ArticlePDF Available

Abstract and Figures

Scaling laws are powerful reflectors of the variations of the output of urban economic activities and the number of infrastructures with urban population. However, the difference in spatial definition of cities and data sources by countries leads to different statistical results of scaling law. We aim to analyse the difference in this paper by calculating regression coefficients of scaling law at different spatial scales, combined with census data, urban statistical yearbook data and remote sensing data of China. The conclusions are shown as follows: (1) Scaling coefficients change with both spatial scales and data sources. For spatial scales, scaling law is more agreeable with the data of urban municipal districts than with those of the whole city area in China. As there is a large number of non-urbanized areas within cities; these regions do not meet the assumptions of scaling law model. For data source, remote sensing data have a better fitting result than urban statistical yearbook data. (2) Comparatively speaking, urban population agglomeration contributes more to economic growth in China than it does in the US, but China has lower energy consumption and land- use efficiency. For example, the Gross Regional Product (GRP) scaling indicator of China is 1.22, while it is 1.11 in the United States. (3) Population agglomeration contributes more to the economic growth in large cities than in small cities. This may explain the emerging trends of urban immigrants in large cities of China. However, for energy consumption, small and medium- sized cities are more efficient than large cities. In addition, this paper discusses the potential direction for urban scaling research from three aspects: establishing more effective statistical units, combining traditional survey with big data analysis, and exploring mechanics behind scaling models.
Content may be subject to copyright.
地理学报
ACTA GEOGRAPHICA SINICA
72 卷 第2
2017 2
Vol.72, No.2
February, 2017
城市范围界定与标度律
董 磊
1
王 浩
2, 3
赵红蕊
2, 3
1. 清华大学建筑学院,北京 1000842. 清华大学土木工程系地球空间信息研究所,北京 100084
3. 清华大学 3S 中心,北京 100084
摘要:标度律作为城市发展的重要规律之一,反映了城市经济活动产值、基础设施数量等要素
随城市人口规模的变动情况,在城市研究领域引起了广泛讨论。但由于不同国家的城市统计
数据对应的空间范围各不相同,导致标度律系数受城市边界范围选取影响很大。本文通过比
较中、美两国统计数据对应的空间范围,并结合普查、城市统计年鉴和遥感数据,计算了不同空
间范围对应的标度律系数。结果表明:不同空间尺度和数据源得到的标度律系数有较大差
异。就空间尺度而言,市辖区比市域范围的数据更符合标度律模型,因为中国城市市域范围内
还存在大量的非城市化地区,并不符合标度律模型的适用条件;就数据源而言,遥感数据比城
市统计年鉴数据有更好的拟合优度;与美国城市相比,中国城市人口集聚带来的经济增长率
更高标度律系数更高)市辖区人口每增加一倍,经济规模可增加 122%这一数字在美国是
111%而在家庭能源消耗用水、用电和土地利用方面,中国城市的效率更低;从中国城市
内部对比来看,大城市与中小城市在经济规模、土地利用方面的标度律集聚效率明显不同,
人口集聚效应带来的大城市经济增长率、工资收入要远高于中小城市;能源消耗方面,中小城
市比大城市更有效率。最后,本文还从建立更加有效的统计单元、传统统计数据与大数据结
合、模型机制探索 3个方面阐述了城市标度律未来可能的研究方向。
关键词:城市范围;标度律;异速增长;集聚效率
DOI: 10.11821/dlxb201702003
1引言
近些年,得益于智能手机和移动互联网的发展,城市研究领域可获得的微观数据越
来越多;与此同时,城市研究学者从地理学、物理学、经济学等学科中借鉴了大量的模
型思路和方法,促进了城市研究向着模型化、定量化、数据化方向发展[1]。但数据的增多
让人容易陷入一个“误区”:轻信定量数据带来的结论,忽视了数据的产生过程与适用范
围。数据的使用与待研究的问题密不可分,同一个研究问题,采用不同的数据可能得出
相异的结论,即便使用相同的数据,但采用不同的处理方法,结果也可能大相径庭。这
一点在近些年广受关注的城市标度律 (scaling law) 研究中尤为值得注意。
标度律作为城市发展的重要规律之一,是指像道路网络、用地在内的城市属性在空
间的自相似分布 (在不同空间的时间尺度上重复一些模式,反映了城市内在的生长逻辑
和集聚规律[2]。城市标度律的研究借鉴了许多其他领域的成果,像语言学中词频幂率分布
收稿日期:2016-04-25; 修订日期:2016-10-18
基金项目:国家自然科学基金项目(41571414); 清华大学自主科研项目(2015THZ01) [Foundation: National Natural
Science Foundation of China, No.41571414; Tsinghua University Initiative Scientific Research Program,
No.2015THZ01]
作者简介:董磊(1988-), ,博士生,主要从事城市数据与空间分析研究。E-mail: arch.dongl@gmail.com
通讯作者:赵红蕊(1969-), ,教授,研究领域为定量遥感、遥感与 GIS 应用。E-mail: zhr@tsinghua.edu.cn
213-223
地理学报 72
的现象启发了对城市序位规模分布的探索[3-5];而生物体新陈代谢率与个体尺寸的异速增
长关系 (allometric growth) 则引导了对于城市异速增长的研究。生物领域的研究者发
现,将不同物种的新陈代谢频率与生物体体积画到对数坐标上,每个哺乳动物几乎都落
在一条斜率为 3/4 的直线上 (这一规律通常被称为 Kleiber law,这个关系式揭示了生物
能量消耗的规模效应,质量越大的组织,单位质量单位时间的能耗更低[6]。这一规律最早
Naroll [7]从生物学领域引入社会科学领域,后经由 Nordbeck[8]
Lee[9]
、陈彦光等[10- 11]
Bettencourt [12-13]多位学者从模型和实证的角度在城市研究领域将其发展完善,建立了更
为系统的城市标度律分析框架。如果把人口 N(t)作为时间 t上城市规模的度量,城市标度
律可以表示如下:
Y(t) = Y0N(t)β1
式中:Y表示物质资源 (例如基础设施数量) 或者社会活动的度量 (例如经济产出、污
染量等N(t)代表 t时刻的城市人口,用来衡量城市的规模;β作为指数反应了城市系
统中的标度律;Y0是一个标准化常量。根据β的大小,表征城市属性的变量可分为 3组:
β1,是亚线性关系 (sub-linear,通常是关于基础设施的参数,以人口增加导致的
规模经济为特征。例如尽管纽约有 4倍于休斯敦的人口,但纽约并不需要休斯敦 4倍数量
的加油站 (加油站与人口的幂律指数为 0.77β1,是线性关系,通常和工作、住
房、家庭用水量等个体需求相关,故与人口增加呈线性关系。β1,是超线性关系
super-linear,通常城市的社会经济属性 (也包括疾病、犯罪量等) 属于此类,表明相
应指标增长率要高于人口规模增长率,体现了城市的集聚效应。大量实证研究表明,标
度律是城市组织的一种普遍性质,从横跨不同国家 (美国、德国、瑞典、日本和中国
) 的数据来看,都符合此规律[10- 12]。图 1a 2010 年美国大都市统计区 (Metropolitan
Statistical Area, MSA) 范围内人口数量和地区生产总值 (Gross Regional Product, GRP
的双对数回归,图 1b 是人口与路网的双对数回归。这两组数据从数据的角度证明了经济
随人口增长呈超线性关系,而路网呈亚线性关系 (1
结合中国的数据,有许多关于城市标度律的实证研究。赖世刚等[14]分析了中国城市
人口分布的位序—规模法则,指出 1999-2009 年中国城市人口分布的幂次现象日渐加强,
城市趋于不均衡发展;李郇等[15]分析了 1990 年、2000 年、2005 3年的中国城市用地与
城镇人口之间的异速增长关系,认为 1990 年中国城市增长是负异速增长,2000 年和 2005
年是正异速增长;陈彦光[16]认为异速生长与分形和自组织网络理论相互融合,并以河南
为例,分析了城市人口和城市用地、城市产出等指标的标度关系,发现城市生长服从异
1美国大都市统计区人口与经济及道路总长的关系
Fig. 1 Scaling of urban socioeconomic and road length in the metropolises of USA
214
2 磊 等:城市范围界定与标度律
速生长规律[10-11]。还有许多学者针对不同地区的数据进行了相似的分析[17-20]
尽管这些研究都得到了相应的标度律系数,但使用的数据统计口径、回归模型各不
相同,得到系数也有较大差别。究其原因,很重要的一点在于城市标度律的提出基于一
个类比
—城市可以类比成生物体。其中,城市的人口规模相当于生物体的质量,城市
的经济活动相当于生物的新陈代谢。但城市与生物体的区别在于,城市并不存在一个清
晰的边界,因此,如何找到合理的空间范围对应生物体的个体 (body) 就成了一个很重
要的问题。特别地,城市边界选取的不同,分析结果可能迥异。例如在对于城市规模效
应的讨论中,Gleaser [21]认为大 城市对环境更友好 (体现在交通 CO2放量随人口数量
的增长呈亚线性关系,而 Louf [22]通过分析数据发现,只是改变一下 Gleaser 等用的城
市定义的边界,则会得出完全相反的结论。同样的数据集,如果在联合统计区
Combined Statistical Area, CSA) 的尺度下分析,系数小于 1,结论是大城市更节约能
源,而如果用城市区域 (Urbanized Area 当做分析单元,得出来的系数是 1.37,结论反
而是大城市不节能。随后又有一系列研究针对对英国[23-24]、法国[25]的城市进行讨论,分析
标度律的边界选取效应,发现边界选取对回归系数的影响很大。
这些争议的出现促使作者思考,不同规律或者模型适用的空间尺度是什么?尤其是
不同国家和地区的城市数据对应的空间范围各不相同,哪些数据和空间尺度更能反映城
市标度律?标度律在不同国家又有什么异同?本文希望通过分析不同空间范围和数据源
对应的标度律系数,并结合中美两国的对比研究,为此类问题提供新的视角。
2城市数据对应的空间范围
数据在城市研究中一个很重要的维度是其所对应的空间范围,不同的城市模型适用
的空间尺度范围不尽相同,因此在实证检验城市标度律模型时,非常有必要对比不同国
家和地区对于区域边界的界定及其对应的数据统计口径,以便在空间尺度上有可比性。
本文主要对比中国和美国的数据,原因有二:中美两国的城市数量、人口规模、国土
面积大致在一个量级,有可比性;现有的标度律研究主要集中在这两个国家,对这两
个国家数据口径和统计结果的比较,有助于在进行比较研究时找到模型共同的适用范围。
1实验观察到的城市尺度的指数分类
Tab. 1 The classification of scaling law indicator by empirical analysis
Y
基础设施
城市化地区
道路面积
加油站
人类个体需求
家庭用电量
家庭用电量
社会经济属性
GDP
工资
犯罪量
艾滋病新病例
专利数量
β
0.63
0.85
0.77
1.00
1.05
1.13
1.12
1.16
1.23
1.27
95%置信区间
[0.62, 0.64]
[0.81, 0.89]
[0.74, 0.81]
[0.94, 1.06]
[0.89, 1.22]
[1.11, 1.15]
[1.07, 1.17]
[1.11, 1.19]
[1.17, 1.29]
[1.22, 1.32]
观测量()
329
451
318
377
295
363
363
287
93
331
国家
美国
美国
美国
德国
中国
美国
美国
美国
美国
美国
城市单元
大都市统计区
大都市统计区
城市区域
-
地级市
大都市统计区
大都市统计区
大都市统计区
大都市统计区
大都市统计区
年份
1980-2000
2006
2001
2002
2002
2006
1969-2009
2003
2002-2003
1980-2001
注:数据来源于参考文献[7-8]
215
地理学报 72
2.1 美国统计数据对应的空间范围
美国对区域范围的定义大致分为两类:一类是以管理需要为基础划分的行政区边界
例如县、州,另一类是以统计需要为基础的统计区域 (Statistical Area,后者由美国
行政管理和预算局 (the United States Office of Management and Budget) 制定2013
全美范围内共划分了 1098 个统计区域,这些统计区域又可按区域人口规模和经济、交通
联系聚合成 541 个小都市统计区 (Micropolitan Statistical Area, µSA,也可译为都市圈
388 个大都市统计区 (Metropolitan Statistical Area, MSA) 和 169 个联合统计区
Combined Statistical Area, CSA。其中大都市统计区是指至少有一个人口规模 5万以
上的城市化地区 (Urbanized Area,加上周边与之有紧密经济联系的地理区域,联系强
度用通勤量来衡量;小都市统计区其他条件不变,只是将核心区的人口规模定义在1~5
万人之间;联合统计区则进一步根据经济和通勤联系将大都市统计区和小都市统计区聚
合而成。这里提到的城市化地区被美国人口普查局定义为人口密度不低于 1000 /平方英
里 (390 /km2
) 的连续的普查区组,并且任何与之相邻的普查区组的人口密度都不低于
500 /平方英里 (190 /km2
,城市化地区与行政边界无关。对于都市区、城市化地
区、城市 3个空间尺度的范围及对应的人口规模可见纽约—纽瓦克地区的例子 (2a
大小都市统计区共涵盖了美国 94%的人口 (大都市统计区涵盖 84%,小都市统计区
10%。常见的美国的区域数据,如人口、地区生产总值、失业率、建成区面积等多是以
大都市统计区为单元,因为这个统计单元更能反映区域的经济和发展情况。在这里需要
说明的是,针对美国的情况本文使用了 (统计) 区域而非城市一词,因为大小都市统计
区并不是城市与农村的划分。美国的 3142 个县中,1100 个在大都市统计区,688 个在小
都市统计区,1354 个县不在任何统计区
2.2 中国统计数据对应的空间范围
详细讨论中国城市的定义超出了本文的范畴,本文着重分析常见的中国区域统计数
据对应的空间范围,中国的区域统计数据是基于行政区划进行的。截至 2013 年底,全国
共有地级行政区划 333 个 (其中地级市 286 个、地区 14 个,自治州 30 个、盟 3,县级
行政区划 2853 个 (其中市辖区 872 个、县级市 368 个、县 117 个,另有自治县、旗等县级
行政单位。常见的人口普查数据给出了各区县尺度的人口数据,城市统计年鉴则列出
了地级市的统计数据,并进一步将地级市分成了市辖区范围和市域范围两个统计口径。
仅就数量来看,中国的区县一级数量与美国县的数量是大致相当的,地级市的数量
与大都市统计区比较类似,大量针对中国的研究使用的数据也都是基于地级市市域范
围。但与美国不同的是,在中国,即便是地级市范围内还有大量的非城市化地区,所以
真正与美国大都市统计区更有可比性的是地级市的市辖区范围。许多研究都认为,市辖
区或者是建成区范围而不是市域范围更适合对中国城市进行异速生长分析[15, 26]
2b 是与纽约—纽瓦克地区相对应的北京市空间范围示意图。北京市自内向外可分
Update of Statistical Area Definitions and Guidance on Their Uses, https://www.whitehouse.gov/sites/default/files/omb/
assets/bulletins/b10-02.pdf
通勤量的阈值被定为:居住在外围县但工作在中心县的就业人口占外围县总就业人口的 25%,或者外围县 25%
就业人口居住在中心县。
About Metropolitan and Micropolitan Statistical Areas, http://www.census.gov/population/metro/about/
Statistical Area Definitions and Guidance on Their Uses, https://www.whitehouse.gov/sites/default/files/omb/assets/
bulletins/b10-02.pdf
民政部社会服务发展统计公报:http://www.mca.gov.cn/article/zwgk/mzyw/201406/20140600654488.shtml
中国的建成区范围 (built-up area) 则类似于美国的城市化地区,只不过前者比后者面积要小。
216
2 磊 等:城市范围界定与标度律
为中心城区 (一般指东城、西城、朝阳、丰台、石景山、海淀、市辖区和市域范围,市
辖区包括除延庆、密云外的 16 个区 (2015 年北京市行政区划进行调整,延庆、密云也已
撤县设区。这 3个圈层的人口数量分别为 1172 万人、1883 万人、1961 万人 (2010 年人
口普查数据,其中,中心城区、市辖区人口数量大致分别相当于纽约市和纽约大都市区
的人口规模。
3中国城市的标度律
3.1 回归方程
标度律有一个重要特征
—系数的计算要在同一空间尺度范畴下进行,比如省是一
个尺度,地级市是一个尺度,而针对中国的特殊情况,地级市市辖区往往又是另一个尺
度。本文把中国地级市按市域和市辖区两个空间范围分别进行了回归,以比较不同空间
尺度的影响。同时考虑到城市规模的差别,在市辖区的回归中,城市被分为了全部城
市、大城市和中小城市 3组,其中大城市对应市辖区人口规模大于 100 万人,在本文采
用的数据集里共有 135 个,剩余为中小城市,共有 149 个 (地级市一共有 286 个,因丽江
市和东营市建成区数据缺失,本文实际使用的数据集包括 284 个地级市及其市辖区
对方程 (1) 取双对数后,得到回归检验方程:
2014 年 《关于调整城市规模划分标准的通知》一文,中国对大中小城市的划分标准是:人口<50 万的城市为
小城市;城区常住人口 50~100 万的城市为中等城市;城区常住人口 100~500 万的城市为大城市;城区常住人口 500~
1000 万的城市为特大城市;城区常住人口>1000 万的城市为超大城市。这里相当于把特大城市和超大城市都划归为
大城市分组。
2纽约—纽瓦克地区和北京市不同空间范围示意及对应的人口规模
Fig. 2 Different city boundaries and population of New York-Newark and Beijing
217
地理学报 72
log Y=βlog N+ log Y02
式中:对于 Y,本文分别测算了代表城市经济活动指标的 GRP、工资;代表城市用地指
标的建成区面积;代表能源消耗的家庭用水量、用电量;代表城市公共设施的中学数
量、医院床位数、公交车数量。N是城市人口;logY0是常数;β作为回归系数反映了城市
系统中各指标的标度律,回归采用最小二乘法。
3.2 数据来源
数据来源主要为:人口方面,常住人口比户籍人口更能反映城市的实际人口规
模。像深圳这种户籍人口只占城市常住人口不到 1/3 的城市,如果用城市统计年鉴中的户
籍人口数量,将会造成较大的偏误。本文使用了第六次人口普查数据,从中整理出了各
市市域及市辖区范围内的常住人口数。普查数据的好处在于准确性高,缺点在于人口普
10 年一次,数据更新慢,很难研究标度律的逐年变化,不过本文重点关注数据的空间
维度,这一缺点并不会对本文造成影响。区域经济数据来自2011 年 《中国城市统计年
》 和各市统计公报;建成区面积使用了两组数据,一组来自城市统计年鉴,另一组
来自世界银行 PUMA 数据库 (http://puma.worldbank.org/,后者是根据遥感影像测算的
城市建成区面积,计算方法参见[27],两组数据便于比较不同数据源对回归结果的影响。
其他数据,如家庭用水、用电量,学校数量等均来自 2011 年 《中国城市统计年鉴
3.3 回归结果
为了更好地理解数据的规律,把人口规模与 GRP 数据作出散点图 (3。通过图 3
可以发现:无论是散点图还是回归结果的 R2的数值都表明,以市辖区为空间范围的数
据拟合程度更好;因为异常值的存在,图 3a 中 (市域范围) 的拟合直线在人口规模较
大的城市偏离严重。通过将坐标从对数刻度转换为标准刻度 (3c3d) 更容易发现
—重庆作为异常值极大影响了回归结果。重庆市的特殊性在于,作为一个直辖市,它
的行政区划面积和人口数量居 4个直辖市之首,但其下辖多个城市化率很低的县市,所
以在市域尺度上,重庆就属于人口数量多,但 GRP 总量低的“异常值”,类似的“异常
值”在以市域为分析单元时还有很多。因此,一些对中国的标度律现象的分析会剔除异
常值,但这种剔除本身是值得商榷的,因为在使用数据验证模型的时候,首先要考虑的
是什么样的数据更符合模型的假设,而不是为了得到一个更好的回归结果而剔除异常
值。在本例中,市辖区范围无论是从与美国都市统计区的空间相似度还是回归结果,都
是比市域范围更好的数据集。
同时,通过对比城市统计年鉴中的建成区数据 (4a) 与遥感测算数据 (4b
发现,一些城市的年鉴数据可能严重低估了真实的建成区范围 (比如东莞市,中山市
。无论是从图的直观展示还是回归R2都表明,遥感测算的数据有更好的拟合优度。
在证明市辖区是分析标度律更好的空间范围后,本文将工资、家庭用水、用电量等
数据也在市辖区的范围内进行了测算。主要的回归结果如表 2所示。
3.4 回归结果分析
数据分析表明,尽管中国的城市体系、数据统计口径和国外不尽相同,但在各项统
计指标上都有比较相似的标度律。这一规律不像生物界有一个特定的指数,而是可被归
3类:β1super-linear,超线性β=1 linear,线性β1sub- linear,亚线
,每个大类中的β值比较接近。但除了普遍意义上的共性外 (GRP,工资等经济领
域的β1,表明相应指标的增长率要高于人口规模的增长率,中国城市与发达国家也有
第六次人口普查中对常住人口的定义:居住在本乡镇街道且户口在本乡镇街道或户口待定的人;居住在本乡镇街
道且离开户口登记地所在的乡镇街道半年以上的人;户口在本乡镇街道且外出不满半年或在境外工作学习的人。
218
2 磊 等:城市范围界定与标度律
许多差异,主要体现在以下几个方面:
1) 中国城市市辖区的 GRP 指数为 1.22,大于美国都市统计区的 1.11,就城市分组
来看,中国大城市的 GRP 指数为 1.36,远大于中小城市的 0.962,后者甚至没有体现出城
市在经济上的集聚效应
—中小城市的经济增长率反而低于人口规模增长率。大中小城
市的工资指数也反映了类似的规律,也就是说,中国大城市的经济产出效率>美国大都
市统计区>中小城市,这也可以解释为什么近些年大量的人口会涌入大城市。
2) 从家庭总用水、用电量的数据来看,中国大城市的家庭水、电消耗指数 (1.3
) >中小城市 (1=发达国家 (1左右。部分原因可能是,中国还处于发展阶
段,大小城市的发展很不平衡,大城市居民生活条件更好,因此城市越大,人均用水用
3市域及市辖区范围人口与地区生产总值的回归
Fig. 3 Regression between population and gross regional products
4市辖区范围人口与建成区面积的回归
Fig. 4 Regression between population and built-up areas
219
地理学报 72
电更多,而不是更少。
3) 从建成区面积指数来看,中国城市的土地利用效率低于发达国家 (中国城市统
计年鉴数据计算结果为0.89,遥感数据计算结果为 1.05,发达国家一般为 0.6~0.8 左右
也反映了一个中国城镇化过程中的现象:土地城镇化的速度与人口城镇化的速度不匹
配,前者高于后者,建设用地粗放低效。与用地类似,中国的公共设施配置 (中学数
量、医院床位数、公交数量) 也没有体现集聚效应 (学校数量除外,城市规模扩大,需
要投入的人均公共资源反而更多。
4) 中国城市的数据离散度更高,特别是中小城市 (大城市的离散度与美国比较接
。一个可能的解释是中国的中小城市正处于发展阶段,城市社会、经济形态并不成
熟。而大城市经过了较长时间的发展,社会、经济形态相对成熟,这一点在巴西、印度
等发展中国家的数据中也得到了印证[28]。此外,大城市的数据比中小城市相对更加准确
也是一个可能的原因。总之,中国的城市尽管有较高的经济集聚效率,但资源利用效率
较低。这一点在中国的大城市身上体现的尤为明显。
2不同空间范围对应的标度律统计结果
Tab. 2 The statistical results of scaling law within different spatial boundaries
因变量 Y
GRP
工资
家庭用水量
家庭用电量
建成区面积 1
建成区面积 2
中学数量
医院床位数
公交车数量
区域类型
地级市
市辖区
市辖区
市辖区
市辖区
市辖区
市辖区
市辖区
市辖区
市辖区
城市分组
全部城市
全部城市
大城市
中小城市
全部城市
大城市
中小城市
全部城市
大城市
中小城市
全部城市
大城市
中小城市
全部城市
大城市
中小城市
全部城市
大城市
中小城市
全部城市
大城市
中小城市
全部城市
大城市
中小城市
全部城市
大城市
中小城市
β
0.993
1.22
1.36
0.962
1.16
1.39
0.845
1.13
1.30
0.953
1.15
1.29
1.00
0.888
0.967
0.775
1.05
1.15
1.03
0.693
0.624
0.756
0.933
0.953
0.965
1.17
1.30
0.979
标准差
0.0492
0.0375
0.0613
0.126
0.0435
0.0778
0.131
0.0406
0.0584
0.147
0.0312
0.0472
0.112
0.0366
0.0601
0.123
0.0375
0.0549
0.128
0.0439
0.0741
0.148
0.0270
0.0467
0.0892
0.0471
0.0812
0.154
R2
0.560
0.790
0.785
0.279
0.717
0.703
0.214
0.733
0.787
0.216
0.830
0.847
0.352
0.678
0.663
0.207
0.752
0.764
0.299
0.466
0.343
0.145
0.808
0.756
0.440
0.687
0.655
0.210
观测量()
284
284
135
149
284
135
149
284
135
149
284
135
149
284
135
149
284
135
149
284
135
149
284
135
149
284
135
149
注:建成区面积:1代表城市统计年鉴数据;2代表世界银行 PUMA 数据库数据。
220
2 磊 等:城市范围界定与标度律
4结论与讨论
4.1 结论
本文比较了中美两国统计数据对应的空间范围,并利用普查数据、城市统计年鉴数
据和遥感数据,在地级市市域和市辖区两个空间尺度上测算了中国城市的标度律。结果
表明,中国城市市辖区范围的数据比市域范围更符合标度律。因为标度律模型侧重于刻
画城市实际的经济、社会活动。像许多美国城市标度律的研究多采用大都市统计区为边
界,而大都市统计区本身就是根据人口规模和经济、社会联系划定的,这比较有利于消
除地理和行政边界造成的影响。而在中国,地级市市域范围内还存在大量的非城市化地
区,所以相对来说市辖区数据更为适合,因为这部分区域城市化率普遍较高,第二、第
三产业占比重较高,其数据更能反映城市特征。
尽管中国市辖区的数据整体上符合标度律,但在数据离散度、社会经济类指标、家
庭消费指标等多方面与发达国家有所不同,而且中国大城市和中小城市也有较多差异。
与美国城市相比,中国城市人口集聚带来的经济增长率更高,但能源和土地利用效率更
低。从中国城市内部对比来看,人口集聚带来的大城市经济增长率要远高于中小城市,
但在能耗方面,中小城市比大城市更有效率。
4.2 讨论
尽管本文在现有数据的基础上分析了空间范围的界定与标度律关系,但这个问题仍
然没有从根本上得到解决,未来有 3个重要的方向可以在后续工作中完善:可以建立
更加有效的统计单元 (比如经济统计区,以获取更为精细 、准确的数据。虽然早在 20
世纪 80 年代,有学者就曾提出过在中国建立城市经济统计区的设想,但在实际中并未实
[29-30],有效的统计单元和统计数据不仅有利于政府政策的制定和实施,也是学者们进行
包括城市地理学在内的各项学术研究的基础。标度律与空间尺度的关系涉及到地理学
领域的 MAUP 问题 (Modifiable Areal Unit ProblemMAUP 是指不同的空间尺度会对统
计回归结果造成偏误[31]。研究者拿到的数据往往都是聚合后的数据,而不同的聚合方式
可能会对分析造成影响[32-33]。在数据更加精细的今天,以后的城市数据分析,可以用大数
据与细粒度的普查数据结合,在多尺度下分别验证分析结果。比如,现在许多研究都采
100 m 或者是 1 km 网格,但不同的网格下对应的结果是否稳健?需要更进一步检验。
亦或是标度律随网格尺度变化本身也会有统计规律呈现。城市标度律背后的机制是什
么还有待进一步探索,更重要的是从机制和模型上解释标度律的形成与背后的意义,数
据并不能代替机制与模型。
周一星[34]的研究中认为:城市研究的第一科学问题是基本概念的正确性,这在数据
时代尤为重要。特别是现在对于中国城市研究的很多数量方法与模型研究还是基于欧美
的研究体系,而中国不像欧美
—传统的调查统计非常完备,概念的界定也比较清晰,
中国的统计数据存在来源多样、口径不一等诸多问题,所以在研究中国城市问题时,在
选择研究数据和处理方法上需要特别小心。
参考文献(References)
[ 1 ] Batty M. The New Science of Cities. Cambridge: MIT Press, 2013.
[ 2 ] Batty M. The size, scale, and shape of cities. Science, 2008, 319(5864): 769-771.
[ 3 ] Zipf G K. Human Behavior and the Principle of Least Effort. Cambridge: Addison-Wesley Press1949.
[ 4 ] Gabaix X. Zipf's law and the growth of cities. The American Economic Review, 1999, 89(2): 129-132.
[ 5 ] Li Xiaojian, Xu Jiawei, Hai Beibei. The changing distribution patterns of rural settlements during the process of
urbanization: The case of Gongyi (1929-2013), China. Acta Geographica Sinica, 2015, 70(12): 1870-1883. [李小健,
221
地理学报 72
家伟,海贝贝.县域聚落分布格局演变分析:基于 1929-2013 年河南巩义的实证研究.地理学报, 2015, 70(12): 1870-
1883.]
[ 6 ] Enquist B, Brown J, West G. Allometric scaling of plant energetics and population density. Nature, 1998, 395(6698):
163-165.
[ 7 ] Naroll R S, Bertalanffy L. The principle of allometry in biology and social sciences. General Systems Yearbook, 1956, 1
(2): 76-89.
[ 8 ] Nordbeck S. Urban allometric growth. Geografiska Annaler Series B, Human Geography, 1971, 53(1): 54-67.
[ 9 ] Lee Y. An allometric analysis of the US urban system. Environment and Planning A, 1989, 21(4): 463-476.
[10] Chen Yanguang, Liu Minghua. Study on fractal dimension of size distribution of cities. Bulletin of Science and
Technology, 1998, 14(6): 395-400. [陈彦光,刘明华.区域城市规模分布的分维研究.科技通报, 1998, 14(6): 395-
400.]
[11] Chen Yanguang, Zhou Yixing. A study of fractional dimension on the relationship between population sizes and
economic outputs of cities. Economic Geography, 2003, 23(4): 476-481. [陈彦光,周一星.城市规模—产出关系的分
形性质与分维特征.经济地理, 2003, 23(4): 476-481.]
[12] Bettencourt L, Lobo J, Helbing D, et al. Growth, innovation, scaling, and the pace of life in cities. Proceedings of the
National Academy of Sciences, 2007, 104(17): 7301-7306.
[13] Bettencourt L. The origins of scaling in cities. Science, 2013, 340(6139): 1438-1441.
[14] Lai Shih- Kung, Han Haoying, Fang Ming. Power law distribution of urban population in China. Urbanization and Land
Use, 2014, 2(1): 1-7. [赖世刚,韩昊英,方明.中国城市人口分布的幂次现象.城镇化与集约用地, 2014, 2(1): 1-7.]
[15] Li Xun, Chen Gangqiang, Xu Xueqiang. Urban allometric growth in China: Theory and facts. Acta Geographica Sinica,
2009, 64(4): 399-407. [李郇,陈刚强,许学强.中国城市异速增长分析.地理学报, 2009, 64(4): 399-407.]
[16] Chen Yanguang. The rise, fall, and revival process of allometric scaling analysis in urban studies. Geographical
Research, 2013, 32(6): 1033-1045. [陈彦光.城市异速标度研究的起源、困境和复兴.地理研究, 2013, 32(6): 1033-
1045.]
[17] Kuang Wenhui, Zhang Shuwen, Zhang Yangzhen, et al. Analysis of urban land utilization spatial expansion mechanism
in Chuangchun City since 1900. Acta Geographica Sinica, 2005, 60(5): 841-850. [匡文慧,张树文,张养贞,. 1900
以来长春市土地利用空间扩张机理分析.地理学报, 2005, 60(5): 841-850.]
[18] Zhang Li, Lei Jun, Li Xuemei, et al. The features and influencing factors of urban expansion in China during 1997-
2007. Progress in Geography, 2011, 30(5): 607-614. [张利,雷军,李雪梅,. 1997-2007 年中国城市用地扩张特征及
其影响因素分析.地理科学进展, 2011, 30(5): 607-614.]
[19] Zhao Ke, Zhang Anlu. Driving forces behind urban land expansion based on panel data for 14 cities in Liaoning.
Resources Science, 2013, 35(5): 928-934. [赵可,张安录.城市建设用地扩张驱动力实证研究:基于辽宁省 14 市市辖
区数据.资源科学, 2013, 35(5): 928-934.]
[20] Fu Jianchun, Li Gang, Zhao Hua, et al. Relation between the population and the land area of urban built- up area in
China: An empirical research of 652 counties. China Land Sciences, 2015, 29(2): 46-53. [傅建春,李钢,赵华,.中国
城市人口与建成区土地面积异速生长关系分析:基于 652 个设市城市的实证研究.中国土地科学, 2015, 29(2): 46-
53.]
[21] Glaeser E, Kahn M. The greenness of cities: Carbon dioxide emissions and urban development. Journal of Urban
Economics, 2010, 67(3): 404-418.
[22] Louf R, Barthelemy M. Scaling: Lost in the smog. Environment and Planning B, 2014, 41(5): 767-769.
[23] Masucci A P, Arcaute E, Hatna E, et al. On the problem of boundaries and scaling for urban street networks. Journal of
The Royal Society Interface, 2015, 12(111): 20150763.
[24] Arcaute E, Hatna E, Ferguson P, et al. Constructing cities, deconstructing scaling laws. Journal of The Royal Society
Interface, 2015, 12(102): 20140745.
[25] Cottineau C, Hatna E, Arcaute E, et al. Paradoxical interpretations of urban scaling laws. arXiv preprint arXiv:
1507.07878, 2015.
[26] Chen Y. Characterizing growth and form of fractal cities with allometric scaling exponents. Discrete Dynamics in
Nature and Society, 2010, Article ID 194715, 22 pages.
[27] Schneider A, Mertes C M, Tatem A J, et al. A new urban landscape in East-Southeast Asia, 2000-2010. Environmental
Research Letters, 2015, 10(3): 034002.
[28] Bettencourt L, West G. A unified theory of urban living. Nature, 2010, 467(7318): 912-913.
[29] Zhou Yixing. Defining the conception of city and city population in China. China Planning Review, 1986(3): 10-15. [
222
2 磊 等:城市范围界定与标度律
一星.关于明确我国城镇概念和城镇人口统计口径的建议.城市规划, 1986(3): 10-15.]
[30] Wang Guoxia, Cai Jianming. Research on the method of quantifying the spatial boundary of metropolitan regions.
Economic Geography, 2008, 28(2): 191-195, 249. [王国霞,蔡建明.都市区空间范围的划分方法.经济地理, 2008, 28
(2): 191-195, 249.]
[31] Chen Yanguang. Simplicity, complexity, and mathematical modeling of geographical distributions. Progress in
Geography, 2015, 34(3): 321-329. [陈彦光.简单、复杂与地理分布模型的选择.地理科学进展, 2015, 34(3): 321-329.]
[32] Chen Jiangping, Zhang Yao, Yu Yuanjian. Effect of MAUP in spatial autocorrelation. Acta Geographia Sinica, 2011, 66
(12): 1597-1606. [陈江平,张瑶,余远剑.空间自相关的可塑性面积单元问题效应.地理学报, 201166(12): 1597-
1606.]
[33] Qi Lili, Bo Yanchen. MAUP effects on the detection of spatial hot spots in socio- economic statistical data. Acta
Geographica Sinica, 2012, 67(10): 1317-1326. [齐丽丽,柏延臣.社会经济统计数据热点探测的MAUP 效应.地理学
, 2012, 67(10): 1317-1326.]
[34] Zhou Yixing. The primary scientific issue of urban research in China is the correctness of basic urban concepts. Urban
Planning Forum, 2006, 1(1): 1-5. [周一星.城市研究的第一科学问题是基本概念的正确性.城市规划学刊, 2006, 1
(1): 1-5.]
The definition of city boundary and scaling law
DONG Lei1, WANG Hao2,3, ZHAO Hongrui2, 3
(1. School of Architecture, Tsinghua University, Beijing 100084, China; 2. Institute of Geomatics,
Department of Civil Engineering, Tsinghua University, Beijing 100084, China;
3. 3S Center, Tsinghua University, Beijing 100084, China)
Abstract: Scaling laws are powerful reflectors of the variations of the output of urban
economic activities and the number of infrastructures with urban population. However, the
difference in spatial definition of cities and data sources by countries leads to different
statistical results of scaling law. We aim to analyse the difference in this paper by calculating
regression coefficients of scaling law at different spatial scales, combined with census data,
urban statistical yearbook data and remote sensing data of China. The conclusions are shown as
follows: (1) Scaling coefficients change with both spatial scales and data sources. For spatial
scales, scaling law is more agreeable with the data of urban municipal districts than with those
of the whole city area in China. As there is a large number of non-urbanized areas within cities;
these regions do not meet the assumptions of scaling law model. For data source, remote
sensing data have a better fitting result than urban statistical yearbook data. (2) Comparatively
speaking, urban population agglomeration contributes more to economic growth in China than
it does in the US, but China has lower energy consumption and land- use efficiency. For
example, the Gross Regional Product (GRP) scaling indicator of China is 1.22, while it is 1.11
in the United States. (3) Population agglomeration contributes more to the economic growth in
large cities than in small cities. This may explain the emerging trends of urban immigrants in
large cities of China. However, for energy consumption, small and medium- sized cities are
more efficient than large cities. In addition, this paper discusses the potential direction for
urban scaling research from three aspects: establishing more effective statistical units,
combining traditional survey with big data analysis, and exploring mechanics behind scaling
models.
Keywords: city boundary; scaling law; allometric growth; agglomeration effect
223
... Previous studies have revealed that the urban scaling law is more applicable to m nicipal districts than the entire spatial scope of a city [29]. Therefore, the municipal d tricts of 293 prefectural cities serve as the basic research unit in this paper, and the urb land areas extracted from satellite images are used to characterize the scale of urban lan The urban property data used in this paper, including urbanization rate, GDP, and p manent resident population, are obtained from the China Statistical Yearbook of Urb Construction (2000-2020), the Shenzhen Statistical Yearbook (2021), and land cover d taset provided by Yang and Huang [30]. ...
... Previous studies have revealed that the urban scaling law is more applicable to municipal districts than the entire spatial scope of a city [29]. Therefore, the municipal districts of 293 prefectural cities serve as the basic research unit in this paper, and the urban land areas extracted from satellite images are used to characterize the scale of urban land. ...
Article
Full-text available
Rapid urban expansion caused by vigorous urban population growth brought up various socioeconomic and eco-environmental problems, which have important ramifications for sustainable development across the world. Along with the accelerated urbanization process, accurate and realistic prediction of urban expansion is of great importance to optimize urban planning and urban development. This study proposed a new hybrid model, which combined the urban scaling law (USL) with the ANN-CA model to predict urban expansion. To employ urban scaling law in the model, we innovatively calculated the law exponent at a single-city scale. Based on USL, we estimated urban land demand in the future by panel data regression. Finally, we added the area constraint and ecological constraint into the ANN-CA model to simulate urban expansion spatially. This frame of urban expansion has been successfully applied in Shenzhen, of which the urban land area would increase from 816.45 km2 in 2020 to 842.48 km2 in 2025. By comparing this model with the traditional prediction method, we proved its effectiveness and accuracy. Besides, we found that the scaling exponent can reflect urbanization level and distinguish overconstructed cities.
... unbalanced growth of two subsystems within a city [8,9], such as urban population-rural population [10], urban scale-economic output [11], urban scale-infrastructure density, industrial output value-land area, and population size-medical resources [12]. The existing literature uses the coupling coordination model [13,14] and the decoupling model [15] to describe the relationship between UL and the UP. ...
Article
Full-text available
Maintaining a balance between urban land (UL) expansion and urban population (UP) growth is one of the goals of sustainable development, and maintaining this balance requires more theoretical exploration and regional experience. This paper re-evaluated the imbalances in growth between urban land and urban population (IGULUP) from the perspective of allometric growth and explored its influencing mechanism, taking urban agglomerations (UAs) in China as a case. This paper reveals that the growth rate of UL in China is slightly higher than that of the UP. However, the IGULUP vary according to development stages. UAs in the primary stage and the early growth stage face the dilemma where UL grows faster than the UP. Conversely, for UAs in the later growth stage and the mature stage, the growth rate of the UP is higher than that of UL. Finally, an increase in economic development level, population agglomeration, fiscal expenditure, and urban compactness can help mitigate the gap between UL and UP. In contrast, industrial structure, urbanization level, and foreign direct investment may hinder the improvement of IGULUP by accelerating the rate of land expansion. These findings may make theoretical contributions to the formulation of more targeted land use control policies and urban population growth strategies.
... This suggests that tier-2 and tier-3 cities in China may still be in the developing stage, as their urban social and economic patterns are not yet mature. In contrast, tier-1 cities exhibit a maturity of socioeconomic development, given that the system has reached economies of scale Dong et al. 2017). ...
Article
High-speed rail (HSR) has become a significant mode of intercity travel due to its numerous socioeconomic and environmental benefits. Although some countries, such as China and Spain, have developed comprehensive HSR networks, many other nations are also investing or interested in developing their HSR infrastructure. However, concerns are arising about the high cost of HSR development and operation and whether the massive investment in HSR is justified. Furthermore, it remains unclear what HSR scale is appropriate to maintain travel demand and financial sustainability , considering the socioeconomic heterogeneity among different regions. To address this research gap, this study introduces a novel analytical framework that examines the relations between the scales of HSR and socioeconomic and demographic characteristics, including the size of the urban population, using 294 Chinese cities as an example. By applying the machine learning method based on the theory of urban scaling law, we find that the existing scale of HSR varies considerably among different tiers of cities. Although some large cities have underutilized the scale of the HSR systems in terms of service frequencies, platforms and rail lines, others have overutilized it, especially in terms of the building area of HSR stations in Tier-1 cities. In addition, by comparing the predicted scale changes of HSR in 2035 and 2050 with the actual system planning documents, we confirm that while some cities' (i.e., Shanghai and Shenzhen) predicted outcomes of system's scale are consistent with that of the actual civic HSR development plans, inconsistencies are also found in other cities, such as Yangzhou and Wuhan. Overall, the study provides valuable empirical evidence and implications for policymakers to improve decision-making on the future HSR scale and guidance for future development.
... In China, urban districts are synonyms for urban economy, which include highly urbanized areas in the urban administrative area. Compared with urban administrative regions, they are closer to physical cities (Dong et al., 2017). Therefore, the changes of the urban population over a long period of time can partially reflect the growth and shrinkage trends of cities. ...
Article
Urban shrinkage is a global phenomenon, and it will coexist with urban growth for many years. At the same time, the network connection between cities continuously improved due to the construction of the transportation and information networks. However, the relationship between urban network externalities and urban population growth/shrinkage remains unclear. Therefore, based on high-speed railway (HSR) flow data, a spatial econometric model is used to explore the mechanism behind urban population growth and shrinkage from the perspective of network externalities in China. The results indicate that: 1) The urban network experiences a certain clubbing effect. Growing cities that are strongly connected are concentrated along China's main railway lines and the southeastern coastal areas, while shrinking cities that are weakly connected are distributed at the periphery of the network. 2) Moreover, The network externality disregards spatial distance and together with the agglomeration externality influences the growth and shrinking of cities. 3) Urban economic development still promotes the development of Chinese cities. However, the improvement of the urban economy has a negative cross-regional spillover effect on neighboring cities due to urban competition. 4) Lastly, Local spillovers of urban network externalities are positive, while cross-regional ones are negative. Consequently, the government needs to promote the construction of multi-dimensional network connections between cities to promote cities’ sustainable development. This study reveals the relationship between urban network externalities and urban development, enriches the theories of network externalities and urban growth/shrinkage, and provide a reference for regional coordinated development.
... 式中:Y 为城市指标 (如县级单元截至某日 COVID-19 累计确诊/死亡病例) ;N 为城市人 口规模;Y0和β为参数,其中β为标度指数。城市标度律分析需要关注城市范围的界定, 进而确定城市人口规模和城市指标数值 [46][47] 。由于美国城市化率超过 80%,本文采用县总 人口表征城市人口规模。 对 (1) 式两边同时取以 10 为底对数,可以得到: lgY = β × lg N + lg Y 0 (2) 式 (2) 为线性函数。对确诊/死亡病例和人口规模取对数后,采用线性函数拟合, 拟合直线的斜率为标度指数 (β) ,分别记为β确诊和β死亡。虽然有其他非线性拟合方法 [48][49] , 但线性拟合因其简单易操作在标度律拟合中被广泛采用 [50][51] Abstract: Urban scaling law quantifies the disproportional growth of urban indicators with urban population size, which is one of the simple rules behind the complex urban system. Infectious diseases are closely related to social interactions that intensify in large cities, resulting in a faster speed of transmission in large cities. ...
Article
Full-text available
Urban scaling law quantifies the disproportional growth of urban indicators with urban population size, which is one of the simple rules behind the complex urban system. Infectious diseases are closely related to social interactions that intensify in large cities, resulting in a faster speed of transmission in large cities. However, how this scaling relationship varies in an evolving pandemic is rarely investigated and remains unclear. Here, taking the COVID-19 epidemic in the United States as an example, we collected daily added cases and deaths from January 2020 to June 2022 in more than three thousand counties to explore the scaling law of COVID-19 cases and city size and its evolution over time. Results show that COVID-19 cases super-linearly scaled with population size, which means cases increased faster than population size from a small city to a large city, resulting in a higher morbidity rate of COVID-19 in large cities. Temporally, the scaling exponent that reflects the scaling relationship stabilized at around 1.25 after a fast increase from less than one. The scaling exponent gradually decreased until it was close to one. In comparison, deaths caused by the epidemic did not show a super-linear scaling relationship with population size, which revealed that the fatality rate of COVID-19 in large cities was not higher than that in small or medium-sized cities. The scaling exponent of COVID-19 deaths shared a similar trend with that of COVID-19 cases but with a lag in time. We further estimated scaling exponents in each wave of the epidemic, respectively, which experienced the common evolution process of first rising, then stabilizing, and then decreasing. We also analyzed the evolution of scaling exponents over time from regional and provincial perspectives. The northeast, where New York State is located, had the highest scaling exponent, and the scaling exponent of COVID-19 deaths was higher than that of COVID-19 cases, which indicates that large cities in this region were more prominently affected by the epidemic. This study reveals the size effect of infectious diseases based on the urban scaling law, and the evolution process of scaling exponents over time also promotes the understanding of the urban scaling law. The mechanism behind temporal variations of scaling exponents is worthy of further exploration.
... Bettencourt et al. [8] were the first to discuss the concept of the scaling law in urban systems. Since then, a number of studies have focused on the importance of urban population size in urban systems, focusing mainly on the concept of urban scaling law [7,9,10], causes and mechanism of formation [11][12][13][14][15], verification [16][17][18][19][20][21], application [22][23][24][25], and questioning [26][27][28][29][30]. The abovementioned research was conducted at the municipal scale. ...
Article
Full-text available
The law of urban scaling implies that there is a universally applicable nonlinear scaling relationship between population size and urban indicators, which is a method of quantitative analysis that can reflect the growth law and internal logic of the urban system. However, most present research is conducted at the municipal scale, and studies of scaling law in the inner-city system are scarce, especially from the perspective of compact urban form development. The goal of this paper is to discover the scaling law within urban systems from a multi-scale perspective. Through the empirical analysis of Wuhan, this paper examines the internal scale law of the urban system from the municipal and district scales. Moreover, we use the landscape expansion index to perform spatial autocorrelation analysis. In this way, we assess the relationship between the compactness of urban morphological development and the urban scaling law. The results indicate that the temporal scaling law on the city scale has a more significant linear law than the single-year scaling law. The analysis also shows the scaling law relationship within the inner-city system. Nevertheless, there is a deviation between the temporal scaling law and the cross-section scaling law. Namely, the time series development of a district does not follow the section scaling law of the urban system. Furthermore, the urban scaling law shows a negative correlation with the compactness of the urban form development. It is crucial to understand the current economic development and resource endowment of an urban system in the urbanization process, as it significantly contributes to urban development and regional coordinated planning.
Article
Full-text available
Utilizing location data from USPTO patents spanning 1790 to 2022, this study incorporates theories from complexity science, lifecycle analysis, and geographical nature. It employs methods including urban scaling laws, scale- adjusted metropolitan indicators, and geographic detectors to construct a performance index for United States science and technology innovation (STI) centers, and analyze their spatiotemporal pattern evolution and driving mechanisms. The results reveal that: (1) Innovation scaling laws across cities have evolved into significant super- linear relationships, indicating that larger cities have higher per capita innovation outputs, thus demonstrating a pronounced scale-increasing effect of STI. (2) The performance index facilitates cross-scale comparisons of urban STI capabilities, illustrating the complete lifecycle of STI centers from incubation, expansion, decline, to revival. Throughout the study period, San Francisco CSA progressively surpassed New York CSA, Boston CSA, and Chicago CSA, becoming a leading STI center in the United States. (3) The spatial distribution of the performance index for US STI centers reveals a "north- superior, south inferior" pattern. Factors of second nature, such as urbanization, internationalization, and socio economic development, along with urban polycentricity, positively influence the performance index. In contrast, population density and first nature factors like climate and topography negatively correlate with the performance index. Under the influence of path dependence and cumulative causation, the performance of US STI centers exhibits a "Matthew effect," exacerbating the north-south divide.
Article
The cities of China, the world’s largest CO2 emitting country, play a crucial role in mitigating global climate change by their pursuit of carbon neutrality. However, assessing their decarbonization levels with internationally recognized metrics such as the OECD cities’ criteria is a major challenge, as the fuzzy definitions of China’s cities – including by city administrative boundary (UB1), city district boundary (UB2), and urban built-up area (UB3) – create misconceptions and ambiguities over the integrity and accuracy of their reported emissions. In this study, we develop a new China high-resolution emission database (CHRED) as a means of quantifying urban emissions based on dynamic city limits defined by population density boundary (UB4). Employing a 1 km grid dataset built from point-emission sources, it is showed that: (1) the reported emissions from the adoption of different urban boundaries can have differences as large as 17.77 %; (2) the spatial structure of China’s urban CO2 emissions has a clustered pattern reflecting the country’s spatial urban agglomeration structure; (3) the CO2 emissions per capita of China mega UB4 cities are similar to OECD C40 cities; and (4) that nonindustrial emissions are influenced by geographical location and population density. We further argue that the boundaries of UB4 cities need to be continuously adjusted to reflect the growth of urban built-up land across China. This is further evidenced by a counterfactual estimation and scenario analysis, which indicates that the differences in per capita nonindustrial emissions under different scenarios of population growth and urban sprawl could be as large as 27.6 % by 2040. The findings show that a bottom-up accounting database based on the UB4 boundary is significant for building high-resolution urban emissions inventories to mitigate nonindustrial carbon emissions in China.
Chapter
Full-text available
Fractal growth is a kind of allometric growth, and the allometric scaling exponents can be employed to describe growing fractal phenomena such as cities. The spatial features of the regular fractals can be characterized by fractal dimension. However, for the real systems with statistical fractality, it is incomplete to measure the structure of scaling invariance only by fractal dimension. Sometimes, we need to know the ratio of different dimensions rather than the fractal dimensions themselves. A fractal-dimension ratio can make an allometric scaling exponent (ASE). As compared with fractal dimension, ASEs have three advantages. First, the values of ASEs are easy to be estimated in practice; second, ASEs can reflect the dynamical characters of system's evolution; third, the analysis of ASEs can be made through prefractal structure with limited scale. Therefore, the ASEs based on fractal dimensions are more functional than fractal dimensions for real fractal systems. In this paper, the definition and calculation method of ASEs are illustrated by starting from mathematical fractals, and then, China's cities are taken as examples to show how to apply ASEs to depicting growth and form of fractal cities.
Article
Full-text available
Power law is a common phenomenon in natural and social sciences, including cities. Though many attempts have been made to understand the underlying mechanisms of the power law phenomenon , it means a lot to cities. This phenomenon is a manifestation of an emerging, objective law in urban population distribution, which can serve as a basis not only for theory construction from the top down, but also for simulations on urban complex systems from the bottom up, in particular as a reference for urban and regional policy making in China. Through mass urban population statistics , we found in the present paper that the R 2 values for the urban population distribution in China were high, indicating that distribution fulfills the requirement of the power law distribution. In addition, the q values scattered around unity, fulfilling the requirement of the rank-size rule. Starting from 1999 through 2009, the evolution toward the power law of the urban population distribution in China intensified, with the slope of q moving further away from −1, indicating im-balanced development of cities. 中国城市人口分布的幂次现象 赖世刚 1 ,韩昊英 2 ,方 明 2 1 国立台北大学不动产与城乡环境学系,新北市 2 浙江大学土地管理系,杭州
Article
Full-text available
Chen Yanguang, Liu Minghua. Study on fractal dimension of size distribution of cities. Bulletin of Science and Technology, 1998, 14(6): 395-400
Article
Carbon dioxide emissions may create significant social harm because of global warming, yet American urban development tends to be in low density areas with very hot summers. In this paper, we attempt to quantify the carbon dioxide emissions associated with new construction in different locations across the country. We look at emissions from driving, public transit, home heating, and household electricity usage. We find that the lowest emissions areas are generally in California and that the highest emissions areas are in Texas and Oklahoma. There is a strong negative association between emissions and land use regulations. By restricting new development, the cleanest areas of the country would seem to be pushing new development towards places with higher emissions. Cities generally have significantly lower emissions than suburban areas, and the city-suburb gap is particularly large in older areas, like New York.
Article
This paper extracted urban boundary information in six periods of 1900, 1930, 1954, 1976 1990 and 2004 by integrating TM and SPOT images, relief maps, history maps and other spatial data of the last 100 years. Based on spatial center of gravity transfer model, fractal model and urban land-population different speed rising model, urban land utilization spatial expansion process and its mechanism for 100 years. The research results show that Changchun urban land area was 5.05 km2 in 1900 and expanded 56.77 times of its area from 1900 to 2004. Changchun urban land area expansion is the quickest from 1990 to 2004 with a speed of 7.89 km2 annually. Then the expansion speed was 3.60 km2 every year from 1930 to 1954. The slowest period was from 1900 to 1930 with a speed of 0.41 km2. Changchun urban land utilization fractal character became more and more complicated with edge being filled mostly from 1976 to 1990. During other periods of time urban land utilization mainly expanded along edge area. Changchun urban population increased 87.57 times from 1900 to 2004. Then per capita available urban land area gradually reduced from 1900 to 2004. Based on urban land and population different speed rising model, the relationship between Changchun urban land and population was analyzed. The result indicated that Changchun urban land and population different speed rising models change irregularly because Changchun urban development and construction was affected by Japanese and Russia empire invasions before 1948.
Article
Based on a review on urban allometric equation, this paper analyzes the allometric growth relationship between urban area and urban population in China by using the data of the urban built-up area and the total population of cities and towns in 1990, 2000 and 2005. After an in-depth discussion of the standard value of the scaling factor, the paper concludes that urban growth in China is negative allometry in 1990 and positive one in 2000 and 2005. Sample cities in the eastern and western China also show the same law. In contrast to this, cities in central China maintain the original proportion of built-up land growth with population growth; however, they show positive allometry in 2005 because of the promotion policy for central China. The conclusion is closely related to the changes in migrant population, housing system reform and the rapid expansion of development zones. E-mail: [email protected] /* */