Abstract
This paper introduces a new city-level panel dataset constructed using satellite nighttime light imagery and grid population data. The data contain over 1500 cities covering 43 Asian and Pacific countries/economies from 1992 to 2016. With the dataset, we perform a variety of analyses for the region as a whole as well as the five largest countries in the region. The exercise produces some novel evidence on several policy-relevant topics including urbanization status and patterns, relations between urbanization and economic growth, evolution of urban systems, primate cities, testing Zipf’s law and Gibrat’s law, the drivers of city growth, and emergence of city clusters.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For instance, 59 countries use administrative designations as the sole criterion for defining a city, while another 62 combine the administrative criterion with others to distinguish between urban and rural areas.
The countries include 42 developing member countries of the Asian Development Bank (ADB) along with Japan. In the paper, we refer to these 43 countries as Asia and the Pacific region.
The data were accessed at the website of the National Geophysical Data Center of NOAA, https://ngdc.noaa.gov/.
It is noted that different thresholds are adopted to draw urban boundaries in the literature. Examples include 5 in Zhang and Seto (2011); 13 in Ellis and Roberts (2015) and in Zhou et al. (2015); 33 in Tewari et al. (2017); and 35 in Harari (2016). A positive threshold is needed if one uses the pre-deblurring data to delineate urban extent. However, a uniform positive threshold across years may not yield consistent definition of urban scope given that different sensors (OLS vs. VIIRS) were used over the period and the same sensor performed differently over its lifecycle. Moreover, a sensible threshold value to define urban extent, if exists, should probably vary across countries.
GRUMP data were generated by the Columbia University Center for International Earth Science Information Network in collaboration with the International Food Policy Research Institute, the World Bank, and Centro Internacional de Agricultura Tropical, through combining census and geospatial datasets. It can be accessed at http://sedac.ciesin.columbia.edu/data/collection/grump-v1. It is worth noting that the units in the GRUMP database are not from the same administrative level within, as well as across, countries.
Clearly, our natural cities are distinct from administratively-defined cities and may cover broader areas as compared to other definitions applying density and/or economic indicators. Note that the size criterion adopted to select GRUMP units was population reported in GRUMP, which was generally sourced from official statistics. A natural city’s population is by no means the same as that of the corresponding GRUMP unit since they have distinct space and use distinct methods to obtain their population counts. Thus, a natural city could still be below 100,000 in 2000, or even after 2000.
Although the number of natural cities is fixed across time, our data can still capture newly urbanized areas if these were near the existing natural cities and grew to become part of the natural cities.
The LandScan data can be accessed at https://www.landscan.ornl.gov.
We did not apply population as weights to the regressions because our attempt here is to understand the relationship at country level, so each country is treated equally.
Regressions for 2000 yield highly similar results.
Soo (2005) found the coefficients greater than 1 in magnitude for China, India, Indonesia, and Japan, and statistically equal to 1 for Pakistan if using administrative-based city data.
To reduce the influence of extreme values on the estimation, we excluded cities with an average annual growth rate greater than 0.2, despite that our data validation does not identify problems with them. They were primarily very small cities in 2000.
Indonesia and the group of 38 countries have positive estimates for the distance variable. This may be due to the fact that some of these countries like Indonesia, Philippines, and Viet Nam have many seaports, so the distance to seaport does not necessarily capture the advantage of access to international markets.
Given the way we identified natural cities, some of them could have in fact been city clusters in 1992. Such examples include Guangzhou, Osaka, Seoul, and Tokyo, which contain multiple administrative cities that could have been independent natural cities if traced further back. However, these were rare cases, especially in the developing world at that time.
There were 19 natural cities reaching population of 10 million or more in 2016, of which 18 belonged to some city clusters. Karachi of Pakistan was the only natural city with more than 10 million population that did not form a city cluster with other cities.
In Guangzhou–Huizhou cluster, the leading cities should be Guangzhou and Shenzhen, each with population greater than 10 million, despite that Shenzhen was considered part of the natural city of Guangzhou according to the NTL in 1992.
References
Abrahams A, Oram C, Lozano-Gracia N (2018) Deblurring DMSP nighttime lights: a new method using Gaussian filters and frequencies of illumination. Remote Sens Environ 210:242–258
Ades AF, Glaeser EL (1995) Trade and circuses: explaining urban giants. Q J Econ 110(1):195–227
Brueckner JK, Sridhar KS (2012) Measuring welfare gains from relaxation of land-use restrictions: the case of India’s building-height limits. Region Sci Urban Econ 42(6):1061–1067
Buhnik S (2010) From shrinking cities to Toshi no Shukushō: identifying patterns of urban shrinkage in the Osaka metropolitan area. Berk Plann J 23(1):134
Chan KW, Henderson V, Tsui KY (2008) Spatial dimensions of Chinese economic development. In: Brandt L, Rawski TG (eds) China’s great transformation: origins, mechanisms, and consequences of the post-reform economic boom. Cambridge Unviersity Press, pp 776–828
Chauvin JP, Glaeser E, Ma Y, Tobio K (2017) What is different about urbanization in rich and poor countries? Cities in Brazil, China, India and the United States. J Urban Econ 98:17–49
Colmer J (2016) Urbanisation, growth, and development: evidence from India. Unpublished review paper
da Mata D, Deichmann U, Henderson JV, Lall SV, Wang HG (2007) Determinants of city growth in Brazil. J Urban Econ 62(2):252–272
Duranton G (2007) Urban evolutions: the fast, the slow, and the still. Am Econ Rev 97(1):197–221
Duranton G (2008) From cities to productivity and growth in developing countries. Can J Econ/Revue Can d’écon 41(3):689–736
Duranton G (2015) Growing through cities in developing countries. World Bank Res Obser 30(1):39–73
Duranton G (2016) Determinants of city growth in Colombia. Pap Region Sci 95(1):101–131
Duranton G, Puga D (2001) Nursery cities: urban diversity, process innovation, and the life cycle of products. Am Econ Rev 91(5):1454–1477
Eeckhout J (2004) Gibrat’s law for (all) cities. Am Econ Rev 94(5):1429–1451
Ellis P, Roberts M (2015) Leveraging urbanization in South Asia: managing spatial transformation for prosperity and livability. The World Bank, New York
Gabaix X (1999) Zipf’s law for cities: an explanation. Q J Econ 114(3):739–767
Gabaix X, Ibragimov R (2011) Rank − 1/2: a simple way to improve the OLS estimation of tail exponents. J Bus Econ Stat 29(1):24–39
Goldblatt R, Stuhlmacher MF, Tellman B, Clinton N, Hanson G, Georgescu M, Wang C et al (2018) Using landsat and nighttime lights for supervised pixel-based image classification of urban land cover. Remote Sens Environ 205:253–275
Gollin D, Jedwab R, Vollrath D (2016) Urbanization with and without industrialization. J Econ Growth 21(1):35–70
Harari M (2016) Cities in bad shape: urban geometry in India. Processed, Wharton School of the University of Pennsylvania
Hasan R, Jiang Y, Kundu D (2018) The growth of Indian cities and “good” jobs: evidence from the 2000s. In: Shah S, Bosworth B, Muralidharan K (eds) Indian policy forum, vol 2017/18. Sage, London
Hattori K, Kaido K, Matsuyuki M (2017) The development of urban shrinkage discourse and policy response in Japan. Cities 69:124–132
Henderson JV (2005) Urbanization and growth. In: Aghion P, Durlauf S (eds) Handbook of economic growth, vol 1. Elsevier, London
Henderson JV, Quigley J, Lim E (2009) Urbanization in China: policy issues and options. Unpublished manuscript, Brown University
Henderson JV, Storeygard A, Weil DN (2012) Measuring economic growth from outer space. Am Econ Rev 102(2):994–1028
Hsu F-C, Baugh K, Ghosh T, Zhizhin M, Elvidge C (2015) DMSP-OLS radiance calibrated nighttime lights time series with intercalibration. Remote Sens 7(2):1855–1876
Kohli HS, Sharma A, Sood A (eds) (2011) Asia 2050: realizing the Asian century. Sage, London
Martin P, Ottaviano GIP (1999) Growing locations: industry location in a model of endogenous growth. Eur Econ Rev 43(2):281–302
Mellander C, Lobo J, Stolarick K, Matheson Z (2015) Night-time light data: a good proxy measure for economic activity? PLoS ONE 10(10):e0139779
Michalopoulos S, Papaioannou E (2013) National institutions and subnational development in Africa. Q J Econ 129(1):151–213
Rosen KT, Resnick M (1980) The size distribution of cities: an examination of the Pareto law and primacy. J Urban Econ 8(2):165–186
Schaffar A, Dimou M (2012) Rank-size city dynamics in China and India, 1981–2004. Reg Stud 46(6):707–721
Small C, Pozzi F, Elvidge CD (2005) Spatial analysis of global urban extent from DMSP-OLS night lights. Remote Sens Environ 96(3–4):277–291
Soo KT (2005) Zipf’s law for cities: a cross-country investigation. Region Sci Urban Econ 35(3):239–263
Soo KT (2014) Zipf, Gibrat and geography: evidence from China, India and Brazil. Pap Region Sci 93(1):159–181
Tewari M, Alder S, Roberts M (2017) India’s urban and spatial development in the post-reform period: an empirical analysis based on nightlight data. Unpublished
United Nations, Department of Economic and Social Affairs, Population Division (2018) World urbanization prospects: the 2018 revision (ST/ESA/SER.A/420). New York: United Nations
Zhang Q, Seto KC (2011) Mapping urbanization dynamics at regional and global scales using multi-temporal DMSP/OLS nighttime light data. Remote Sens Environ 115(9):2320–2329
Zhou N, Hubacek K, Roberts M (2015) Analysis of spatial patterns of urban growth across South Asia using DMSP-OLS nighttime lights data. Appl Geogr 63:292–303
Acknowledgements
The paper benefits greatly from guidance and comments of Gilles Duranton and Rana Hasan. The author thanks Marjorie Remolador for outstanding GIS work and Renz Adrian Calub for excellent research assistance. Seminar participants at ADB also provided very useful suggestions. The views expressed in this publication are those of the author and do not necessarily reflect the views and policies of ADB or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of its use.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix: Additional details of development of the natural city dataset
Appendix: Additional details of development of the natural city dataset
1.1 Defense meteorological satellite program-operational linescan system nighttime lights data
The United States Air Force Defense Meteorological Satellite Program (DMSP) initiated nighttime visible imaging that recorded the intensity of Earth-based lights, with satellites installed with Operational Linescan System (OLS) sensors in the 1970s. The data were archived in digital format from 1992 to 2013. As Table 12 shows, for several years, two satellites were flying simultaneously to collect imaging data. For this study, we utilized data from the following satellites: F10 for 1992, F12 for 1995, F15 for 2000, F16 for 2005, and F18 for 2010. When a year was covered by two satellites, the newer satellite was chosen because satellites tend to deteriorate over time, which affects the quality of information derived from them.
Scientists from the National Geophysical Data Center of the National Oceanic and Atmospheric Administration developed an automatic algorithm to filter the nighttime light (NTL) observations and remove unwanted lights, noises, and cloud presence from the raw DMSP-OLS NTL data (Hsu et al. 2015). Among the products released to the public, we used the annual raw average visible lights, which are available for all years of interest and allow for identification of dimmer suburban and rural areas.
DMSP-OLS NTL images suffer from significant blurring known as “blooming” or “overglowing.” According to Small et al. (2005), blooming is the result of the relatively coarse spatial resolution of the OLS sensor, the large overlap in the footprints of adjacent OLS pixels, and the accumulation of geolocation errors in the compositing process. With these blurring in the NTL images, the digital number values of pixels outside the actual illuminated areas are still positive. This is a major issue in terms of urban area delineation, since it makes it hard to identify urban areas from non-urban.
To resolve the issue, several methodologies were developed and these can be classified mainly into two types: (1) thresholding based (Small et al. 2005; Abrahams et al. 2018) and (2) classification based (Goldblatt et al. 2018). The thresholding method is done by applying frequency thresholds to the image to reduce the spatial exaggeration on the lighted area, while the classification method uses other auxiliary data such as daytime satellite images.
In this study, we employed the latest method developed in Abrahams et al. (2018), which is independent from auxiliary data and coded through MATLAB. The method deblurs the images by applying two filters. The first filter tries to invert the blurring process in a noise-sensitive manner using a standard Weiner deconvolution. In this filter, it is assumed that the light was blurred via a symmetric Gaussian point-spread function. The second filter uses the PCT image wherein the pixel at which a light source is located will always be a local maximum in the PCT image. It is implied that when a pixel is not a local maximum in the PCT image, the light recorded for those pixels are considered erroneous. The authors showed that when a 20% PCT threshold is applied, most areas that are infrequently lit are removed and the remaining pixels approximate urban area light well. According to the assessment in the Abrahams et al. (2018), the deblurred NTL data are comparable to the stable lights version of the National Oceanic and Atmospheric Administration products, but they have much less exaggeration on urban areas.
Interannual calibration was performed on the deblurred data following Zhang and Seto (2011). However, inter-calibration does not make significant difference in our exercise as we only applied luminosity cutoff value at zero to delineating polygons. Not many pixels have changed from a positive luminosity to zero or negative or from zero to positive values after inter-calibration.
1.2 Visible infrared imaging radiometer suite nighttime lights data
Nighttime lights from Visible Infrared Imaging Radiometer Suite (VIIRS) succeeded the widely-used DMSP-OLS NTL data in 2012. This dataset has a finer resolution of 15 arc seconds and gives the actual radiance captured by the sensor. The monthly data are available from 2012 to the present day, and annual composite data are available for 2015 and 2016. We use the “vcm-orm-ntl” (VIIRS Cloud Mask—Outlier Removed—Nighttime Lights) version of the 2016 annual composite VIIRS NTL data, since it contains cloud-free average radiance values that have undergone an outlier removal process to filter out fires and other ephemeral lights, with background (non-lights) set to zero.
Unlike the DMSP-OLS NTL images, the VIIRS NTL data do not need deblurring and inter-calibration, since it already has an on-board calibration and the movement of the VIIRS satellite prevents overlaps in the images, which mainly caused the blurs in the DMSP-OLS product.
1.3 Construction of natural city sample
We took the following steps to construct the natural city sample for this study, with the deblurred DMSP-OLS NTL images for data up to 2010 and VIIRS NTL images for 2016.
-
1.
Delineating extents of human settlements We implemented a practical definition of human settlements, i.e., all pixels with a positive luminosity value (i.e., digital number). Those with luminosity value equal to zero are not taken as human settlements. The delineated polygons with small gaps between them (1 km for DMSP-OLS NTLs and 0.5 km for VIIRS NTLs) are joined together as one settlement.
-
2.
Identifying urban agglomerations (natural cities) We used Global Rural Urban Mapping Project (GRUMP)’s population of year 2000 to select urban units (1964 units selected). We did not use population of earlier years, say 1990, because that may leave out many cities which were small in 1990 and grew faster in later years unless we use a very low threshold. We then linked these units with NTL-based polygons of 1992 through GIS software and manual matching. Most GRUMP units are contained in the polygons or find polygons close to them. These linked polygons are referred to as natural cities. We track them from 1992 to 2016 in terms of spatial extent with NTL.
-
3.
Including major cities and large polygons We included the major cities from the countries with no city meeting the criterion of 2000 GRUMP population specified in (2) and polygons greater than 100 square kilometers (km2) in 2000, although the associated administrative cities do not meet the criterion.
-
4.
Estimating population of the natural cities We overlaid the natural cities to the grid population from LandScan and aggregated the population count per polygon. Population of natural cities are dynamic as both grip population and the spatial extent of each natural city change over time.
We performed considerable amount of data validation focusing on the following cases: (1) city polygons with an area less than 2 km2; (2) natural cities with less than 100,000 population count for at least one of the analysis years; (3) natural cities with extraordinary area growth or shrinking from 1992 to 2016; and (4) natural cities with extraordinary population growth or shrinking from 2010 to 2016. Some issues were detected and fixed including:
-
1.
Some GRUMP settlement points may have been placed incorrectly and therefore the wrong polygons were tagged.
-
2.
When the settlement points did not fall in any NTL polygons, we assigned the nearest polygons to them. This does not always yield correct matching. We conducted extensive visual checking on the luminosity, population agglomerations, and daytime satellite imagery to identify and fix problematic cases.
-
3.
Several urban areas in 2016 included roads connecting to other settlements, which caused extraordinary area and population growth as opposed to 2010. We redefined the urban area in 2016 by cutting roads and the connected settlements from the city if the roads were obvious and/or the connected settlements were at least 20 km away from the city being analyzed.
-
4.
Some polygons with area greater than 100 km2 were found to be oil fields with no administrative city located in them. We eliminated these from the sample.
We obtained a more logical dataset after conducting all these checks and revisions. While extraordinary caution was exercised in processing the data, our dataset would not be free of errors. Measurement errors are most likely to arise from imperfection of the NTL data and the grid population data. For instance, the quality of satellite imagery deteriorates with the length of the satellite’s service years, which adversely affects the accuracy of the NTL data. The LandScan data are taken as given because validating LandScan’s algorithm for the project purpose is beyond the capacity of the team.
