This PhD thesis deals with urbanization-induced changes in land use/cover (LULC), associated local-level warming and the role of vegetation in mitigating the local thermal climate change in cities of Ethiopia. The thesis consists of four research papers that address these environmental changes and consequent problems. Through these papers, the project contributes to: 1) the science of remote sensing and Geographic Information System (GIS) by introducing methods that improve LULC classification accuracies, and an improved method of spatial thermal climate analyses, 2) better understanding of urban warming in tropical dry climates and effects of different urban vegetation in mitigating local warming, and 3) knowledge of human bioclimatology in a tropical warm climate by examining influences of urban thermal environment on workplace comfort and wellbeing.
The changes in urban LULC as well as anthropogenic heat releases in cities, often lead to a local thermal climatic change described as an Urban Heat Island (UHI) effect. In tropical warm climate environments, additional warming caused by the UHI effect may intensify the already warm climate and may have detrimental effects on human health, socio-economic activities and general wellbeing. Worsening the UHI effects, the rapid urbanization in developing countries seems to ignore the immediate and long term impacts of the changes LULC on human and general ecosystem functioning. This could be evidenced by the haphazard urban expansion in many cities of developing countries, intensifying areas of impervious surfaces and shrinking of green spaces within cities and the surrounding landscape (Jorgenson et al. 2010; Kumar et al. 2010).
Despite a large and increasing literature on the problems of urban environmental changes, a limited amount of such research has addressed thermal consequences of the ongoing rapid urbanization in African contexts. The magnitude of UHI and its impacts on human comfort and wellbeing, as well as effectiveness of urban vegetation in mitigating the UHI effects in the context of tropical Africa, are rarely studied. This project, therefore, contributes to the research gaps by considering examples from two cities in Ethiopia, namely Addis Ababa and Adama. The major sources of data used in this study were remotely sensed multi-temporal digital imagery from Landsat TM and ETM+ sensors, ground surveying of LULC, measurements of air temperature and humidity, and questionnaire surveying. Remote sensing and GIS techniques were applied to analyze urbanization-induced dynamics of LULC and associated urban warming in five-year intervals between 1985 and 2010. LULC dynamics were analyzed applying post-classification change analysis using the Landsat imagery.
In relation to the use of Landsat data for LULC mapping, problems related to classification accuracy and consistency were detected during the initial phase of the project. Hence, three new indices were developed to improve the LULC classification accuracy and consistency: water index, built-up index, and soil index. Paper I provides details of accuracy improvement achieved in surface water mapping and the results showed that the new water index consistently improved classification accuracy compared to two commonly used classifiers. The use of the new indices has improved the accuracy and consistency of urban LULC classification considerably, and a comparative assessment of improvements is presented in Paper II.
Results from LULC analysis in Paper II showed that rapid urbanization in Addis Ababa has led to enormous LULC changes during the period from 1985 to 2010. An increasing trend of thermal aggregation was observed during this period and the thermal intensity variation was shown to be attributable to urbanization-induced changes in the LULC. Intra-urban surface heat intensity differences of up to 5.3ºC were detected through spatial aggregation analyses. Unlike many surface UHI studies, results from this study indicated no clear pattern of urban environments being warmer than the surrounding rural areas. It was, therefore, concluded that comparison across broad urbanrural land use dichotomies may not be an efficient indicator for quantification of surface UHI in tropical dry conditions like those of Addis Ababa. More meaningful analysis of thermal impacts of urbanization could be achieved by analyzing spatial patterns of surface heat intensities and by detailed examination of relationships between temperature and urban surface properties.
Cooling effects of urban parks, as an option to mitigate UHI, was examined in Paper III. The data sources were ground-based automated air temperature sensors and thermal remote sensing. The results showed that tree species, canopy cover, size and shape of parks were the main determinants of cooling effects. Compact parks and large-sized parks showed higher thermal contrast with the surrounding non-park environment. Species with highest and least cooling effect were, respectively, Eucalyptus and Grevillea.
In Paper IV, influence of urban environment on human thermal comfort at workplaces in Adama was explored. The role of tree vegetation in improving workplace thermal comfort was also studied in this paper. Moreover, spatial pattern and intra-urban variation in surface heat intensity was assessed. The results revealed that there was large intra-urban variation in thermal comfort. A considerable number of working hours in indoor and outdoor workplaces were characterized by temperature that exceeded the human comfort range. Thermal comfort in outdoor and indoor workplaces was improved by tree vegetation; particularly in outdoor workplaces, the influence of tree vegetation was large. The urban warming was also shown to have negative impacts on work activities by reducing effective working hours as people take extra breaks due to thermal stress.
In summary, this PhD project contributes to the science of urban environmental studies by integrating data from a wide range of sources and using diverse analytical approaches. The project has demonstrated that urbanization-induced changes in LULC contributed significantly to local urban warming, which was shown to be negatively affect human comfort and wellbeing at workplaces. By providing important findings on the immediate detrimental thermal effects and indications for long term environmental impacts of urbanization, the results of this PhD project offer an improved basis for planning of ecologically healthy cities in Ethiopia and countries with similar conditions.