In 2011, the global population passed the 7 billion mark, and more than half of the population is living in urban areas. Between 2011 and 2050, the world population is expected to increase by 2.3 billion, increasing to 9.3 billion. At the same time, the urban population is projected to gain 2.6 billion, passing from 3.6 billion in 2011 to 6.3 billion in 2050. The population growth in urban areas will be concentrated in the cities and towns of the less-developed countries. Asia, in particular, is projected to see its urban population increase by 1.4 billion, Africa by 0.9 billion, and Latin America and the Caribbean by 0.2 billion [1]. Population growth is therefore becoming largely an urban phenomenon concentrated in the developing world. Much of this growth is taking place as informal settlements in the urban fringes or disaster prone areas (e.g. slopes or floodplains), where people are living in substandard dwellings without access to drinking water, sanitation, health and education facilities.

The figures alone are alarming enough to understand that we are facing major challenges to manage the development of urban areas in a sustainable way. A central issue in this respect is the availability of up-to-date information on the extent and quality of the urban settlement. Urban planners need to know where new settlements are growing and how many people have to be supported. However, in less-developed countries, such information is largely unavailable or outdated. Cities are often growing at a pace that cannot be controlled by the local or regional mapping agencies.

Census data are the most common source of population information, but the quality, coverage and time span between census records is often poor, particularly in the fastest-growing parts of the world. Remote sensing of dwellings is increasingly used to provide detailed, up-to-date information. However, these remotely sensed datasets are mostly limited to local studies. For example, Taubenb̦ck et al. [2] monitored the growth of different megacities using space-borne Earth observation data. Baud et al. [3] assessed the status of sub-standard housing in Delhi. Kemper et al. [4] counted the dwellings in a camp for internally displaced persons (IDP’s) in Darfur. Ehrlich et al. [5] used remote sensing to assess the exposure of buildings in the context of earthquake vulnerability.

There are some global datasets derived from remote sensing that are applicable to urban studies [6], but they do not provide the necessary detail for many applications. Figure 2 shows the city of Harare (Zimbabwe) as it is mapped by the MODIS satellite, with 500 meter pixel size. The gridded global population provided by the LandScan data set [7] provides more detail in particular in the hinterland (Figure 3) by combining census and land use information. However, the spatial resolution of 30 arc-seconds (approximately 1 kilometer) is even coarser.

Although high resolution (HR, < 10 meter spatial resolution) and even very high resolution (VHR, < 1 meter) imagery with an almost global coverage are available, no consistent global coverage of settlements derived from those datasets exists. There are two main causes for the absence of a consistent global layer with higher spatial resolution. First, the availability of HR/VHR satellite data is limited since most HR/VHR satellite missions are operated on a commercial basis and consequently complete global coverage is costly. The only relevant exception is the CBERS-2B platform releasing 2.5 meter resolution panchromatic imagery with a very open data sharing policy in Brazil.

Second, and perhaps more importantly, to date no system has demonstrated the capacity to automatically extract global information layers about human settlement from HR/VHR satellite data with the necessary accuracy.

To alleviate this lack of information, the Joint Research Centre (JRC) of the European Commission has recently developed an automated image information query system to produce a Global Human Settlement Layer (GHSL) from high-resolution optical imagery. This system is able to ingest high and very-high spatial resolution imagery from heterogeneous sources even with low quality (e.g. lacking metadata, radiometric and geometric distortions). It provides a globally-consistent settlement layer at different scales [8]. Figure 4 shows the same extent as the previous figures of Harare, Zimbabwe. The GHSL output shows a much higher detail and a much larger, more realistic extent compared to the MODIS and LandScan datasets (Figure 5 ).

The GHSL is quality controlled and validated against visually collected reference data and global datasets. It has an average classification accuracy of 91.5 percent (plus or minus 10 percent) compared to the visually collected reference information. A limitation of the GHSL is that it is not yet a global dataset. So far, only 24 million square kilometers (16 percent of the global landmass) have been processed, covering parts of Europe, Latin America, Africa and Asia (Figure 6). Nevertheless, according to the LandScan dataset, this area is inhabited by approximately 1.3 billion people (19 percent of the global population).

In addition, the German Aerospace Center (DLR) is working towards a Global Urban Footprint (GUF) based on the TanDEM-X mission radar data [9]. Given the very different sensor characteristics of both systems (optical vs. radar), it is expected that the layers will provide complementary information that will significantly increase the knowledge about the presence of mankind on the planet in the near future.

Urbanization is inevitable, and we have to be ready to face this challenge and guide it in the right direction. Urban planning should include a sustainable, inclusive dimension that also benefits all citizens. Remote sensing will play an increasingly important role in this area and the frequently updated information also will be useful for other applications like crisis management, disaster risk reduction and population monitoring.

Thomas Kemper, who has a Ph.D. in geosciences, is scientific officer at the Joint Research Centre (JRC) of the European Commission in Ispra, Italy. He is part of the Global Security and Crisis Management Unit, where he provides scientific support to European Union (EU) policies enhancing the EU’s prevention, preparedness and response capabilities to natural and humanitarian disasters, conflict and other risks.

References

[1] U. N. Department of Economic & Social Affairs Population Division, ‰ÛÏWorld urbanization prospects: the 2011 revision,‰Û April 2012. [Online]. Available: http://esa.un.org/unpd/wup/pdf/WUP2011_Highlights.pdf

[2] H. Taubenb̦ck, T. Esch, A. Felbier, M. Wiesner, A. Roth, and S. Dech, ‰ÛÏMonitoring urbanization in mega cities from space,‰Û Remote Sensing of Environment, vol. 117, p. 162‰ÛÒ176, 2012.

[3] I. Baud, M. Kuffer, K. Pfeffer, and R. Sliuzas, ‰ÛÏUnderstanding heterogeneity in metropolitan India: The added value of remote sensing data for analyzing substandard residential areas,‰Û International Journal of Applied Earth Observation and Geoinformation: JAG, vol. 12, no. 5, pp. 359‰ÛÒ374, 2010.

[4] T. Kemper, M. Jenerowicz, M. Pesaresi, and P. Soille, ‰ÛÏEnumeration of dwellings in Darfur camps from geoeye-1 satellite images using mathematical morphology,‰Û IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 4, no. 1, pp. 8‰ÛÒ15, 2011.

[5] D. Ehrlich, T. Kemper, X. Blaes, and P. Soille, ‰ÛÏExtracting building stock information from optical satellite imagery for mapping earthquake exposure and its vulnerability,‰Û Natural Hazards, 2012.

[6] A. Schneider, M. Friedl, and D. Potere, ‰ÛÏMonitoring urban areas globally using MODIS 500m data: new methods based on urban ecoregions,‰Û Remote Sensing of Environment, vol. 114, no. 8, pp. 1733‰ÛÒ1746, Aug. 2010.

[7] J. Dobson, E. Bright, P. Coleman, R. Durfee, and B. Worley, ‰ÛÏLandScan: A global population database for estimating populations at risk,‰Û Photogrammetric Engineering and Remote Sensing, vol. 66, no. 7, pp. 849‰ÛÒ857, 2000.

[8] M. Pesaresi, X. Blaes, D. Ehrlich, S. Ferri, L. Gueguen, F. Haag, M. Halkia, J. Heinzel, M. Kauffmann, T. Kemper, G. Ouzounis, M. Scavazzon, P. Soille, V. Syrris, and L. Zanchetta, ‰ÛÏA Global Human Settlement Layer from optical high resolution imagery: concept and first results,‰Û EUR ‰ÛÒ Scientific and Technical Research series ‰ÛÒ ISSN 1831-9424, 2013. [online] Available: doi:10.2788/73897.

[9] H. Taubenb̦ck, A. Roth, T. Esch, A. Felbier, A. MÌ_ller, and S. Dech, ‰ÛÏThe vision of mapping the global urban footprint using the terrasar-x and tandem-x mission,‰Û in Urban and Regional Data Management. CRC Press, Sep. 2011, pp. 243‰ÛÒ251. [Online]. Available: http://dx.doi.org/10.1201/b11647-25