| dc.description.abstract | Several studies have reported an increase In air pollution in urban centers especially in
developing co untries, leading to cardio-respiratory diseases. According to World Health
Organisation (WHO), air pollution is estimated to cause about two million premature deaths
worldwide annually.
This study sought to assess the impacts of climate change on particulate matter (PM) levels and
distribution and the concomitant human health effects in Kenya. The data used included
temperature, rainfall, aerosol optical depth and respiratory-related diseases. These data sets were
collected and analysed for four selected areas within the entire study domain. Short-Cut Bartlett
test was employed in the data homogeneity test. Temporal distribution of these datasets was
determined using time series analysis. The spatial variation of the same data was obtained by the
use of Surfer software (version 8). Pearson's correlation analysis method was used to establish
the relationship among the variables. The combined effect of weather and PM on human
respiratory health was assessed through the use of multivariate linear regression analysis.
From time series analysis seasonality is evident in all the mean monthly datasets. Highest well
defined peak in aerosol optical depth (AOO) was observed during June-June-August season due
to thermal inversion that results in less atmospheric mixing, increased fuel combustion for
domestic heating. Over all the stations except Garissa AOD loading depict significant decreasing
trend. Generally, a peak in the number of respiratory cases was observed during March-April-
May season over Nyeri due to pollen. The observed seasonality implied that weather has a direct
effect on health outcome. Hospital admission was higher among male cohort at about 82% due to
increased occupational exposure. Generally, more hospital admission and mortality was reported
in the under five due to their vulnerability.
From the analysis of minimum temperature, well-defined urban heat island is observed over
Eastleigh due city's population build-up. The northern and north eastern parts of the country are
centres of heavy AOO loadings due to geographical effects i.e. semi-arid dust. A peak in the
number of respiratory cases-was reported over Nyeri due to heavy AOD loading as a result of
tree and grass pollen, high population, increased occupational exposure, rural solid fuel use and
high reporting rates. A minimum was recorded over Garissa even with cold weather and heavy
dust loading due to relatively low urban population, less windy conditions, less dust suspension,
natural immunity to pathogens and low reporting rate.
The nature of correlation obtained between variables varied in both strength and direction.
Temperature showed a stronger correlation with AOD compared to rainfall over Nairobi. The
negative relationship was attributed to enhanced pollutant mixing due to instability. Rainfall
depicted a weaker but significant correlation. This implied that atmospheric pollutant mixing is
dominant over rain/wash-out AOD removal mechanism. Although the individual correlations
were not particularly high, they illustrate that one variable is affected differently by different
parameters over different locations due to varying seasonal conditions.
The present study showed weather and PM accounts for over 84% variance in the number of
reported respiratory cases over Nairobi and Nyeri. The least variation in respiratory incidences of
about 53% is explained over Mombasa. Even though the slope varied in both strength and
direction, AOD had the most strong and positive influence. However the weather variables made
a statistically significant contribution. The fact that temperature and rainfall explained the pattern
of respiratory diseases showed that thermal environment has significant influence on the
occurrence of respiratory diseases. Moreover, the seasonal variation in the occurrence of
respiratory diseases considered in this study confirmed the role of seasonal weather conditions in
the occurrence of respiratory diseases. | en_US |
