Logo Logo
Switch language to English
Smollich, Susan (2015): An integrative approach using remote sensing and social analysis to identify different settlement types and the specific living conditions of its inhabitants: the case study of mega city Delhi, India. Dissertation, LMU München: Fakultät für Geowissenschaften



Someday in 2007, the world population reached a historical landmark: for the first time in human history, more than half of the world´s population was urban. A stagnation of this urbanization process is not in sight, so that by 2050, already 70 percent of humankind is projected to live in urban settlements. Over the last few decades, enormous migrations from rural hinterlands to steadily growing cities could be witnessed coming along with a dramatic growth of the world’s urban population. The speed and the scale of this growth, particularly in the so called less developed regions, are posing tremendous challenges to the countries concerned as well as to the world community. Within mega cities the strongest trends and the most extreme dimensions of the urbanization process can be observed. Their rapid growth results in uncontrolled processes of fragmentation which is often associated with pronounced poverty, social inequality, socio-spatial and political fragmentation, environmental degradation as well as population demands that outstrip environmental service capacity. For the majority of the mega cities a tremendous increase of informal structures and processes has to be observed. Consequentially informal settlements are growing, which represent those characteristic municipal areas being subject to particularly high population density, dynamics as well as marginalization. They have quickly become the most visible expression of urban poverty in developing world cities. Due to the extreme dynamics, the high complexity and huge spatial dimension of mega cities, urban administrations often only have an obsolete or not even existing data basis available to be at all informed about developments, trends and dimensions of urban growth and change. The knowledge about the living conditions of the residents is correspondingly very limited, incomplete and not up to date. Traditional methods such as statistical and regional analyses or fieldwork are no longer capable to capture such urban process. New data sources and monitoring methodologies are required in order to provide an up to date information basis as well as planning strate¬gies to enable sustainable developments and to simplify planning processes in complex urban structures. This research shall seize the described problem and aims to make a contribution to the requirements of monitoring fast developing mega cities. Against this background a methodology is developed to compensate the lack of socio-economic data and to deduce meaningful information on the living conditions of the inhabitants of mega cities. Neither social science methods alone nor the exclusive analysis of remote sensing data can solve the problem of the poor quality and outdated data base. Conventional social science methods cannot cope with the enormous developments and the tremendous growth as they are too labor-, as well as too time- and too cost-intensive. On the other hand, the physical discipline of remote sensing does not allow for direct conclusions on social parameters out of remote sensing images. The prime objective of this research is therefore the development of an integrative approach − bridging remote sensing and social analysis – in order to derive useful information about the living conditions in this specific case of the mega city Delhi and its inhabitants. Hence, this work is established in the overlapping range of the research topics remote sensing, urban areas and social science. Delhi, as India’s fast growing capital, meanwhile with almost 25 million residents the second largest city of the world, represents a prime example of a mega city. Since the second half of the 20th century, Delhi has been transformed from a modest town with mainly administrative and trade-related functions to a complex metropolis with a steep socio-economic gradient. The quality and amount of administrative and socio-economic data are poor and the knowledge about the circumstances of Delhi’s residents is correspondingly insufficient and outdated. Delhi represents therefore a perfectly suited study area for this research. In order to gather information about the living conditions within the different settlement types a methodology was developed and conducted to analyze the urban environment of the mega city Delhi. To identify different settlement types within the urban area, regarding the complex and heterogeneous appearance of the Delhi area, a semi-automated, object-oriented classification approach, based on segmentation derived image objects, was implemented. As the complete conceptual framework of this research, the classification methodology was developed based on a smaller representative training area at first and applied to larger test sites within Delhi afterwards. The object-oriented classification of VHR satellite imagery of the QuickBird sensor allowed for the identification of five different urban land cover classes within the municipal area of Delhi. In the focus of the image analysis is yet the identification of different settlement types and amongst these of informal settlements in particular. The results presented within this study demonstrate, that, based on density classes, the developed methodology is suitable to identify different settlement types and to detect informal settlements which are mega urban risk areas and thus potential residential zones of vulnerable population groups. The remote sensing derived land cover maps form the foundation for the integrative analysis concept and deliver there¬fore the general basis for the derivation of social attributes out of remote sensing data. For this purpose settlement characteristics (e.g., area of the settlement, average building size, and number of houses) are estimated from the classified QuickBird data and used to derive spatial information about the population distribution. In a next step, the derived information is combined with in-situ information on socio-economic conditions (e.g., family size, mean water consumption per capita/family) extracted from georeferenced questionnaires conducted during two field trips in Delhi. This combined data is used to characterize a given settlement type in terms of specific population and water related variables (e.g., population density, total water consumption). With this integrative methodology a catalogue can be compiled, comprising the living conditions of Delhi’s inhabitants living in specific settlement structures – and this in a quick, large-scaled, cost effective, by random or regularly repeatable way with a relatively small required data basis.The combined application of remotely sensed imagery and socio-economic data allows for the mapping, capturing and characterizing the socio-economic structures and dynamics within the mega city of Delhi, as well as it establishes a basis for the monitoring of the mega city of Delhi or certain areas within the city respectively by remote sensing. The opportunity to capture the condition of a mega city and to monitor its development in general enables the persons in charge to identify unbeneficial trends and to intervene accordingly from an urban planning perspective and to countersteer against a non-adequate supply of the inhabitants of different urban districts, primarily of those of informal settlements. This study is understood to be a first step to the development of methods which will help to identify and understand the different forms, actors and processes of urbanization in mega cities. It could support a more proactive and sustainable urban planning and land management – which in turn will increase the importance of urban remote sensing techniques. In this regard, the most obvious and direct beneficiaries are on the one hand the governmental agencies and urban planners and on the other hand, and which is possibly the most important goal, the inhabitants of the affected areas, whose living conditions can be monitored and improved as required. Only if the urban monitoring is quickly, inexpensively and easily available, it will be accepted and applied by the authorities, which in turn enables for the poorest to get the support they need. All in all, the listed benefits are very convincing and corroborate the combined use of remotely sensed and socio-economic data in mega city research.