Monitoring urbanization dynamics in India using DMSP/OLS night time lights and SPOT-VGT data

Abstract

India is a rapidly urbanizing country and has experienced profound changes in the spatial structure of urban areas. This study endeavours to illuminate the process of urbanization in India using Defence Meteorological Satellites Program – Operational Linescan System (DMSP-OLS) night time lights (NTLs) and SPOT vegetation (VGT) dataset for the period 1998–2008. Satellite imagery of NTLs provides an efficient way to map urban areas at global and national scales. DMSP/OLS dataset however lacks continuity and comparability; hence the dataset was first intercalibrated using second order polynomial regression equation. The intercalibrated dataset along with SPOT-VGT dataset for the year 1998 and 2008 were subjected to a support vector machine (SVM) method to extract urban areas. SVM is semi-automated technique that overcomes the problems associated with the thresholding methods for NTLs data and hence enables for regional and national scale assessment of urbanization. The extracted urban areas were validated with Google Earth images and global urban extent maps. Spatial metrics were calculated and analyzed state-wise to understand the dynamism of urban areas in India. Significant changes in urban proportion were observed in Tamil Nadu, Punjab and Kerala while other states also showed a high degree of changes in area wise urban proportion.

Introduction

The global urban land extent covers a very small fraction of Earth's land surface with urban area estimates ranging from 0.3 × 10⁶ km² (Danko, 1992) to 3.53 × 10⁶ km² (CIESIN, 2004).

The relatively small fraction of the total Earth's land surface hold more than half of the current world population, and the total urban population is projected to increase by more than 2 billion in 2050 (United Nations, 2010). Increasing urban population due to natural growth of population in urban areas, rural to urban migration and reclassification of rural areas as urban in course of time (Mitra and Murayama, 2009), stands as a major driver for urban growth (Torrens, 2006). A study on global urban expansion of 120 cities showed that the annual urban land-use change rate is 3.3% which is even higher than the increase in urban population (Angel et al., 2005). The expansion of urban areas contributes significantly to regional and global environmental change (Mills, 2007, Seto and Satterthwaite, 2010). In fact, the contribution of urban areas on global environmental change is likely to become more pronounced in the future due to global urban expansion (Mills, 2007). Currently, urban areas contribute 70% of the total carbon dioxide emissions and more than 70% of the world's total energy consumption (Seto and Satterthwaite, 2010, Madlener and Sunak, 2011). Urbanization has severe implications to environmental systems (Grimm et al., 2008) and human health (Dye, 2008).

According to United Nations, the increase in the world's urban population is concentrated in few countries with China and India projected to account for about a third of increase in urban population in the coming decades (United Nations, 2010). This significant contribution towards increasing world's urban population by China and India, reveals the alarming rates of urbanization in these countries. According to Census of India 2011 report, the population of India in the beginning of 20th century was 238.4 million which has increased more than four times in a period of 110 years to reach 1210 million in 2011. This phenomenal increase is observed to be three fold in the latter half of the period 1901–2011 (Census of India, 2011). The decadal growth of urban population in India between 2001 and 2011 is 31.80%, which is much higher than the decadal growth of rural population in India, i.e. 12.18% (Census of India, 2011). All these demographic figures clearly indicate high rate of urbanization in India. The rapid urbanization in India has also led to several problems such as lack of appropriate infrastructure and basic amenities (Ramachandra et al., 2012, Goli et al., 2011, Vij, 2012), increase in flood prone areas (Suriya and Mudgal, 2012), heat island formation (Deosthali, 2000), vegetation clearing and fragmentation (Nagendra et al., 2012) and others. While the impacts of urbanization are multifaceted and are at multiple spatiotemporal scales, it is essential to timely assess urban growth.

Remote sensing makes large amount of data available with continuous spatial and temporal coverage that enables for periodic monitoring of large urban agglomerations at several spatial scales (Griffiths et al., 2010). Previous studies have utilized fine and medium resolution imageries to study changes in the urban land-use for individual cities or metropolitan areas (Sokhi et al., 1989, Weber and Puissant, 2003, Sudhira et al., 2004, Yuan et al., 2005, Xian and Crane, 2005, Jat et al., 2008, Taubenböck et al., 2012). Apart from remote sensing data in optical domain, Radar data is also utilized to estimate human population and analyze urban growth (Henderson and Xia, 1997, Dell’Acqua and Gamba, 2003, Taubenböck et al., 2012). Coarse resolution remote sensing data has been used to produce static global urban extent maps. Examples include DMSP-OLS night-time imageries (Elvidge et al., 2001, Schneider et al., 2005, CIESIN, 2004), MODIS data (Schneider et al., 2005, Schneider et al., 2010) and others. Urban areas are typically analyzed using techniques pertaining to pixel based (Sudhira et al., 2004, Schneider et al., 2005, Schneider et al., 2010, Cao et al., 2009, Taubenböck et al., 2012) and object based classification (Taubenböck et al., 2012, Jacquin et al., 2008). With respect to DMSP-OLS night-time lights (NTLs), most commonly utilized methods to extract urban areas are global fixed and local-optimized thresholding (Imhoff et al., 1997, Henderson et al., 2003, Cao et al., 2009). These methods are although complex and time consuming to map urban areas at regional or global scales (Lu et al., 2008, Cao et al., 2009). Coarse resolution urban area mapping is also possible with multiple sources dataset (Schneider et al., 2005, Schneider et al., 2010, Cao et al., 2009). Cao et al. (2009) have shown the effectiveness of using DMSP-OLS NTLs data along with SPOT-VGT data to extract regional urban extents using semi-automatic support vector machines (SVM) based classification algorithm. Schneider et al. (2005) used DMSP-OLS NTLs along with MODIS data to map urban areas. Lu et al. (2008) developed a human settlement index based on DMSP-OLS NTLs and MODIS NDVI data to map regional human settlements. The increasing research on utilizing DMSP-OLS NTLs to map urban extents is attributed to high statistical correlation of human induced light emissions with urban population, economic activity and urban land cover (Sutton et al., 2001, Zhang and Seto, 2011, Ma et al., 2012). Although previous studies differ in the scope and methods to analyze NTLs in estimating city-scale, regional or global urbanization attributes, the results of these analyses indicate high potential of NTLs data to monitor urbanization dynamics in absence of socio economic data needed to characterize the urbanization process (Ma et al., 2012, Ghosh et al., 2010).

In the context of developing countries like India where there is absence of high temporal frequency socio-economic data, remote sensing data can be utilized as a proxy data for monitoring urbanization dynamics. Given the impending urban demographic shifts in India and other developing countries, and considering the importance of urbanization to social, economic and environmental processes, this study aims to monitor urban growth using DMSP-OLS NTLs and SPOT vegetation (VGT) datset for nationwide urbanization assessment. The timely assesment of urbanization dynamics is important to the scientific and policy community for addressing the linkages of urbanization to various social, economic and environmental processes. Our study thus addresses the following questions: Is the DMSP-OLS NTLs and NDVI data apt to analyze urbanization process? What are the spatio-temporal patterns in growth and distribution of urban areas in different states of India? The specific objectives of this study correspond to extracting urban areas in India for the year 1998 and 2008 and monitoring nationwide patterns of urban growth using spatial metrics.