Modelling The Spatial Relationship Between Built-Up Volumes And Surface Urban Heat Islands In Upper Hill, Nairobi City County, Kenya

ABSTRACT

Urban heat island refers to the thermal temperature differences between rural and urban areas. The factor that has been attributed to this phenomenon is urbanization, where natural land cover is replaced by concrete and other man-made impervious surfaces. Causes and effects of heat islands and urban climates are varied. The urbanization process dislocates the natural solar and hydrologic balances by transforming the radiative, thermal, moisture and aerodynamic characteristics of a region. Lack of indicators and frameworks on three dimensional development in urban environment possess a challenge to urban authorities when formulating policies on thermal comfort, pollution, assessing health-related risks such as heat stress, respiratory diseases and vector-borne diseases. Therefore, this study aims at modelling the relationship between built-up morphology and topographical features with land surface temperatures using time series data in Upper Hill, Nairobi. Upper Hill, Nairobi, Kenya has rapidly developed over the years, transforming to commercial, office blocks and high-rise flats from low density residential area. Changes in zoning regulations is informed by businesses choosing to locate their offices away from the Central Business District because of traffic jams, inadequate parking space and high costs of renting office space. The research uses case study and correlational research design. Random sampling is used to collect ground control points and verify analysed data. Land surface temperature, land cover, NDVI and albedo is analysed for the years 1986, 1995, 2000, 2005, 2011 and 2017. Road and building information is extracted from stereo-aerial photographs for the period 1978, 1998 and 2017, which is utilized to obtain building height, ground coverage ratio and built-up volume density index. The spatial relationship of land surface temperature with built-up morphology is modelled using geographically weighted regression and ordinary linear regression. Findings show that land surface temperature is influenced by type of land cover and albedo; with sparse grassland having an albedo of 0.18 and mean surface temperature of 28 0C while water has an albedo of 0.09 and mean surface temperature of 25 0C during the day. At night, water and sparse grassland have mean surface temperatures of 18.62 0C and 18 0C respectively. Contribution index of built-up and forest areas has reduced with increased impervious surfaces. Mean built-up volume density and ground coverage ratio in Upper Hill in 1978, 1998 and 2017 is low density while mean building heights are low-rise. Geographically weighted regression model findings indicate that building height in 1978 has an r2 of 70% and built-up volume density in 1998 and 2017 have r2 of 72% therefore having a great relationship with land surface temperature. Combined modelling tools capture the physical dynamics and interaction built-up forms have with land surface temperature. Built-up volume densities should be integrated in the development control frameworks, with building heights being an important variable in urban development. It further recommends that green design strategies be made an integral part of urban development with percentage vegetation cover specified in the zoning policy as surface urban heat islands is a spatial temporal phenomena.