Inter annual variability of onset and cessation dates of the long rains in Kenya
Abstract
The overall objective of the study was to determine the inter-annual variability of onset
and cessation dates of the long rains season in Kenya.
The specific objectives that were addressed to achieve the overall objective
included determination of the onset and cessation dates of long rains (MAM) during
the period 1961-2001, determination of the seasonal rainfall performance during early,
normal and late onset years, identification of the anomalous wet and dry years and
examination of the circulation pattern associated with early, normal and late onset of
the long rains.
The data used in the study were daily rainfall, monthly rainfall, pentad rainfall,
and wind data. The rainfall data was obtained from Kenya Meteorological Department
while the wind data was obtained from the National Centre for Environmental
Prediction (NCEP) website.
In order to achieve the overall and specific objectives, the data was subjected
to various analyses including quality control, correlation analysis, and principle
component analysis. The core methodology lay in determination of onset, cessation
and duration of the rainfall season using time series analysis, mass curve analysis
and the quantitative determination of onset/cessation dates using instat program.
Wind analysis for early, normal and late onset periods was performed.
Stations located over the western and southern parts of Kenya received earlier
onset than others. This was attributed to the fact that the rainfall during this season is
associated with both the northward progression of the zonal arm of the ITCZ and the
eastward extension of the meridional arm.
Although onset progresses across the region, the mean onset date for the whole
region is the 74th day of the year that is 14th of March. The average standard
deviation, i.e. the interannual variability of the onset, is 24 days.
The zone with the highest onset variability is zone 1 (Lodwar) with a standard
deviation of 33 days and zone 10 (Narok) which has the lowest standard deviation of
15 days has the lowest variability.
The study showed that in general the onset month is March while the cessation
month is May. However, early onset will mostly occur in February and late cessation is
June. It was quite evident that for years with early onset, rainfall withdrew late while for
years of late onset, withdrew early. Thus on average, the years with early /Iate onset
had longer/shorter rainfall duration.
The analysis of the wind system in reference to early/late onset years indicated
that during early onset there were stronger easterlies in the upper troposphere with
relatively weak easterlies or westerlies in the mid/lower troposphere. Most of the late
onset will be recovery to the normal flow pattern after experiencing a time delay in one
phase. The high pressure in southern Africa helps in providing a mechanism for the
advection of moisture towards the equator. In some cases westerlies develop from the
incursions of the Congo air mass which are forced eastwards by high pressure over
the eastern Atlantic Ocean.
The study has demonstrated that there is both spatial and temporal variability
of onset and cessation dates over Kenya. It may be concluded that when the onset of
the rains is early (negative onset date anomaly), the rainfall tend to be heavier than
normal.
These findings may be used to develop a model for monitoring and forecasting
of evolutions of anomalous rainfall over the region including the extreme events like
droughts and floods. Such efforts would contribute enormously to regional/national
early warning and preparedness fOL extreme events and food security. This would
therefore contribute significantly to the effective management and sustainable
development of the regional/national social-economic activities which are heavily raindependent
Citation
Master of Science in MeteorologySponsorhip
University of NairobiPublisher
Department of Meteorology University of Nairobi