Elsevier

Energy

Volume 100, 1 April 2016, Pages 177-189
Energy

Detecting urban-scale dynamics of electricity consumption at Chinese cities using time-series DMSP-OLS (Defense Meteorological Satellite Program-Operational Linescan System) nighttime light imageries

https://doi.org/10.1016/j.energy.2016.01.058Get rights and content

Highlights

  • Spatiotemporal pattern of urban EC (electricity consumption) in China was analyzed.

  • Time series of pixel-based EC in China were estimated using night-light intensity.

  • The ratio of EC in urban areas rose from 50.6% in 2000–71.32% in 2012.

  • Suburban areas contributed more to the increase of total EC than urban cores.

  • Suburban areas are more crucial in Chinese energy sustainability.

Abstract

A better understanding of the spatiotemporal pattern of energy consumption at the urban scale is significant in the interactions between economic activities and environment. This study assessed the spatiotemporal dynamics of EC (electricity consumption) in UC (urban cores) and SR (suburban regions) in China from 2000 to 2012 by using remotely sensed NTL (nighttime light) imagery. Firstly, UC and SR were extracted using a threshold technique. Next, provincial level model was calibrated yearly by using Enhanced Vegetation Index and population-adjusted NTL data as independent variables. These models were then applied for pixel-based estimation to obtain time-series EC data sets. Finally, the spatiotemporal pattern of EC in both UC and SR were explored. The results indicated that the proportion of EC in urban areas rose from 50.6% to 71.32%, with a growing trend of spatial autocorrelation. Cities with high urban EC were either located in the coastal region or belonged to provincial capitals. These cities experienced a moderate to a rapid growth of EC in both UC and SR, while a slow growth was detected for the majority of western and northeastern cities. The findings suggested that EC in SR was more crucial for sustainable energy development in China.

Introduction

Global warming, urbanization, energy consumption, and carbon dioxide (CO2) emission are crucial, interconnected themes in the 21st century [1], [2], [3], [4], [5]. Specifically, urbanization, which is usually accompanied by economic growth, population increase, and improved living standards, contributes heavily to the increase of energy consumption and carbon emission [1], [6]. Urban areas accounted for approximate 67% of energy use and 70% of CO2 emissions in 2008, and these values are projected to 73% and 76% by 2030, respectively [7]. How to improve energy efficiency and reduce per capita carbon emission at urban areas becomes a compelling issue for the sake of the adaption and mitigation of climate change [5]. The knowledge of energy consumption in urban areas and its spatiotemporal dynamics is thus crucial, yet this type of information is usually unavailable [5], [8], [9].

As an important component of energy, EC (electricity consumption) relates to every aspect of commercial activities, industrial productions, and urban residents' daily activities. As such, EC is often one of the largest sources of carbon emission [10]. Therefore, a better knowledge of the spatiotemporal pattern of EC at urban areas is of significance to understand urban energy consumption and carbon emission. However, EC data sources are scarce at the urban scale, especially in developing countries like China [5], [11], [12]. Census data, the primary source of EC data, is usually acquired based on administrative units (e.g., country, province, or county) and at the temporal interval of years, and can hardly satisfy the needs for urban scale applications [11], [12].

A large number of models have been proposed to estimate EC. Commonly used methods include linear regression models [13], support vector machines [14], and artificial neural networks [14], [15]. Despite their relatively high accuracy, these models can hardly be applied to estimate urban scale EC because (1) they are usually developed for modeling at the country scale and (2) the multiple inputs for modeling in the urban scale cannot be satisfied. NTL (nighttime light) signal recorded by DMSP (Defense Meteorological Satellite Program)'s OLS (Operational Linescan System) has been proved effective in the estimation of urban EC, because NTL can provide spatially explicit information of EC [12], [16], [17], [18], [19], [20], [21], [22], [23].

The usage of DMSP-OLS NTL imagery for estimating EC patterns at multiple scales has been extensively documented [16], [17], [18], [19], [20], [24], [25], [26]. Elvidge et al. [17], [24] reported the log–log relationship between lit area and EC for 21 and 200 countries, respectively; Amaral et al. [19] modeled EC in Brazilian Amazon from extracted lighted area at municipal level; Lo [18] established the logarithmic relationship between EC and lit area for 35 Chinese capital cities. Additionally, the examination of spatiotemporal dynamics of EC at different scales has been conducted. For instance, by using time series of DMSP-OLS NTL imageries, Chand et al. [23] characterized spatiotemporal pattern of EC in major cities and states of India; He et al. [21], [22] analyzed EC pattern at the county level in the Mainland China. Furthermore, to detect the spatiotemporal pattern of EC at the sub-county level, Cao et al. [12] proposed a top-down method to model pixel-based EC dynamics in China, using NTL, population density, and GDP (gross domestic product) as the variables. Although these studies have documented the effectiveness of NTL for predicting EC with varying degree of success, most of these studies focused on global, continental, or national level estimates. What is missing in the literature is the evaluation of NTL for characterizing spatiotemporal dynamics of EC at the urban and suburban scales. In addition, the capacity of DMSP-OLS NTL imagery for evaluating the relationship between urbanization and EC is seldom conducted. This is especially true for the time period of the first decade of the 21st century when urbanization in China accelerated [27].

The objective of this study is to detect and assess the spatiotemporal pattern of EC at the urban and suburban scale by using time-series NTL imageries and auxiliary data sets in China from 2000 to 2012, so that a better understanding of the interactions between urbanization and EC can be achieved. The rest of the paper was organized as follows. Section 2 described data collection and the study area. The methods for delineating urban cores and suburban regions, estimating pixel-based EC, and analyzing spatiotemporal pattern of urban EC were provided in Section 3. Section 4 provided an analysis of the results, followed by discussion in Section 5 and conclusions in Section 6.

Section snippets

Study area and data sets

The Mainland China has experienced a rapid urbanization since the economic reform in 1978 [6]. This is especially true for the first decade of the 21st century when urbanization in China accelerated [27]. In particular, population size reached to 1.34 billion at the end of 2010 and urban population increased from 36.22% to 51.27% between 2000 and 2011 [6]. The rapid economic growth associated with urbanization steadily raised the demand of electricity, with EC increased by 3.6 times, from 1356

Methodology

There are three major analytical steps: data preprocessing, extraction of urban extent (including urban cores and suburban regions) and pixel-based EC estimates, and spatiotemporal pattern analysis of urban EC (Fig. 3).

Pixel-based EC for the mainland China

Fig. 5 presented the estimated result in 2000 and 2012. Visually, the overall pattern of EC in 2000 and 2012 were similar. There were some regions with remarkably higher EC, such as Jing-Jin-Tang Metropolitan Region, the Yangtze River Delta, the Pearl River Delta, the Sichuan Basin, and most of the provincial capitals. Gridded EC in the urban cores could be as high as 38.5 million kWh and 58.4 million kWh for 2000 and 2012, respectively. It was also demonstrated in Fig. 5 that the urban cores

Urban EC growth in relation to socioeconomic development and policy implications

Table 3 showed the linear regressions between urban EC growth and the increases of urban population and GDP at the provincial level. The linear regression between urban EC growth and the increase of urban GDP was statistically significant (F = 37.33, sig. = 0.000) and had an R2 of 0.58. Another regression with the incorporation of urban population growth improved the modeling with R2 of 0.66 (F = 24.77, sig. = 0.000). The results in Table 3 indicated that population urbanization contributed to

Conclusions

Time-series DMSP-OLS nighttime light imageries revealed that the proportion of EC in the urban areas rose from 50.6% to 71.32% in China in the first decade of the 21st century, with a growing trend of spatial autocorrelation. Cities with high and rapidly increasing urban EC were either located in coastal region or belonged to provincial capitals, while a slow growth was detected for the majority of western and northeastern cities and part of central China. Due to the intensive demand of

Acknowledgments

The authors are grateful of comments and suggestions provided by anonymous reviewers and the editor, which helped to improve this manuscript. We further thank National Oceanic and Atmosphere Administration/National Geophysical Data Center (NOAA/NGDC) and Global Administrative Areas for provision of DMSP-OLS night-time light products and administrative boundary data. Weng acknowledges a visiting chair professorship awarded to him by South China Normal University.

References (59)

  • C.D. Elvidge et al.

    Night-time lights of the world: 1994–1995

    ISPRS J Photogramm

    (2001)
  • S. Amaral et al.

    Estimating population and energy consumption in Brazilian Amazonia using DMSP night-time satellite data

    Comput Environ Urban

    (2005)
  • Q. Huang et al.

    Detecting the 20 year city-size dynamics in China with a rank clock approach and DMSP/OLS nighttime data

    Landsc Urban Plan

    (2015)
  • C. Justice et al.

    An overview of MODIS Land data processing and product status

    Remote Sens Environ

    (2002)
  • Q. Zhang et al.

    The vegetation adjusted NTL urban index: a new approach to reduce saturation and increase variation in nighttime luminosity

    Remote Sens Environ

    (2013)
  • Z. Liu et al.

    Extracting the dynamics of urban expansion in China using DMSP-OLS nighttime light data from 1992 to 2008

    Landsc Urban Plan

    (2012)
  • C.N. Doll et al.

    Estimating rural populations without access to electricity in developing countries through night-time light satellite imagery

    Energy Policy

    (2010)
  • Y. Su et al.

    China's 19-year city-level carbon emissions of energy consumptions, driving forces and regionalized mitigation guidelines

    Renew Sust Energy Rev

    (2014)
  • Q. Wang et al.

    Exploring the relationship between urbanization, energy consumption, and CO2 emissions in different provinces of China

    Renew Sust Energy Rev

    (2016)
  • C. Zhang et al.

    Panel estimation for urbanization, energy consumption and CO2 emissions: a regional analysis in China

    Energy Policy

    (2012)
  • C. Kontgis et al.

    Monitoring peri-urbanization in the greater Ho Chi Minh city metropolitan area

    Appl Geogr

    (2014)
  • J.A. Cherni et al.

    Renewable energy policy and electricity market reforms in China

    Energy Policy

    (2007)
  • S. Xu et al.

    The reform of electricity power sector in the PR of China

    Energy Policy

    (2006)
  • H. Ngan

    Electricity regulation and electricity market reforms in China

    Energy Policy

    (2010)
  • V. Oikonomou et al.

    Energy saving and energy efficiency concepts for policy making

    Energy Policy

    (2009)
  • S.C. Lillemo

    Measuring the effect of procrastination and environmental awareness on households' energy-saving behaviours: an empirical approach

    Energy Policy

    (2014)
  • C.S. Hayles et al.

    Social housing tenants, climate change and sustainable living: a study of awareness, behaviours and willingness to adapt

    Sustain Cities Soc

    (2015)
  • Y. Zhou et al.

    A cluster-based method to map urban area from DMSP/OLS nightlights

    Remote Sens Environ

    (2014)
  • H. Lu et al.

    Spatial effects of carbon dioxide emissions from residential energy consumption: a county-level study using enhanced nocturnal lighting

    Appl Energy

    (2014)
  • Cited by (95)

    • Modelling spatiotemporal carbon dioxide emission at the urban scale based on DMSP-OLS and NPP-VIIRS data: A case study in China

      2022, Urban Climate
      Citation Excerpt :

      China's rapid carbon emissions from urban development have generated substantial attention (Chen et al., 2019; Wu et al., 2022; Cai et al., 2021). Focus has been placed on county-level cities (Wei et al., 2021; Huo et al., 2021; Wang and Huang, 2019; Xie and Weng, 2016), major cities (Liu et al., 2021; Li et al., 2013a, 2013b; Meng et al., 2014; Miao, 2017; Mi et al., 2016), the Yangtze River Economic Zone (Chen et al., 2018a, 2018b), and the Yellow River Delta (Xu and Xu, 2021; Sun et al., 2014). However, most studies on urbanization in China have focused on already built up coastal areas, with relatively few studies on emerging urban centres in the west (Liu et al., 2012).

    • Automated Extraction of Energy Systems Information from Remotely Sensed Data: A Review and Analysis

      2022, Applied Energy
      Citation Excerpt :

      A number of studies in this space estimate energy consumption utilizing NTL but tend to vary in the geographic coverage and resolution of their response (output) data. Xie et al. [210] reported sub-city level estimation in China where urban and suburban areas were studied separately. Sahoo et al. [215] reported sub-state level estimation in India where the State of Uttar Pradesh was divided into 19 zones based on their respective power suppliers.

    View all citing articles on Scopus
    View full text