Journal of Agricultural Economics and Rural Development AJAERD

Vol. 4(1), pp. 326-334, January, 2018. © www.premierpublishers.org. ISSN: XXXX-XXXX

Research Article

Long-term observed Precipitation Trends in Arid and

Semi-arid Lands, Baringo County, Kenya
1,2Gideon Muriithi*, 2Daniel Olago, 2Gilbert Ouma, 2Silas Oriaso
1*
Beef Research Institute-Lanet, Kenya Agricultural and Livestock Research Organization (KALRO)
PO Box 3840-20100 Nakuru, Kenya
2
Institute for Climate Change and Adaptation, University of Nairobi, PO Box 30197-00100, Nairobi, Kenya

The research was conducted to validate the pastoralists’ and agro-pastoralists’ claim that there
has been an increasingly variable and changing climate in the study area. The station average
and Theissen polygon methods were used to estimate the mean areal precipitation of the small
(Mogotio and Baringo South Sub-counties) and the large area (Baringo County), respectively. The
aim of the current study is to analyse rainfall time series over long term observed precipitation
and a wide area, detecting potential trends and assessing their significance. Monthly precipitation
data for the period 1974-2003 from six weather stations, located mainly in Mogotio and Baringo
South sub-counties and covering 3906km2 were used in the analysis. The data were quality
controlled to ensure no missing data and any inconsistencies. Linear regression analysis of the
database highlighted that; the trends were predominantly negative, both where the average and
Theissen polygon methods were used and over the whole reference period. The negative trends
are not significant. This finding implies that the study area has been suffering a precipitation
decrease especially in the period under review.

Keywords: climate change, Theissen Polygon, Weather Stations, Variability

INTRODUCTION

Erratic rainfalls and increasing temperatures are the most
serious global problems associated with climate change.
They are considered to be some of the most serious
threats to sustainable development with adverse impact on
the environment, human health, food security, economic
activities, natural resources and physical infrastructure (Jiri
et al., 2014). The arid and semi-arid lands (ASALs) are
hardest hit by the above-mentioned problem. It accounts
for more than 40% of the world’s land area and are home
to over two billion people, of whom 325million are in Africa.
Yet they are among the regions in the world where climate
change impacts on ecosystems, livelihoods and human
health are potentially the greatest (IPCC, 2014) as cited by
(Nyasimi, 2014). The main challenge that confronts
farmers in arid and semi-arid areas of Sub-Saharan Africa
is managing unreliable rainfall (Tumbo et al., 2010). In
Kenya’s ASALs, for example, more than three million
pastoralist households are regularly hit by drought, costing
Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya
the economy an estimated $12.1 billion in 2008–2011
(ILRI, 2014). The risks expose the ASALs in the region to
the challenge of high costs of inputs occasioned by
unreliable, highly variable, and scarce rainfall for their
livestock and crop production (Tol et al.,1998; Bhatt et al.,
2006; Cooper et al., 2012).
Netherlands Development Organization-SNV (2012)
reported that climate change and variability in the ASAL is
evidenced by more extreme oscillations between dry and
wet periods, and between the availability of natural
resources in different areas, leading to greater variability
in livestock production.
*Corresponding author: Gideon Muriithi, Beef Research
Institute-Lanet, Kenya Agricultural and Livestock
Research Organization (KALRO) PO Box 3840-20100
Nakuru, Kenya. E-mail: murithigm@gmail.com
Muriithi et al. 327
Therefore, climate change and variability are increasingly
becoming serious problems to pastoralists. Kibria et al.
(2017) reported that pastoralist communities live in ASALs
where their livelihoodsare highly dependent on natural
resources which are sensitive to changes in climatic
conditions. The aforementioned livelihood makes
pastoralists more vulnerable to climate change and
variability. Saitabau (2014) noted that pastoralists use
indigenous knowledge to model weather events and
livelihood management. However, indigenous knowledge
systems are now failing to predict the weather events; for
example, weather variations have become so phenomenal
that droughtswhichused to occur every ten years now
occur every two years or less, and the trend continues to
worsen. Annual rainfall is now more erratic than in the past
and figures continue to decline, while people experience
warmer dry months than in the past. Thus, the current
trend of changing climate poses a threat to livelihoods. The
influence of seasonal weather patterns on crop and
livestock production suggests that agricultural applications
of seasonal weather predictions may be highly valuable to
the society (Jones, 2000). Climate information (including
observations, research, predictions, and projections) plays
a central role in both adaptation and mitigation of climate
change (Zillman, 2009).
According to Chin (2007), there are three (3) common
ways to estimate mean areal precipitation: namely the
station average method, Theissen polygon method and
isohy et al method. Each method has its advantage and
constraint. The Theissen method involved drawing
perpendicular bisectors (Figure 2) between two nearby
rain gauge points, converging data point with two other
bisectors to form a polygon for each representative point
(Linsley et al., 2000). The method provides weights
proportional to the size of polygon area, pivoted by each
rain gauge. The method is widely used because of its
practicability and it is less time consuming with relatively
high accurate estimates (Damant et al., 1983). Vaes et al.
(2004) reports that the station average method is
recommended for very small catchments, of up to one (1)
km radius evenly installed with rain gauges. Further, Chin
(2007) recommends Theissen method for large area
coverage.
The objective addressed in this paper is to validate the
pastoralists and agro-pastoralists claim that there has
been an increasingly variable and changing rainfall in the
study area. To achieve this objective rainfall time series
over a long-time interval and wide area was analyzed,
detecting potential trends and assessing their significance.
MATERIALS AND METHODS
The study was conducted in Baringo County, Kenya, and
focused on the sub-counties of Mogotio and Baringo South
Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya
(Figure 1). The monthly precipitation data for the years
1974-2003 was sourced from the Kenya Meteorological
Department (KMD).Specifically, the data came from six
weather stations namely; Snake farm, Chemususu,
Kimose, Lake Bogoria, Perkerra and Talai covering
3906km2(Table 1).The percentage of missing rainfall data
in the aforementioned weather stations was less than 10%
and the data spread over 20 years, hence met the
requirements of World Meteorological Organization with
regard to climatological analysis. The data were quality
controlled to ensure no missing data or inconsistencies.
Radi et al. (2015) recommended imputation approaches to
fill in missing rainfall data. For this study multiple
imputations method were used to overcome
underestimation of standard errors and confidence
intervals typical of single imputation. The method
involvedestimation of missing rainfall data from the
observations of rainfall (rainfall data sets) at the same
station and period but different yearsand filling in of
themissing data with substituted values. The imputed data
sets were combined and average worked out. According
to Ochieng et al. (2017) the missing rainfall data PX was
estimated using the following formula:
𝑃𝑥 = 1/𝑛 ∑𝑖−𝑛
𝑖−0 𝑃𝑖 Equation (1)
Where:
n = the number of rainfall data sets
Pi = rainfall data from the ith data set
PX = missing rainfall data
For this study, the station averages the station average
and Theissen polygon methods were used. Apart from the
two aforementioned methods, annual precipitation for an
individual station was also analyzed. The average method
was applied for four weather station namely; Snake farm,
Kimose, Lake Bogoria and Perkerra all located in Mogotio
and Baringo South Sub-counties, covering an area of
1867km2.For the large area in this study two weather
stations from outside the study area namely; Chemususu
and Talai were added to the four weather stations used in
the small area. Therefore, rainfall data used in the study
came from a total of six weather stations, covering a
cumulative area of3906km 2. Equation 2was used to
estimate the Thiessen values from the Thiessen polygons.
Polygons for Baringo County were generated using
ArcGIS software (Table 1). The data on spatial annual
average rainfall from the five stations was estimated using
the Thiessen polygon method (weighted values) with the
Thiessen value using the Equation 2 below:
𝑃 𝐴 +𝑃 𝐴 +𝑃 𝐴 +𝑃 𝐴 +𝑃 𝐴
𝑃̅ = 1 1 2 2 3 3 4 4 5 5 = ∑5𝑖=1 Pi Ai Equation (2)
𝐴1 +𝐴2 +𝐴3 +𝐴4 +𝐴5
Where;
𝑃̅ is the weighted average
P’s are measurements
A’s are areas of each polygon (Table 1)
 J. Agric. Econ. Rural Devel. 328

Figure 1: Map of Kenya and the study area location, Baringo County. Mogotio and Baringo South are shown with Weather
stations (Source: Author 2017)

Statistical analysis of the data computed from both the RESULTS

station average as well as from Theissen method was
performed by using Microsoft Excel and Statistics Package
for Social Scientists (SPSS). The linear regression
analysis was used to generate the mean areal precipitation
trends, potential trends, their descriptive and assessing
their significance.
Table 1 shows the six weather stations whose long term
observed precipitation data were used in this study. It
further shows the area coverage of each station in square
kilometers. This data was used in the construction of the
Theissen polygons (Figure 2) in the process of computing
the Theissen values/weighted averages.

Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya
Muriithi et al. 329

Table 1: Area in Km2 covered by respective weather Figure 2 shows the distribution and location of seven
stations in Baringo County weather stations in Baringo County. The long -term
observed precipitation data for Kabimoi Agricultural
S/No Station Name Area in Kilometers
Holding could not meet the threshold for the analysis since
squared
1. Snake Farm - Lake Baringo 306.724
it lacked data for five years unlike the other six weather
2. Chemususu Forest Station 348.842 stations. It was therefore, omitted in any further process
3. Kimose Agricultural Holding Ground 889.485 and analysis. Figure 2, further demonstrates the bisected
4. Lake Bogoria National Reserve 880.314 perpendiculars, generated polygons and km 2 coverage of
5 Perkerra Agricultural Research 680.039 each polygon. The area covered by each polygon was
6 Talai Agricultural Station 800.355 used in Equation 2 for the computation of the weighted
averages/Theissen values, which were further used in
plotting the mean areal precipitation trends.

Figure 2: Map of Baringo County showing generated polygons and area each weather station is covering
(Source: Google map and Author 2016)

Mean Areal Precipitation Trend by use of Station
Average Method
Figure 3 shows the estimated mean areal precipitation
trend for four weather stations namely; Snake farm,
Kimose, Lake Bogoria and Perkerra for the period 1974-
2003 and covering 2757km2 in Mogotio and Baringo South
sub-counties. It further shows the highest and lowest
estimated mean areal precipitation of 1081.10mm in 1977
and 277.90mm in 1984 respectively. The range between
the highest and the lowest is 803.20mm, implying that
there has been a big variation in the amount of
precipitation in the years under review. Further analysis of
the long term observed precipitation data for the
aforementioned weather stations showed that 706.07mm
± 178.13mm, n = 30. This implies that the amount ofmean
areal precipitation observed and recorded was between
884.2mm and 527.13 mm within the span period of 30
years. This finding demonstrates the variability of the
precipitation since the standard deviation (SD) reported is
very high relative to the means.
From the trend equation, the predicted mean areal
precipitation for the year 2018 (15 years from 2003) was
707.28 mm. The linear regression line shows a negative
slope of 2.408mm per year. This finding indicates a
general decreasing trend in the mean areal precipitation.
The regression equation, R2=0.014 implies that the time
series explained only 1% of the mean areal precipitation.
Further analysis showed that, there was no significant
relationship (P>0.05) between the mean areal precipitation
and the time series in the study area.

Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya

Figure 3: Estimated mean areal precipitation trend by use of average method for four weather stations within the study area (Snake
Farm, Kimose Agricultural Holding Ground, Lake Bogoria National Reserve and Perkerra Agricultural Research)
Source: KMD Observed Precipitation data (2016)
Mean Areal Precipitation Trend by use of Theissen
polygon method
Figure 4 shows the estimated mean areal precipitation
trend for the six weather stations namely; Snake farm,
Kimose, Lake Bogoria, Chemususu, Talai and Perkerra for
the period 1974-2003 and coverage of 3905.76km 2in the
larger Baringo County. It further shows that the highest and
the lowest estimated mean areal precipitation was
1298.65mm in 1988 and 359.59mm in 1984, respectively.
The range between the highest and the lowest was
939.05mm, implying that there has been a big variation in
the amount of precipitation over the years under review.
Further analysis of the long term observed precipitation
data for the aforementioned weather stations showed
868.12 mm ± 207.56mm, (n = 30) clearly demonstrates the
variability of the precipitation since the SD reported was
very high relative to the means. The findings clearly
indicate that there has been variation in the amount of
precipitation experienced and eventually contributed to the
climate variability in the Baringo County.
From the trend equation, the predicted mean areal
precipitation for the year 2018 (15 years from 2003) was
expected to be 870.47 mm. The linear regression line
shows a negative slope of 4.72mm per year, indicating a
general decreasing trend in mean areal precipitation,
which suggests that the rainfall was decreasing in the
study area. For regression line equation, the R2 = 0.040,
indicates that the time series could explain only 4% of the
mean areal precipitation. When a further analysis was
conducted, the findings showed that there were no
significant relationships (P>0.05) between the mean areal
precipitation and time series in the study area.
Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya
J. Agric. Econ. Rural Devel. 330
Individual Annual Precipitation Trends
The annual rainfall trends for the individual five weather
stations namely; Chemususu, Perkerra, Kimose, Snake
Farm and Lake Bogoria were plotted in one graph
(Figure5). The Chemususu station trend recorded the
highest precipitation of 1716 mm in 1977 and lowest of 656
mm in 1984 relative to all the others under review. It further
showed 1139.16 mm ± 221.37mm, n = 30. The findings
imply that the SD is very high relative to the mean within
the span period of 30 years. Therefore, the annual
precipitation received within the period under review
(1974-2003) demonstrates a big variation. The linear
regression shows a negative slope of 5.393mm per year
indicating that the general amount of rainfall received in
the station reduced progressively. The trend yielded an R 2
= 0.046, indicating that the time series predicts only 4.6%
of the annual rainfall. There was no significant relationship
(P>0.05) between the mean areal precipitation and the
time series in the study area.
From the amount of precipitation observed in Kimose
weather station, a trend was generated as indicated in
Figure 5. From the trend, it is clear that there were
variations in the observed amount of precipitation
throughout the 30 years period (1974-2003). This scenario
is an indication that climate variability is tantamount to
climate change, since precipitation is a climatic parameter.
The highest annual rainfall observed in the station was
1341 in 1997 while the minimum was 365 mm in 1981.
From the annual rainfall observed from the station, the
annual rainfall trend from 1974-2003 showed 857.19mm ±
236.82mm, (n = 30). This shows that there has been a big
variation in the amount of rainfall received over the period
Muriithi et al. 331

Figure 4: Estimated mean areal precipitation trend by use of Theissen polygon method forsix weather stations, 1974-2003
Source: KMD Observed Precipitation data (2016)

Figure 5: Combined Individual Stations Annual Rainfall Trends from 1974-2003

Source: KMD Observed Precipitation data (2016)

in reference. This is because the SD is relatively high to decrease of the annual rainfall received. There was no
the mean. The linear regression line shows a negative significant relationship (P>0.05) between the mean areal
slope of 2.413mm per year, indicating a continuous precipitation and time series in the study area.

Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya

A trend was generated for the annual rainfall observed in
Snake Farm around Lake Baringo as presented in Figure
5. There was evidence of variation in the annual rainfall
amount over the 30 years (1974-2003). From the trend, the
highest amount received was 1115 mm in 1988 where the
lowest 362 mm was observed in 1984. From the recorded
annual rainfall in Snake Farm Station, showed 668.31mm
± 194.57mm, n = 30. This indicated that there has been
great variation in the annual rainfall received for the 30
years. The linear regression line shows a negative slope
of 4.071mm per year indicating a continuous decline of
annual rainfall. There was no significant relationship
(P>0.05) between the mean areal precipitation and time
series in the study area.
Figure 5 above shows the trend of annual rainfall recorded
in Lake Bogoria National Reserve for 30 years (1974-
2003). The trend shows clearly that, there has been
variation in the annual rainfall received during the study
period, the maximum annual rainfall recorded was 1010
mm in 1978 and the minimum was 247 mm in 1984.
Further, the result shows (658.36mm ± 198.36mm, n = 30).
This shows that there has been highly dispersed rainfall
observed over the 30 years. The slope of the regression
line is negative 0.851mm per year indicating the annual
rainfall is reducing over the 30 years period under review.
There was no significant relationship (P>0.05) between the
mean areal precipitation and time series in the study area.
As indicated in Figure 5 the intensity of variation in the
annual rainfall recorded in Perkerra Agricultural Research
Station is noted in every year within the entire study period
of 30 years. The maximum annual rainfall recorded was
1087 mm in 1977 and a minimum 264.8 in 1984. The
analysed annual rainfall recorded for the 30 years showed
612.32 ± 179.27mm, n = 30, indicating that the variation of
annual rainfall is evident for the study period (1974-2003).
The trend yielded a linear regression line showing a
negative slope of 9.178mm per year depicting that the
annual rainfall observed is continually decreasing. There
was no significant relationship (P>0.05) between the mean
areal precipitation and time series in the study area.
DISCUSSION
The trend demonstrated by Figure 3 on the mean areal
precipitation, corroborates Lelenguyah (2013) findings that
Baringo South/Marigat has been experiencing an
oscillating trend in rainfall with a significant peak observed
in 1977. Other observable peaks are those of 1988-89 and
1997-98 from 1971-2007. These are the four major peaks
with the highest amounts of rainfall over the last four
decades. It is worth noting that the area has been receiving
heavy rainfall in every ten years over 1971-2010 periods.
Also, this study finding agrees with Ochieng et al. (2017)
who reported a declining long-term seasonal rainfall trend
in the drylands of Baringo County. Further, Wakachala et
al. (2015) reported a decreasing trend in annual rainfall
during March-April-May season and high variability within
seasonsin the Great Rift Valley of Kenya.
Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya
J. Agric. Econ. Rural Devel. 332
Although the study area has been experiencing an
oscillating trend in rainfall, it is important to note that, the
amount of annual average rainfall has been decreasing
with time. The linear regression lines with negative slopes
in Figure 3, 4 and 5 supports the decrease in rainfall. It is
perhaps associated with the aspect of climate change and
variability.
Figure 4 shows the trend of variation for amount of rainfall
observed annually that demonstrates a decrease in the
amount rainfall throughout the study period. It corroborates
Omoyo et al. (2015) where rainfall trend analysis for 1994–
2008 revealed that four of the six weather stations in
Machakos County, Kenya were declining up to 3 mm pa.
A similar study by Mwaura et al. (2017) reported that
rainfall in semi-arid Ijara in Garissa, Kenya, has
increasingly become uncertain and the trend analyzed
using KMD data (40 years) period indicates a definite
decline. However, Recha et al. (2017) also analyzed
precipitation data for 10 years in Semi-arid Tharaka
District, Kenya and reported similar results that showed a
declining trend between the 1970s and mid-1980.
From the combined individual station's annual observed
precipitation trend at Chemususu Forest Station (Figure 5)
it was evident that there is climate change and variability
as depicted by some observed rainfall variations. This
indicates that the annual rainfall received over the 30 years
(1974-2003) has not been consistent as indicated by the
negative slope of 5.393mm per year. The general amount
of rainfall received in the station has been reducing during
the study period
From the amount of rainfall observed in Kimose
Agricultural Holding Ground weather station, a trend was
generated as indicated in Figure 5. From the trend, it is
clear that there are variations in the observed amount of
rainfall throughout the 30 years period (1974-2003). This
scenario is an indication of climate variability, which is
tantamount to climate changes. This shows that there has
been a big variation in the amount of rainfall received in a
year for the 30 years in question. From the trend equation,
the predicted annual rainfall amount for the year 2018 is
expected to be 785.92mm. This confirms that the annual
rainfall amount was decreasing consistently from 1974.A
trend was generated for the annual rainfall observed in
Snake Farm around Lake Baringo as presented in Figure
5. There was evidence of variation in the annual rainfall
amount over the 30 years (1974-2003).
Figure 5 above shows the trend of annual rainfall recorded
in Lake Bogoria National Park Reserve for the 30 years
(1974-2003). The trend clearly shows that there has been
variation in the annual rainfall received. For the study
period, the maximum annual rainfall recorded was 1010
mm in 1978, and the minimum was 247 mm in 1984. The
standard deviation computed from the observed rainfall is
highly dispersed over the 30 years.
Muriithi et al. 333
As indicated in Figure 5, the intensity of variation in the
annual rainfall recorded in Perkerra Agricultural Research
Station is noted in every year within the entire study period
of 30 years. The p-value of 0.012<0.05 which shows that
for Perkerra Agricultural Research Institute Station, the
relationship between the time series and the annual rainfall
recorded is significant at 5% levels of significance.
The individual annual rainfall trends for the combined 5
stations from 1974-2003 clearly shows that there has been
variation in the annual amount of rainfall received in the
five weather stations (see Figure 5). From the trends of the
stations, it is clear that there is an annual rainfall variation
across the stations and eventually reduced annual
precipitation. The findings of this study confirm Luseno et
al. (2002) citation in (Galvin et al., 2001) that climatic
variability is especially pronounced and crucial in dry
lands. It also supports Easterling et al. (2000) that, results
of observational studies suggest that in many areas that
have been analysed, changes in total precipitation are
amplified at the tails. Comparing the annual rainfall as
displayed in the trends, it was evident that Chemususu
recorded the highest amount of rainfall, followed by
Kimose, Snake Farm and Lake Bogoria while Perkerra
recorded the lowest. The findings in this study corroborate
IPCC (2007) who reported that, as climate changes,
observations show that changes are occurring in the
amount, intensity, frequency and type of precipitation.
These aspects of precipitation generally exhibit large
natural variability.
CONCLUSION
Based on all the trends analysed using the two methods
for estimating the mean areal precipitation namely; Station
Average and Theissen polygon methods it is clear that
precipitation was decreasing respectively across and
within the area covered by the four stations and for the six
stations in Baringo County, Kenya. This is evidenced by
the negative slope along the linear regression line. It is also
reported that in both methods there was no significant
relationship between the mean areal precipitation and time
series.
The same scenario is reported in the individual combined
stations annual precipitation trends. It is evident that the
general amount of rainfall received in all stations is
decreasing and from all the trends it is apparent that there
are variations. This confirms that the annual rainfall
amount is decreasing consistently from 1974 to 2003 and
all the linear regression equation generated predicted that
the amount of rainfall expected in 2018 in all the stations
is generally decreasing in amount. This aspect of
decreasing amount of precipitation for the long period of
analyses confirms that the pastoralist and agro-
pastoralists claim of the presence of climate change and
variability in the study area. Therefore, it is prudent to
Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya
conduct further research on the adverse effect posed on
the pastoralist and agro-pastoralist and come up with
appropriate interventions to counteract the effects of the
climate variation and change on the livelihoods in the
ASALs.
ACKNOWLEDGEMENT
The authors wish to acknowledge the National
Commission for Science, Technology and Innovation
(NACOSTI) for providing the funds to undertake this study.
Further, the authors appreciate the Kenya Meteorological
Department (KMD) for the provision of the observed
precipitation data used in the study.
REFERENCE
Bhatt Y, Bossio D, Gordon L, Kongo V, Kosgei JR,
Makurira H, Masuki K, Mul M, Tumbo SD (2006).
Smallholder System Innovations in Integrated
Watershed Management (SSI): Strategies of Water for
Food and Environmental Security in Drought-Prone
Tropical and Subtropical Agro- Ecosystems.
http://ageconsearch.umn.edu/bitstream/91943/2/WOR
109. Accessed on June 19, 2017.
Chin LW (2007). Study on Malaysian urban rainfall-runoff
characteristics: Case study of Sungai Kayu Ara,
Damansara, Selangor.Master of Science (Civil
Engineering) Thesis of the UniversitiSains Malaysia.
Cooper P, Stern R, Noguer M, Gathenya J, Osbahr H,
Dorward P, Coe R, Greatrex H. (2012). Evidence-
based adaptation to climate change in East Africa.
Cambridge University Press, Cambridge
Damant C, Austin G, Bellion A, Osseyrane M, Nguyen N.
(1983). Radar rain forecasting for waste water control.
J. Hydraul. Div. ASCE, 109(2):293-297.
Easterling DR, Meehl GA, Parmesan C, Changnon SA,
Karl TR, Mearns LO (2000). Climate Extremes:
Observations, Modelling, and Impacts. Science
289:2068–2074.
Galvin KA, Boone RB, Smith NM, Lynn SJ (2001). Impacts
of climate variability on East African pastoralists: linking
social science and remote sensing.
https://courses.eas.ualberta.ca/eas523/Week 7/Paper
2. Accessed on October 4, 2017.
ILRI (2014). Kenya ending drought emergencies policy
review: Scenarios for building resilience in the ASALs.
http://www.technicalconsortium.org/wp-. Accessed on
September 6, 2017.
IPCC (2007). Climate Change 2007: Synthesis Report.
Contribution of working groups I, II and III to the Fourth
Assessment Report of the Intergovernmental Panel on
climate change [Core Writing Team, Pachauri RK and
Reisinger A.(eds.)]. IPCC, Geneva, Switzerland, 104
pp.

IPCC (2014). IPCCWGIIAR5. Summary for Policy Makers.
http://www.climatechange2013. Accessed on January
20, 2017.
Jiri O, Paramu L, Mafongoya, Chipo, Mubaya1,
Mafongoya O. (2014). Seasonal climate prediction and
adaptation using indigenous knowledge systems in
agriculture systems in Southern Africa: J. Agric. Sc. Vol.
8, No. 5; 2016.Review
Jones JW, Hansen 1, Royce FS, Messina CD (2000).
Potential Benefits of Seasonal Weather Prediction to
Agriculture: Univ.of Florida(USA). J. Agric. Ecosystems
andEnv. 82: 169–184.
Kibria G, Haroon AKY1, Nugegoda D. (2017). Climate
Change Impacts on Tropical and Temperate Fisheries,
Aquaculture, and Seafood Security and Implications.
Journal of Livestock Research for Rural Development
29 (1) 2017. http://www.lrrd.org/lrrd29/1/kibr29012.htm.
Lelenguyah GL (2013). Effects of Climate Variability on
Pastoral Livelihoods in Marigat District, Baringo
County, Kenya. MasterThesis, Kenyatta Univ. Nairobi,
Kenya.
Linsley BK, Ren L, Dunbar RB, Howe SS (2000). El Niño
Southern Oscillation (ENSO) and Decadal-Scale
Climate Variability at 10øN in the Eastern Pacific from
1893 to 1994: A Coral-Based Reconstruction from
Clipperton.
Http://www.atmos.albany.edu/daes/faculty/howe/Linsle
y et al. Accessed on June 21, 2017.
Luseno WK, John G, McPeak JG, Barrett CB, Little PD,
Gebru G. (2002). Assessing the value of seasonal
weather predictions information for pastoralists:
Evidence from Southern Ethiopia and Northern Kenya.
Cornell University, USA.
Mwaura J, KoskeJ, Kiprotich B. (2017). Economic Value of
water harvesting for climate smart adaptation in Semi
and Arid Ijara Garissa, Kenya. Environ Syst
Res.doi: 10.1186/s40068-017-0088-3.
Nyasimi, M. (2014). East Africa learning event for
practitioners, researchers and policy makers on
community-based adaptation and resilience in East and
Southern Africa’s Drylands.September 1- 4, 2014 Addis
Ababa, Ethiopia Learning Event Report.
Netherlands Development Organisation-SNV (2012). SNV
Annual Report. http://www.snv.org/public. Accessed on
August 4, 2017.
Ochieng R, Recha C, Bebe BO, Ogendi GM (2017).
Rainfall variability and droughts in the drylands of
Baringo County, Kenya. Open Access Library Journal,
4: E3827. https://doi.org/10.4236/oalib.1103827.
Accessed, October 20, 2017.
Omoyo NN, Wakhungu J, Oteng S. (2015). Effects of
Climate Variability on Maize Yield in the Arid and Semi
Arid Lands of Lower Eastern Kenya. Agriculture and
Food Security. DOI 10.1186/S40066-015-0028-2
Long-term observed Precipitation Trends in Arid and Semi-arid Lands, Baringo County, Kenya
J. Agric. Econ. Rural Devel. 334
Radi NFA, Zakaria R, Azman MAZ (2015). Estimation of
missing rainfall data using spatial interpolation and
imputation methods. The 2nd ISM International
Statistical Conference 2014 (ISM-II ): Empowering the
Applications of Statistical and Mathematical Sciences ,
1643, 42-48. https://doi.org/10.1063/1.4907423.
Accessed on August 15, 2017.
Recha CW, Makokha GL, Shisanya CA and Mukopi MN
(2017). Climate Variability: Attributesand indicators of
adaptive capacity in Semi-Arid Tharaka Sub-County,
Kenya. Open accesslibrary journal, 4:
e3505.https://doi.org/10.4236/oalib.1103505
Saitabau H.(2014). Impacts of Climate Change on the
Livelihoods of Loitamaasai Pastoral Community and
Related Indigenous Knowledge on Adaptation and
Mitigation.http://africanclimate.net/en/publications.
Accessed on October 5, 2017.
Tol R, Fankhauser S, Joel B. (1998). The Scope for
Adaptation to Climate Changes. Glob. Environ. Change
8(2):109–123
Tumbo, SD, Rwehumbiza FB, Kahimba, FC, Enfors E,
Mahoo HF, Mbilinyi BP, Mkoga Z Churi A. (2010).
Agronomic Management Strategies for Adaptation to
the Current Climate Variability: The Case of North-
Eastern Tanzania. Final Proc.2nd Int. Conf. On Climate,
Sustainability, and Development in Semi-Arid Regions.
Brazil, Pp.16-20
Wakachala, FM, Zablon, W, Shilenje, ZW, Nguyo, J,
Shaka and Apondo, W. (2015). Statistical Patterns of
Rainfall Variability in the Great Rift Valley of
Kenya.Journal of Environmental and Agricultural
Sciences (ISSN: 2313-8629).
https://www.researchgate.net/publication/282641024.
Accessed on December 4, 2017.
Zillman J. (2009). Adaptation to a Variable and Changing
Climate: Challenges and Opportunities for NMHSs. In
Scientific Lecture, World Meteorological Organisation,
EC-LXI, Geneva, 11th June, 2009.
Accepted 9 December 2017
Citation: Muriithi G, Olago D, Ouma G, Oriaso S (2018).
Long-term observed Precipitation Trends in Arid and
Semi-arid Lands, Baringo County, Kenya. Journal of
Agricultural Economics and Rural Development, 4(1): 326-
334.
Copyright: © 2018 Muriithi et al. This is an open-access
article distributed under the terms of the Creative
Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium,
provided the original author and source are cited.