A Time Series Analysis Of Rainfall Patterns In Kenya
There has been a need to study and understand the dynamic rainfall pattern
which is alarming and more so after experiencing little rain; not forgetting the
immense water shortage in January this year which posed a challenge in food
security. In this statistical data analysis, we have presented a study of rainfall
data for observations made over 24 years (1991-2015) in an attempt to capture
the intrinsic features in the rainfall pattern and forecast future uncertain values
on the quantity of rainfall expected. Time series analysis has been used as a
tool for any study of a behavior or process observed over a prescribed period
of time and this paper re
ects how it can be applied in modeling and eective
forecasting of the rainfall pattern in Kenya. Dierent techniques in data analysis
have been employed in studying the components in the data.These include the
trend,seasonality, cyclical component and the error term. The results are further
explained including a 5 year forecast and recommendation.