 |
| Titel |
Quantifying Urban Texture in Nairobi, Kenya and its Implications for
Understanding Natural Hazard Impact |
| VerfasserIn |
Faith E. Taylor, Bruce D. Malamud, James D. A. Millington |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250134198
|
| Publikation (Nr.) |
 EGU/EGU2016-14897.pdf |
|
|
|
|
|
| Zusammenfassung |
| The configuration of infrastructure networks such as roads, drainage and power lines can both
affect and be affected by natural hazards such as earthquakes, intense rain, wildfires and
extreme temperatures. In this paper, we present and compare two methods to quantify urban
topology on approximate scales of 0.0005 km2 to 10 km2 and create classifications of
different ‘urban textures’ that relate to risk of natural hazard impact in an area. The methods
we use focus on applicability in urban developing country settings, where access to
high resolution and high quality data may be difficult. We use the city of Nairobi,
Kenya to trial these methods. Nairobi has a population >3 million, and is a mix of
informal settlements, residential and commercial development. The city and its
immediate surroundings are subject to a variety of natural hazards such as floods,
landslides, fires, drought, hail, heavy wind and extreme temperatures; all of these
hazards can occur singly, but also have the potential for one to trigger another, thus
providing a ‘cascade’ of hazards, or for two of the hazards to occur spatially and
temporally near each other and interact. We use two measures of urban texture: (i) Street
block textures, (ii) Google Earth land cover textures. Street block textures builds
on the methodology of Louf and Barthelemy (2014) and uses Open Street Map
data to analyse the shape, size, complexity and pattern of individual blocks of land
created by fully enclosed loops of the major and minor road network of Nairobi.
We find >4000 of these blocks ranging in size from approximately 0.0005 km2 to
10 km2, with approximately 5 classifications of urban texture. Google Earth land
cover texture is a visual classification of homogeneous parcels of land performed
in Google Earth using high-resolution airborne imagery and a qualitative criteria
for each land cover type. Using the Google Earth land cover texture method, we
identify >40 ‘urban textures’ based on visual characteristics such as colour, texture,
shadow and setting and have created a clear criteria for classifying an area based
on its visual characteristics. These two methods for classifying urban texture in
Nairobi are compared in a GIS and in the field to investigate whether there is a link
between the visual appearance of an area and its network topology. From these urban
textures, we may start to identify areas where (a) urban texture types may indicate a
relative propensity to certain hazards and their interactions and (b) urban texture
types that may increase or decrease the impact of a hazard that occurs in that area. |
|
|
|
|
|
|