A comparison of greenspace loss and urban expansion over time in London and Nairobi

by Mary Langsdale, PhD student in the Department of Geography, KCL

On the 2^nd March at the Earth Observation and Environmental Sensing (EOES) Research Hub meeting, I presented my recent research into urban land use and land cover changes (LULCC) in London and Nairobi, with a particular focus on urban greenspace. The work had been conducted for my MSc Climate Change dissertation at King’s and motivated by the increasing awareness around the important ecosystem services greenspace provides in urban areas. These include (but are not limited to) mitigating the urban heat island (UHI) effect, improvements to local biodiversity and stormwater retention.

To analyse LULCC, I classified Landsat imagery of both locations between 1988 and 2016 using a supervised classifier (Random Forest). Results for London and Nairobi are shown in Figures 1 and 2 respectively. In London, my results suggested densification and industrial development between 1990 and 2016 accompanied by an overall increase in urban areas by 4% and a corresponding decrease in vegetated areas by 4% in that same period.

MLblog — Figure 1. Supervised classification of London in a)1990, b) 2002 and c) 2016 using the Random Forest Classifier.

In Nairobi, urban expansion and subsequent loss of greenspace was evident, particularly to the north, east and west. Deforestation was identified through the classification, in areas corresponding to local deforestation prevention campaigns. This led to an overall increase of 22% in urban areas and corresponding loss of greenspace.

Figure 2: Supervised classifications of Nairobi in a) 1988, b) 2002 and c) 2016, performed using the Random Forest classifier.

Comparisons between the two cities showed strong differences between spatial and temporal patterns of urban growth and loss of greenspace. In London, LULCC occurred primarily through densification of existing urban areas and development of inner-city brownfield sites. The rate of urban growth in London decreased notably in the latter half of the period, following introduction of tighter planning policies in the year 2000. In Nairobi however, both densification and urban expansion occurred throughout the period. Rates of urban growth and loss of greenspace did drop in the latter half of the period but still remained high, with the latter correlating with population increase (R=0.83).

These differences were attributed to planning policies in each city. In London, urban expansion and development was tightly controlled, with an enforced Green Belt and active pursuit of densification through the London Plan. In contrast, there has been no master development plan in Nairobi since 1948 despite rapid population growth and much of the urban expansion in Nairobi has manifested in informal housing areas and slums with little public services provisions. These areas are particularly vulnerable to the effects of climate change, particularly flooding, as was seen in devastating floods during 2016.

Comparisons with recent greenspace estimates of both cities suggested this study had under-estimated greenspace in London by around 14% (GiGL, 2015). This figure corresponded with GiGL’s (2015) estimated contribution from private gardens and was thus attributed to limited spatial resolution of the Landsat imagery. This highlighted the importance of high resolution imagery for LULCC studies and studies into greenspace in urban areas, which are highly heterogeneous.

Reference:

Greenspace Information for Greater London (GiGL) (2015) Key London Figures [Online] Available at: http://www.gigl.org.uk/our-data-holdings/keyfigures [last accessed 20/08/16]