporting people to its surrounding areas to engageengage inin variousvarious socioeconomicsocioeconomic activitiesactivities [[42].42]. By analyzing changes in the population of these surrounding areas,areas, thethe inﬂuencesinfluences ofof thethe QTRQTR onon populationpopulation distributiondistribution inin thethe TPTP cancan bebe determined.determined.

4.3.1. Comparisons at the County ScaleScale The population densities of both Qinghai and Tibet atat thethe countycounty scalescale forfor 2000,2000, 2005,2005, andand 20102010 are presentedpresented inin FigureFigure8 8.. InIn 2000,2000, thethe populationpopulation waswas distributeddistributed mainlymainly inin thethe valleyvalley regionsregions ofof thethe YellowYellow River–HuangshuiRiver–Huangshui River River and and their their near near regions regions and and the valleythe valley regions regions of the of Brahmaputra the Brahmaputra River andRiver its and tributaries. its tributaries. The population The population densities densities of other of other regions regions were lowerwere lower than thesethan these values. values. For 2005 For and2005 2010, and a2010, similar a similar pattern pattern of population of population density de couldnsity be could detected. be detected. In other words,In other between words, 2000between and 2010,2000 and the populations2010, the populations of both Qinghai of both andQinghai Tibet and were Tibet not were concentrated not concentrated in the surrounding in the surrounding areas of theareas QTR. of the QTR.

0500250 0500250 0500250 km km km Population density 2 (inhabitants/km ) 0.2 0.40.6 0.8 1.0 1.5 2.02.5 3.0 3.5 4.0 4.55.0 6.0 7.0 8.0 9.010 15 20 25 30 70 150 2000

Figure 8.8. Population densities of Qinghai and TibetTibet atat thethe countycounty scalescale forfor 2000,2000, 2005,2005, andand 2010.2010.

For comparison, the the mean mean population population densities densities and and the the AGR AGR of populati of populationon of surrounding of surrounding and andnon-surrounding non-surrounding counties counties were ca werelculated calculated for 2000, for 2005, 2000, and 2005, 2010 and (Table 2010 3).(Table The mean3). Thepopulation mean populationdensity of surrounding density of surrounding counties was counties greater than was greaterthat of non-surrounding than that of non-surrounding counties. For 2000, counties. the Forcorresponding 2000, the corresponding values for surrounding values for surrounding and non-surrounding and non-surrounding counties countieswere 4.70 were and 4.70 and3.55 3.55inhabitants/km inhabitants/km2, respectively,2, respectively, a ratio a ratioof 1.32:1. of 1.32:1. By 2010, By 2010, these these values values had increased had increased to 5.05 to 5.05and and4.14 4.14inhabitants/km inhabitants/km2, respectively,2, respectively, a ratio a ratioof 1.22:1. of 1.22:1. The average population density showed an increasing tendency both for surrounding and non-surroundingTable 3. Mean countiespopulation for densities the period and AGR 2000–2010. of population However, of surrounding the AGR and of non-surrounding surrounding counties was lowercounties than of the that QTR of non-surrounding for 2000, 2005, and counties 2010. for the periods 2000–2005, 2005–2010, and 2000–2010. The biggest difference occurredMean Population for the period Density 2000–2005. The AGR of non-surrounding areas was 1.08%, while that for surrounding areas was only 0.05%. From aAGR sustainability of Population perspective, (%) the fact Region (Inhabitants/km2) 2000 2005 2010 2000–2005 2005–2010 2000–2010 Surrounding 4.70 4.71 5.05 0.05 1.40 0.72 counties Non-surrounding 3.55 3.75 4.14 1.08 2.02 1.55 counties

The average population density showed an increasing tendency both for surrounding and non- surrounding counties for the period 2000–2010. However, the AGR of surrounding counties was lower than that of non-surrounding counties for the periods 2000–2005, 2005–2010, and 2000–2010. The biggest difference occurred for the period 2000–2005. The AGR of non-surrounding areas was 1.08%, while that for surrounding areas was only 0.05%. From a sustainability perspective, the fact Sustainability 2016, 8, 776 11 of 17 that population grew more rapidly in non-surrounding counties than surrounding counties indicated that the QTR could be seen as a good example of sustainable transport.

Table 3. Mean population densities and AGR of population of surrounding and non-surrounding counties of the QTR for 2000, 2005, and 2010.

Mean Population Density AGR of Population (%) (Inhabitants/km2) Region 2000 2005 2010 2000–2005 2005–2010 2000–2010 Sustainability 2016, 8, 776 11 of 17 Surrounding counties 4.70 4.71 5.05 0.05 1.40 0.72 Non-surrounding 3.55 3.75 4.14 1.08 2.02 1.55 that populationcounties grew more rapidly in non-surrounding counties than surrounding counties indicated that the QTR could be seen as a good example of sustainable transport. 4.3.2. Comparisons at the 1 km Scale 4.3.2. Comparisons at the 1 km Scale Like GDP analyses at the 1 km scale, the population comparison at the county scale is inadequate Like GDP analyses at the 1 km scale, the population comparison at the county scale is inadequate to reﬂect the details of differences in population changes between surrounding and non-surrounding to reflect the details of differences in population changes between surrounding and non-surrounding areas. Therefore, further comparisons at the 1 km scale were carried out. areas. Therefore, further comparisons at the 1 km scale were carried out. In this study, of the QTR, ten buffer areas of the 1 km population dataset from 0 to 100 km In this study, of the QTR, ten buffer areas of the 1 km population dataset from 0 to 100 km with with 10 km intervals were extracted (Figure9). It can be seen that the surrounding areas of the QTR 10 km intervals were extracted (Figure 9). It can be seen that the surrounding areas of the QTR were were sparsely populated and the population was concentrated mainly in Xining and Lhasa and their sparsely populated and the population was concentrated mainly in Xining and Lhasa and their surrounding areas. In terms of changes in the spatial distribution of population, no obvious changes surrounding areas. In terms of changes in the spatial distribution of population, no obvious changes could be detected for the period 2000–2010. could be detected for the period 2000–2010.

Population (inhabitants/km2) 0 5 10 20 30 40 50 100 200 300 400 500 600 700 800 900 1000 1500 3000 6000 62000 Figure 9. PopulationPopulation distribution distribution of of surrounding surrounding area areass of of the the QTR QTR at at the the 1 1 km km scale scale for for 2000, 2000, 2005, 2005, and 2010.

InIn addition, statistical statistical analysis analysis was was carried carried ou outt for for the the ten ten buffer buffer areas areas centered centered on on the QTR (Table4 4).). This This shows shows that that for for 2000, 2000, 2005, 2005, and and 2010, 2010, population population density density was highwas forhigh the for 0–10 the km 0–10 buffer km bufferregion region and lower and forlower other for buffer other regions. buffer regions. That is toTh say,at is the to populationsay, the population was concentrated was concentrated mainly in mainlythe 0–10 in km the buffer 0–10 km region. buffer region. In terms of the AGR of population between 2000 and 2005, various ﬂuctuations could be detected Table 4. Population densities of buffer regions and AGR of population for 2000, 2005, and 2010. and no regular patterns could be found for any of the buffer regions. For the period 2005–2010, all buffer regions had positivePopulation AGR values Density that were more stable than those for the period 2000–2005. AGR (%) The changesBuffer in AGR for 2000–2010(Inhabitants/km were2 similar) to those of 2000–2005 and 2005–2010. The above analyses2000 indicated2005 that population2010 was 2000 not–2005 concentrated 2005–2010 in surrounding 2000–2010 areas of the QTR for0–10 the km period 2000–201037.51 whether 43.10 at the 49.49 county scale 2.82 or the 1 km 2.80 scale. As an environmentally 2.81 and culturally10–20 km sensitive 11.15 area, the 12.51 TP’s low 12.51 population 2.33 density is actually 0.00 favorable to 1.16 sustainable 20–30 km 10.97 10.69 10.69 −0.51 0.00 −0.25 development. In addition, one possible reason for this is the sustainable development policies 30–40 km 12.56 13.38 13.65 1.27 0.40 0.84 implemented here, including Kekexili (Hoh Xil) and Qinghai Sanjiangyuan nature reserves [5] and 40–50 km 8.14 7.33 8.14 −2.09 2.13 0.00 unstaffed50–60 railway km stations. 7.84 Another 5.68 possible 6.76 reason is that−6.25 the operating3.55 period of the railway−1.47 appears to have60–70 not km been long 8.59 enough to 13.95 generate substantial 14.76 changes 10.20 in population 1.13 distribution. 5.57 70–80 km 9.83 13.29 14.08 6.21 1.17 3.66 80–90 km 5.25 4.20 4.47 −4.36 1.22 −1.61 90–100 km 8.31 7.53 9.08 −1.95 3.83 0.90

In terms of the AGR of population between 2000 and 2005, various fluctuations could be detected and no regular patterns could be found for any of the buffer regions. For the period 2005–2010, all buffer regions had positive AGR values that were more stable than those for the period 2000–2005. The changes in AGR for 2000–2010 were similar to those of 2000–2005 and 2005–2010. The above analyses indicated that population was not concentrated in surrounding areas of the QTR for the period 2000–2010 whether at the county scale or the 1 km scale. As an environmentally and culturally sensitive area, the TP’s low population density is actually favorable to sustainable development. In addition, one possible reason for this is the sustainable development policies implemented here, including Kekexili (Hoh Xil) and Qinghai Sanjiangyuan nature reserves [5] and Sustainability 2016, 8, 776 12 of 17

Table 4. Population densities of buffer regions and AGR of population for 2000, 2005, and 2010.

Population Density (Inhabitants/km2) AGR (%) Buffer 2000 2005 2010 2000–2005 2005–2010 2000–2010 0–10 km 37.51 43.10 49.49 2.82 2.80 2.81 10–20 km 11.15 12.51 12.51 2.33 0.00 1.16 20–30 km 10.97 10.69 10.69 ´0.51 0.00 ´0.25 30–40 km 12.56 13.38 13.65 1.27 0.40 0.84 40–50 km 8.14 7.33 8.14 ´2.09 2.13 0.00 50–60 km 7.84 5.68 6.76 ´6.25 3.55 ´1.47 Sustainability60–70 2016 km, 8, 776 8.59 13.95 14.76 10.20 1.13 5.57 12 of 17 70–80 km 9.83 13.29 14.08 6.21 1.17 3.66 80–90 km 5.25 4.20 4.47 ´4.36 1.22 ´1.61 unstaffed railway stations. Another possible reason is that the operating period of the railway 90–100 km 8.31 7.53 9.08 ´1.95 3.83 0.90 appears to have not been long enough to generate substantial changes in population distribution.

4.4.4.4. Comparisons of Industrial StructureStructure IndustrialIndustrial structurestructure is is the the basis basis and and core core of economic of economic structure, structure, and it and can beit can inﬂuenced be influenced by railway by transportationrailway transportation [43]. The [43]. industrial The industrial structure structure of Qinghai–Tibet of Qinghai–Tibet is illustrated is illustrated in Figure in 10 Figure. 10.

100 90 primary industry secondary industry tertiary industry 80 70 60 50 40 30 20

GDP proportion of each industry (%) 10 0 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 Year Figure 10.10. Industrial structure ofof QinghaiQinghai andand TibetTibet forfor 1984–2012.1984–2012.

TheThe productionproduction values values of of primary primary industry industry (PI), (PI), secondary secondary industry industry (SI), (SI), and and tertiary tertiary industry industry (TI) had(TI) beenhad been increasing increasing for both for Qinghai both Qinghai and Tibet and between Tibet between 1984 and 1984 2012 and [31, 322012], but [31,32], in terms but of in industrial terms of structure,industrial thestructure, proportion the ofproporti PI decreasedon of PI continuously decreased incontinuously the same period, in the while same the period, proportion while of the SI andproportion TI increased of SI overalland TI (Figure increased 10). overall In 1984, (Figure the ratios 10). of In PI, 1984, SI, and the TI ratios were 34.2:32.3:33.5.of PI, SI, and ForTI 2012,were the34.2:32.3:33.5. ratios were For 9.9:51.5:38.6, 2012, the andratios the were combined 9.9:51.5:3 percentage8.6, and of the SI andcombined TI exceeded percentage 90%. of SI and TI exceededIn terms 90%. of the different stages of construction, for 1984–2001 when construction of the GL section had notIn terms begun, of the the percentage different stages of PI decreased of constructi continuously,on, for 1984–2001 that of TI when increased construction continuously, of the andGL thatsection of SIhad remained not begun, fairly the stable.percentage For 2002,of PI decreased the ratios continuously, of the three industry that of TI types increased were continuously, 17.3:35.3:47.4. Comparedand that of with SI thatremained in 1980, fairly the percentage stable. For of TI2002 increased, the ratios by 21.68%. of the Since three the industry GL section types opened were to trafﬁc17.3:35.3:47.4. in 2006, Compared the percentage with ofthat PI decreasedin 1980, the continuously, percentage of that TI ofincreased TI remained by 21.68%. fairly stable,Since the while GL thatsection of SI opened increased to traffic markedly, in 2006, and the it increased percentage by of 16.02% PI decreased in the period continuously, 2002–2012. that The of above TI remained results fairly stable, while that of SI increased markedly, and it increased by 16.02% in the period 2002–2012. indicate that the opening of the QTR probably helped to optimize the industrial structure of the TP, The above results indicate that the opening of the QTR probably helped to optimize the industrial which is consistent with the conclusion drawn by Wu [44]. In addition, as an important component of structure of the TP, which is consistent with the conclusion drawn by Wu [44]. In addition, as an TI, tourism boomed on the plateau ten years after the opening of the QTR to trafﬁc [45]. important component of TI, tourism boomed on the plateau ten years after the opening of the QTR to traffic [45].

4.5. Comparisons of the Urbanization Level Urbanization level (UL), and its changing trends, are important components of socioeconomic activity, and it refers to the proportion of non-agricultural population to total population. Changes in UL can reflect the process of migration of population to cities and urban areas. Non-agricultural population refers to those people engaged in SI and TI and the people they support. Figure 11 illustrates the UL of both Qinghai and Tibet at the county scale. Those counties surrounding the XG section of the QTR had a relatively high UL, whereas most other counties had a low UL for 2000, 2005, and 2010. In terms of changes in the spatial distribution of the UL, an increase can be seen in the surrounding counties of the QTR. Sustainability 2016, 8, 776 13 of 17

4.5. Comparisons of the Urbanization Level Urbanization level (UL), and its changing trends, are important components of socioeconomic activity, and it refers to the proportion of non-agricultural population to total population. Changes in UL can reﬂect the process of migration of population to cities and urban areas. Non-agricultural population refers to those people engaged in SI and TI and the people they support. Figure 11 illustrates the UL of both Qinghai and Tibet at the county scale. Those counties surrounding the XG section of the QTR had a relatively high UL, whereas most other counties had a low UL for 2000, 2005, and 2010. In terms of changes in the spatial distribution of the UL, an increase canSustainability be seen in 2016 the, surrounding8, 776 counties of the QTR. 13 of 17

0500250 0500250 0500250 km km km

Urbanization level (%) 2 4 6 8 10 12 14 16 18 20 30 50 90 FigureFigure 11. 11.Urbanization Urbanization levelslevels ofof QinghaiQinghai and and Tibet Tibet at at the the county county scale scale for for 2000, 2000, 2005, 2005, and and 2010. 2010.

ForFor comparison, comparison, the averagethe average UL valuesUL values of surrounding of surrounding and non-surrounding and non-surrounding areas ofareas the QTRof the for QTR 2000,for 2005, 2000, and 2005, 2010 and were 2010 calculated were calculated (Table5 ).(Table It can 5). be It seen can thatbe seen the ULthat values the UL of values surrounding of surrounding areas wereareas greater were than greater those than of non-surroundingthose of non-surrounding areas for areas the three for the time three points. time The points. average The ULaverage value UL of surroundingvalue of surrounding areas of the areas QTR of for the 2000, QTR 2005, for and2000, 2010 2005, was and 47.06%, 2010 and was that 47.06%, for non-surrounding and that for non- areassurrounding was 12.63%, areas a ratio was of12.63%, 3.73:1. a Theratio ULof 3.73:1. of surrounding The UL ofand surrounding non-surrounding and non-surrounding areas increased areas graduallyincreased over gradually the period over 2000–2010, the period but 2000–2010, the AGR of but non-surrounding the AGR of non-surrounding areas was greater areas than was that greater of surroundingthan that of areas surrounding in the periods areas in 2000–2005, the periods 2005–2010, 2000–2005, and 2005–2010, 2000–2010, and which 2000–2010, means which that the means QTR that hadthe no QTR obvious had impactno obvious on changes impact inon the changes UL of in its the surrounding UL of its surrounding areas in those areas time in periods. those time periods.

TableTable 5. Mean 5. Mean urbanization urbanization level level and and its AGRits AGR of surroundingof surrounding and and non-surrounding non-surrounding counties counties of of the the QTRQTR for 2000,for 2000, 2005, 2005, and and 2010. 2010. Mean Urbanization Level (%) AGR of Mean Urbanization Level (%) Region Mean Urbanization Level (%) AGR of Mean Urbanization Level (%) Region 2000 2005 2010 2000–2005 2005–2010 2000–2010 Surrounding areas 2000 42.74 2005 48.99 49.46 2010 2000–2005 2.77 2005–2010 0.19 2000–2010 1.47 SurroundingNon-surrounding areas areas 42.74 10.99 48.99 12.90 13.99 49.46 3.27 2.77 0.19 1.63 1.47 2.45 Non-surrounding areas 10.99 12.90 13.99 3.27 1.63 2.45 5. Discussion 5. DiscussionIn this study, the evolution of the transport system in the TP following the construction of the QTRIn this was study, analyzed, the evolution followed of theby transport a comparison system of in thefour TP factors following between the construction surrounding of the and QTR non- wassurrounding analyzed, followed areas before by a comparisonand after construction, of four factors and between finally the surrounding changes in and these non-surrounding factors under the areasinfluence before andof the after QTR construction, were examined. and finally However, the changes some uncertainties in these factors still underexist. the influence of the QTR wereFirst, examined. it should However, be noted somethat data uncertainties related to still the exist.QTR are difficult to acquire since both Qinghai andFirst, Tibet it should are remote be noted regions that dataand record related keeping to the QTR for arethese difficult regions to acquirelags behind since other both Qinghairegions of and mid- Tibeteast are China. remote For regions example, and because record keepingof a lack forof GDP these and regions population lags behind data otherat the regionstown scale, of mid-east they could China.only Forbe analyzed example, at because the county of a scale. lack ofHowever, GDP and the population land areas dataof counties at the townin both scale, Qinghai they and could Tibet onlyare be very analyzed large. atSome the countyof them scale. are even However, greater the in land size areas than ofprovinces counties in in mid-east both Qinghai China. and Therefore, Tibet arecomparison very large. Someat the ofcounty them scale are even is inadequate greater in sizeto reflect than provincesthe details in of mid-east the influence China. of Therefore, the QTR on comparisoneconomic atperformance. the county scaleAlthough is inadequate the 1 km GDP to reflect and population the details datasets of the influence were analyzed of the QTRfurther, on the economicuncertainties performance. in these Althoughdatasets might the 1 kmhave GDP extended and population to the results datasets of this were study analyzed since further,they were produced based on an algorithm devised for the national scale [38,39]. Second, changes in infrastructures and economic conditions in Tibet have attracted a large proportion of the floating population since the 1990s [46]. Some studies show that the floating population is becoming the major power for accelerating urbanization in modern times, and it is estimated that temporary residents may account for about half of the urban population in Tibet [47]. However, the population data obtained for this study did not contain the floating population, so the population density, the AGR of population, and the UL of the surrounding areas may therefore be underestimated. Other evidence also shows that Tibet has seen a surge of Han migrants, which has been further boosted by the QTR [4,48]. Once data with high precision and resolution become available, more objective conclusions can be drawn in future studies. Our results indicate that the influences of the QTR on the GDP of its surrounding areas are obvious, which is consistent with the results of Wang and Wu [25]. However, the influences of the Sustainability 2016, 8, 776 14 of 17 the uncertainties in these datasets might have extended to the results of this study since they were produced based on an algorithm devised for the national scale [38,39]. Second, changes in infrastructures and economic conditions in Tibet have attracted a large proportion of the floating population since the 1990s [46]. Some studies show that the floating population is becoming the major power for accelerating urbanization in modern times, and it is estimated that temporary residents may account for about half of the urban population in Tibet [47]. However, the population data obtained for this study did not contain the floating population, so the population density, the AGR of population, and the UL of the surrounding areas may therefore be underestimated. Other evidence also shows that Tibet has seen a surge of Han migrants, which has been further boosted by the QTR [4,48]. Once data with high precision and resolution become available, more objective conclusions can be drawn in future studies. Our results indicate that the influences of the QTR on the GDP of its surrounding areas are obvious, which is consistent with the results of Wang and Wu [25]. However, the influences of the QTR on population and UL of its surrounding areas are not so obvious. Some possible reasons are listed below. In fact, compared with the Qinghai–Tibet Highway, which had operated for 60 years, the traffic extension of the QTR was not so obvious (Figures4 and5). In addition, because the TP is a unique and fragile high-altitude ecosystem, the QTR construction project has raised serious concerns about its possible environmental consequences [15]. As a result, the Chinese government and local officials developed a green and sustainable policy that emphasized protection of soils, vegetation, animals, and water resources. In addition, five nature reserves along the route were established [4,5]. In other words, the construction of the QTR has been achieved with ecological protection at the top of the project’s agenda, which will certainly restrict the economic development of its surrounding areas. Moreover, the QTR is a young railway, and only ten years have passed since the entire line opened to traffic in 2006. Moreover, the construction of the supporting infrastructure of the QTR is incomplete and not comprehensive. Finally, compared with railways located in the mid-east region of China, the natural environmental conditions surrounding the QTR are very harsh and 965 km of the line runs at an altitude greater than 4000 m [49]. Therefore, it will take considerably more time for the QTR to influence the socioeconomic development of its surrounding areas. This comparison work is only an initial study into the influence of the QTR on the economic performance of the TP. In fact, as well as the QTR’s influence, the development of the region is tied to other factors, including financial support from central government [50]. Therefore, on the basis of this study, future analyses of the changes in accessibility to and economic linkages of the TP with surrounding regions will be more significant and overcome these uncertainties. With the passage of time, the transportation system and corresponding infrastructure of the TP are developing rapidly. For example, the Ministry of Transport of the People’s Republic of China has drawn up policies that the initial railway network of Tibet will be constructed and the length of the railway in Tibet will reach 1300 km by 2020 [51]. Recently, the construction of the Lhasa–Xigazê Railway was completed in July 2014 and was formally opened a month later [52]. As an extension of the QTR, designed with a capacity of more than 8.3 million tons of freight per annum, it will promote the economic development of Xigazê, as well as Tibet, by reducing transportation costs and promoting tourism. In addition, the Lanzhou–Xinjiang High-Speed Railway in northwestern China, from Lanzhou in Gansu Province to Ürümqi in the Xinjiang Uyghur Autonomous Region, was opened in December 2014 [53]. It is routed from Lanzhou to Xining before heading northwest into the Hexi Corridor at Zhangye, which makes it the first high-speed railway to go through the TP. In other words, a railway network will exist here in the near future. When the railway network era of the TP finally arrives, Tibet, Qinghai, Xinjiang, Gansu, Sichuan, and Yunnan provinces will be more closely connected by railway, which means that more relevant studies can be carried out. For example, the construction and development of the railway network in the TP is potentially beneficial to the Belt and Road Initiative—where “Belt” refers to the Silk Road Economic Belt and ”Road” refers to the 21st Century Maritime Silk Road [54]. Sustainability 2016, 8, 776 15 of 17

The railway network will boost the socioeconomic development of this region by moving people into it [5,46]. However, it must be mentioned that, from a sustainability perspective, the QTR and the coming railway network will need to be carefully managed to promote and ensure long-term ecosystem health and sustainability in western China. Only then can sustainable railway networks be followed in other regions.

6. Conclusions The major ﬁndings of this study are summarized as follows. (1) In the TP, the highway-dominated transport system will break up and an integrated and sustainable transport system will form with the construction of the QTR; (2) At the county scale, the AGR values of GDP of surrounding counties of the QTR were greater than those of non-surrounding counties for the periods 2000–2005, 2005–2010, and 2000–2010. At the 1 km scale, following the opening of the entire QTR line to trafﬁc, the GDP of surrounding areas had a greater growth rate than before; (3) Analysis at the county and 1 km scales indicated that the population was not aggregated into the surrounding areas of the QTR during the study period; (4) In terms of industrial structure, the proportion of PI decreased continuously, while the proportions of SI and TI increased overall between 1984 and 2012. The QTR had no obvious impact on UL changes in surrounding counties.

Acknowledgments: This work was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB03030500). We are very grateful to the two anonymous reviewers of this paper for providing constructive comments. Many thanks to Hua Zhang from Jiangxi Normal University and Basanta Paudel, Lanhui Li, and Zhongjun Hu from Institute of Geographic Sciences and Natural Resources Research for polishing the language. Author Contributions: Zhaofeng Wang, Yili Zhang and Shicheng Li designed the experiments; Shicheng Li performed the experiments and wrote the paper; Yukun Wang and Fenggui Liu analyzed the data; Zhaofeng Wang and Yili Zhang reviewed the manuscript. Conﬂicts of Interest: The authors declare no conﬂict of interest.

Abbreviations The following abbreviations are used in the main text: AGR Annual Growth Rate GDP Gross Domestic Production GL Golmud-Lhasa NBSC National Bureau of Statistics of China PI Primary industry QTR Qinghai-Tibet Railway SI Secondary industry TI Tertiary industry TP Tibetan Plateau UL Urbanization level XG Xining-Golmud

References

1. Brundtland Commission. Our Common Future; Oxford University Press: Oxford, UK, 1987. 2. ECMT. Assessment and Decision Making for Sustainable Transport. In Proceedings of the European Conference of Ministers of Transportation, Organization of Economic Coordination and Development, Paris, France, 7–8 October 2004; Available online: http://www.oecd.org (accessed on 8 July 2016). 3. Russo, F.; Comi, A. Urban Freight Transport Planning towards Green Goals: Synthetic Environmental Evidence from Tested Results. Sustainability 2016, 8, 381. [CrossRef] 4. Qiu, J. Environment: Riding on the roof of the world. Nature 2007, 449, 398–402. [CrossRef][PubMed] 5. Peng, C.; Ouyang, H.; Gao, Q.; Jiang, Y.; Zhang, F.; Li, J.; Yu, Q. Environment-Building a “green” railway in China. Science 2007, 316, 546–547. [CrossRef][PubMed] 6. Zhou, J.; Yang, J.; Gong, P. Constructing a green railway on the Tibet Plateau: Evaluating the effectiveness of mitigation measures. Transp. Res. Part D Transp. Environ. 2008, 13, 369–376. Sustainability 2016, 8, 776 16 of 17

7. Gutiérrez, J.; González, R.; Gómez, G. The European high-speed train network: Predicted effects on accessibility patterns. J. Transp. Geogr. 1996, 4, 227–238. [CrossRef] 8. Gutiérrez, J. Location, economic potential and daily accessibility: An analysis of the accessibility impact of the high-speed line Madrid–Barcelona–French border. J. Transp. Geogr. 2001, 9, 229–242. [CrossRef] 9. Wu, W.; Liang, Y.; Wu, D. Evaluating the Impact of China’s Rail Network Expansions on Local Accessibility: A Market Potential Approach. Sustainability 2016, 8, 512. [CrossRef] 10. Shaw, S.-L.; Fang, Z.; Lu, S.; Tao, R. Impacts of high speed rail on railroad network accessibility in China. J. Transp. Geogr. 2014, 40, 112–122. [CrossRef] 11. Ozbay, K.; Ozmen-Ertekin, D.; Berechman, J. Contribution of transportation investments to county output. Transp. Policy 2007, 14, 317–329. [CrossRef] 12. Zheng, S.; Kahn, M.E. China’s bullet trains facilitate market integration and mitigate the cost of megacity growth. Proc. Natl. Acad. Sci. 2013, 110, E1248–E1253. [CrossRef][PubMed] 13. Naranjo Gómez, J. Impacts on the Social Cohesion of Mainland Spain’s Future Motorway and High-Speed Rail Networks. Sustainability 2016, 8, 624. [CrossRef] 14. Wang, J.; Jin, F.; Mo, H.; Wang, F. Spatiotemporal evolution of China’s railway network in the 20th century: An accessibility approach. Transp. Res. Part A Policy Pract. 2009, 43, 765–778. [CrossRef] 15. Qin, Y.; Zheng, B. The Qinghai-Tibet Railway: A landmark project and its subsequent environmental challenges. Environ. Dev. Sustain. 2010, 12, 859–873. [CrossRef] 16. Timco, G. The Qinghai–Tibet Railroad. Cold Reg. Sci. Technol. 2009, 59, 1–2. [CrossRef] 17. Ding, M.; Zhang, Y.; Shen, Z.; Liu, L.; Zhang, W.; Wang, Z.; Bai, W.; Zheng, D. Land cover change along the Qinghai-Tibet Highway and Railway from 1981 to 2001. J. Geogr. Sci. 2006, 16, 387–395. [CrossRef] 18. Zhang, H.; Wang, Z.; Zhang, Y.; Ding, M.; Li, L. Identification of traffic-related metals and the effects of different environments on their enrichment in roadside soils along the Qinghai-Tibet highway. Sci. Total Environ. 2015, 521, 160–172. [CrossRef][PubMed] 19. Zhang, H.; Zhang, Y.; Wang, Z.; Ding, M. Heavy metal enrichment in the soil along the Delhi-Ulan section of the Qinghai-Tibet railway in China. Environ. Monit. Assess. 2013, 185, 5435–5447. [CrossRef][PubMed] 20. Sun, H.; Zheng, D.; Tong, Q.; Chen, Y.; Li, D. Recommendations on the Qinghai-Tibet railway construction and socioeconomic development of Tibet. Bull. Chin. Acad. Sci. 2004, 19, 247–249. (In Chinese) 21. Wang, L.; Fang, Y. Influence of the Qinghai-Tibet railway development on economic structure and economic layout along the line. Econ. Geogr. 2005, 25, 49–51. (In Chinese) 22. Jin, S. The impact of Qinghai-Tibet railway on economic development in Tibet. Macroecon. Res. 2006, 7, 47–52. (In Chinese) 23. Bai, G.; Li, W. The role of railway in promoting the economic development in the ethnic minority areas—A study of railways in Guangxi, QInghai, and Tibet. Guangxi Ethn. Stud. 2014, 7, 139–145. (In Chinese) 24. Su, M.M.; Wall, G. The Qinghai-Tibet railway and Tibetan tourism: Travelers’ perspectives. Tourism Manag. 2009, 30, 650–657. [CrossRef] 25. Wang, Y.; Wu, B. Can Transportation Investment Stimulate Local Economy? The Case of Qingzang Railway. China J. Econ. 2014, 1, 55–80. (In Chinese) 26. Yao, T.; Thompson, L.G.; Mosbrugger, V.; Zhang, F.; Ma, Y.; Luo, T.; Xu, B.; Yang, X.; Joswiak, D.R.; Wang, W.; et al. Third Pole Environment (TPE). Environ. Dev. 2012, 3, 52–64. [CrossRef] 27. Qiu, J. The third pole. Nature 2008, 454, 393–396. [CrossRef][PubMed] 28. Ma, W.; Cheng, G.; Wu, Q. Construction on permafrost foundations: Lessons learned from the Qinghai-Tibet railroad. Cold Reg. Sci. Technol. 2009, 59, 3–11. 29. Niu, F.; Lin, Z.; Lu, J.; Liu, H.; Xu, Z.Y. Characteristics of roadbed settlement in embankment–bridge transition section along the Qinghai-Tibet Railway in permafrost regions. Cold Reg. Sci. Technol. 2011, 65, 437–445. 30. Liu, G.D.; Zhang, C.X.; Hu, S.T.; Gang, W. Research on the Oxygen Supply System on Qinghai-Tibet Railway Train; Soc Air-Conditioning & Refrigerating Engineers Korea-Sarek: Seoul, Korea, 2006; pp. 193–196. 31. Zhu, J.; Liu, X.; Liang, Q. Study on the Carrying Capacity of Golmud-Lhasa Section of Qinghai-Tibet Railway. China Railw. Sci. 2007, 28, 128–135. 32. National Geomatics Center of China Spatial Dataset of China Administrative Areas at Scale of 1:4000000. Available online: http://nfgis.nsdi.gov.cn/ (accessed on 4 March 2014). 33. National bureau of statistics of China. Qinghai Statistical Yearbook; China Statistics Press: Beijing, China, 1985–2014. Sustainability 2016, 8, 776 17 of 17

34. National bureau of statistics of China. Tibet Statistical Yearbook; China Statistics Press: Beijing, China, 1985–2014. 35. Public Order Administration of Ministry of Public Security of China. Statistics of City and County Population of China (2000); China Statistics Press: Beijing, China, 2001. 36. Public Order Administration of Ministry of Public Security of China. Statistics of City and County Population of China (2005); China Statistics Press: Beijing, China, 2006. 37. Public Order Administration of Ministry of Public Security of China. Statistics of City and County Population of China (2010); China Statistics Press: Beijing, China, 2011. 38. Fu, J.; Jiang, D.; Huang, Y. 1 km grid population dataset of China (2005, 2010). Glob. Chang. Res. Data Publ. Repos. 2014.[CrossRef] 39. Huang, Y.; Jiang, D.; Fu, J. 1 km grid GDP dataset of China (2005, 2010). Glob. Chang. Res. Data Publ. Repos. 2014.[CrossRef] 40. Du, G. Official Figures: Tibetan Tourism Booming Before Riots. Available online: http://news.xinhuanet. com/english/2008-04/25/content_8051848.htm (accessed on 4 July 2016). 41. An, C. Petroleum Products Can Be Transported to Tibet by Railway, Which Endes the History of Only Depending on Highway to Transport. Available online: http://tibet.news.cn/gdbb/2014-05/05/c_ 133310397_2.htm (accessed on 4 July 2015). 42. Atack, J.; Bateman, F.; Haines, M.; Margo, R.A. Did Railroads Induce or Follow Economic Growth? Urbanization and Population Growth in the American Midwest, 1850–1860. Soc. Sci. Hist. 2010, 34, 171–197. [CrossRef] 43. Haines, M.R.; Margo, R.A. Railroads and Local Economic Development: The United States in the 1850s. 2006. Avaliable online: http://www.nber.org/papers/w12381 (accessed on 21 May 2015). 44. Wu, X. The opening to traffic of the Qinghai-Tibet Railway promoted the economic and social development of Tibet. Tibet Sci. Technol. 2007, 8, 14–16. (In Chinese) 45. Tian, S. Xinhua Insight: Tourism Booms on Plateau after 10 Years of Qinghai Tibet Railway. Available online: http://news.xinhuanet.com/english/2016-06/23/c_135461101.htm (accessed on 4 July 2016). 46. Liu, C.; Otsubo, K.; Wang, Q.; Ichinose, T.; Ishimura, S. Spatial and temporal changes of floating population in China between 1990 and 2000. Chin. Geogr. Sci. 2007, 17, 99–109. [CrossRef] 47. Fan, J.; Wang, H.; Chen, D.; Zhang, W.; Wang, C. Discussion on Sustainable Urbanization in Tibet. Chin. Geogr. Sci. 2010, 20, 258–268. [CrossRef] 48. Harris, T. Trading places: New economic geographies across Himalayan borderlands. Political Geogr. 2013, 35, 60–68. [CrossRef] 49. Zheng, D.; Zhang, Y. Ecological and Environmental Security along the Qinghai-Tibet Railway; Zhengjiang Science and Technology Publishing house: Hangzhou, China, 2009. (In Chinese) 50. Liu, G.; Shen, L. Characteristics and Mechanism of Tibet’s Industrial Structure Evolution from 1951 to 2004. Acta Geogr. Sin. 2007, 62, 364–376. (In Chinese) 51. Ministry of Transport of the People‘s Republic of China The advices of Ministry of Transport to Further Promote the Scientific Development of Tibet Transportation Effectively. Available online: http://www. moc.gov.cn/zhuzhan/jiaotongxinwen/xinwenredian/201407xinwen/201407/t20140724_1653868.html (accessed on 15 April 2015). 52. Liu, Y. The Railway from Lhasa to Shigatse Open to Traffic and It Will Provide New Possibilities for the Development in Tibet. Available online: http://news.xinhuanet.com/politics/2014-08/15/c_1112099527. htm (accessed on 1 July 2015). 53. Cui, J. Lanzhou-Xinjiang High-Speed Line Nears Completion. Available online: http://www.chinadaily. com.cn/china/2014-11/05/content_18872608.htm (accessed on 8 May 2016). 54. Liu, W. Scientific understanding of the Belt and Road Initiative of China and related research themes. Prog. Geogr. 2015, 34, 538–544. (In Chinese)

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).