Journal of Agricultural Economics and Rural Development

Vol. 7(1), pp. 978-996, March, 2021. © www.premierpublishers.org, ISSN: 2167-0477

Research Article

Spatio-Temporal Dynamics of Climatic Parameters

in Togo
*BALAKA Manamboba Mitélama1, YOVO Koffi2, EGBENDEWE Aklesso1
1University of Lomé-Togo / Faculty of Economics and Management (FASEG) / PO Box: 1515 Lomé-Togo
2University of Lomé-Togo / Advanced School of Agronomy (ESA) / PO Box: 1515 Lomé-Togo

The detection of the variability and trends of weather variables is a necessary step in the
explanation of the impacts of climate change on ecosystems and agricultural production
activities. To this end, this study analyses the spatio-temporal trends of precipitation,
temperature, evapotranspiration, relative humidity, wind speed (weather stations in Lomé,
Atakpamé, Sokodé, Kara and Dapaong) and insolation (stations in Lomé, Atakpamé, Sokodé,
Kara and Mango). The data analyzed are from the meteorological directorate-general. Trends
over time were calculated over monthly, annual and seasonal intervals using analysis of
variability (coefficient of variation, precipitation concentration indices) and trend analysis
techniques using Mann-Kendall and Sen slope methods, allowing non-parametric statistical
analysis. The results show a general upward trend in the inter-seasonally, intra and inter-yearly
in precipitation, temperature (maximal and minimal) and evapotranspiration. On the other hand,
a general downward trend is detected for relative humidity (maximal and minimal) and
insolation. The results raise concerns for food security in Togo inasmuch as increases in
rainfall, temperature and evapotranspiration coupled with decreases in relative humidity and
insolation negatively affect agricultural production. It urgently needs that the public authorities
take adaptation and mitigation measures.

Keywords: Spatio-temporal dynamics - Meteorological variables - Mann-Kendall's test – Sen’s slope estimator -Togo.

INTRODUCTION

Partly due to the increase in anthropogenic concentration Indeed, precipitation and temperature are the main
of greenhouse gas (GHG), the warming of the planet components that explain seasons, water availability, the
cannot be denied (Asfaw et al., 2018). This warming is at occurrence of droughts and floods (Chattopadhyay et
the root of the observed global climate change (Kate et al., Edwards, 2016). These two variables, along with
2015; Sonali et Kumar, 2016; Ali et Abubaker, 2019). evapotranspiration, are the most important variables in
Climate change is leading to the increased variability in climate science and hydrology and are often used to
meteorological parameters such as precipitation, measure the extent and magnitude of climate change and
temperature, relative humidity, insolation, wind speed and variability (Parry et al., 2007; Chattopadhyay et Edwards,
evapotranspiration (Hartmann et al., 2013; Hayelom et al., 2016; Mansour et al., 2017).
2017; Coulibaly et al., 2018). As long as the climate
continues to change, the risks associated with climate Known as one of the main components of the hydrological
extremes are increasing (Easterling et al., 2016). Climate cycle, potential evapotranspiration is influenced directly or
change is thus receiving much attention because of its
effects on water resources and social life (El-Nesr, 2010). *Corresponding Author: BALAKA Manamboba
Precipitation and temperature variations are the main University of Lomé-Togo / Faculty of Economics and
determinants of climate change (Hayelom et al., 2017; Management (FASEG) / PO Box: 1515 Lomé-Togo
Pandey et Khare, 2018). These determinants have the *E-mail: [email protected]
2
most important influence and role in the hydrological cycle E-mail: [email protected]
1
(Chattopadhyay et Edwards, 2016; Aawar et al., 2019). E-mail: [email protected]

Spatio-Temporal Dynamics of Climatic Parameters in Togo

Balaka et al. 979

indirectly by other meteorological variables such as

precipitation, temperature, insolation, wind speed and
relative humidity, as well as by soil and plant Analysis of variability
characteristics (Sonali et Kumar, 2016; Mansour et al.,
2017). Potential evapotranspiration, the union of soil The analysis of variability involves the use of statistical
evaporation and plant transpiration, is the main indicator of parameters such as the mean, the standard deviation (σ),
climate change because of its relationship with other the skewness coefficient (Cs), the kurtosis coefficient (Ck),
meteorological variables (Sonali et Kumar, 2016). It affects the coefficient of variation (Cv), the rainfall concentration
crop water requirements and water allocation for index 𝑅𝐶𝐼.
agriculture (Yong-fang et al., 2011). It’s variations also
affect hydrological processes and crop yields (Hobbins et Fisher's asymmetry coefficient (skewness)
al., 2004). It is an important component of the hydrological
cycle and plays an important role in crop growth, water It is determined as follows (Aawar et al., 2019) :
demand and irrigation management (Mansour et al., ∑𝑛 (𝑥 −𝑥̅ )3
𝐶𝑠 = 𝑖=0 𝑖 3 Equation (1)
2017). It is also an important component for climatology, 𝑛∗𝜎
plant growth estimation and agricultural irrigation It can be positive or negative. If -0.5 ≤𝐶𝑠 < 0.5; the series is
management (Sun et al., 2016). fairly symmetrical. If -1≤𝐶𝑠 <-0.5 or 0.5≤ 𝐶𝑠 ≤1; the series is
moderately biased. If 𝐶𝑠 < -1 or 𝐶𝑠 >1; the series is highly
The effects of changes in weather variables on agricultural skewed.
crops, especially in countries where rain-fed agriculture is Pearson Flattening Coefficient (kurtosis)
predominant, make it crucial to examine the space-time
dynamics of these weather variables in the context of It is defined as follows (Aawar et al., 2019) :
climate change. This is useful for assessment, irrigation
∑𝑛 (𝑥 −𝑥̅ )4
planning, monitoring and management of water resources, 𝐶𝑘 = 𝑖=0 𝑖 4 Equation (2)
𝑛∗𝜎
management of agricultural production, and adaptation to
If 𝐶𝑘 =3, the distribution is said to be mesokurtic or normal.
flood and drought conditions (Chattopadhyay et Edwards,
If 𝐶𝑘 <3, the distribution is said to be platykurtic or hypo
2016; Mansour et al., 2017; Ali et Abubaker, 2019). This
normal (less flattened). If 𝐶𝑘 >3, the distribution is said to
analysis also allows us to suggest feasible adaptation
be leptokurtic or hyper normal (more flattened).
strategies (Asfaw et al., 2018). Since the end of the 1960s,
like other West African countries, Togo has been Coefficient of variation
undergoing significant spatio-temporal variations in
weather patterns with consequences on agriculture, the It is calculated as (Asfaw et al., 2018) :
𝜎
main activity of the majority of the population (Adewi et al., 𝐶𝑣 = ∗ 100 Equation (3)
𝑥̅
2010; Maléki et al., 2014). Consequently, as agricultural
production in Togo is mainly rain fed, the inter, intra-annual where 𝐶𝑣 represents the coefficient of variation. According
and seasonal variability of climatic parameters is often to Hare (2003), the 𝐶𝑣 is used to classify the degree of
cited to explain the efficiency of the agricultural production variability of climate events as lower (𝐶𝑣 < 20%), moderate
system and the situation of food insecurity. The (20% <𝐶𝑣 <30%), high (𝐶𝑣 > 30%), very high (𝐶𝑣 > 40%) and
characterization of the inter, intra-annual and seasonal extremely high (𝐶𝑣 > 70 %).
spatio-temporal trend of climatic parameters is important.
Thus, the objective of this study is to analyze the inter, Rainfall concentration index
intra-annual and seasonal variability as well as the spatio-
temporal trend of meteorological variables in Togo. The The annual or seasonal precipitation concentration index
research is intended to provide farmers and decision- 𝑅𝐶𝐼𝑎/𝑠 is used to examine the variability of precipitation at
makers with useful information for decision-making on the different scales. This index is calculated according to
one hand and to suggest feasible adaptation strategies Asfaw et al. (2018) as :
∑12 𝑃2
and agricultural practices on the other hand. 𝑖=1 𝑖
𝑃𝐶𝐼𝑎/𝑠 = ̅̅̅̅̅̅̅̅̅̅̅̅̅̅
12 ∗ 100 Equation (4)
2
(∑𝑖=1 𝑃𝑖 )
where: 𝑃𝑖 represents the amount of rainfall in the ith month.
MATERIALS AND METHODS
According to Oliver (1980), values 𝑅𝐶𝐼𝑎/𝑠 below 10 indicate
Methods of analysis a uniform annual/seasonal distribution of precipitation (low
precipitation concentration), values between 11 and 15
The analysis of the spatio-temporal dynamics of climate indicate moderate concentration, values between 16 and
parameters is based on a number of techniques that are 20 indicate high concentration and values of 21 and above
generally classified into two categories: variability and indicate very high concentration.
trend analysis techniques.

Spatio-Temporal Dynamics of Climatic Parameters in Togo

 J. Agric. Econs. Rural Dev. 980

variable. As a result, trends determined are less affected

Percentage of change over time series
by outliers (Birsan et al., 2005). The Mann-Kendall test
statistics is calculated on the basis of Mann (1945),
It is calculated by using the median slope of Theil and Sen,
Kendall (1975), El-Nesr (2010), Chattopadhyay and
and then the mean is evaluated over the period as follows
Edwards (2016) and Aawar et al. (2019) using the
(Meena et al., 2015; Aawar et al., 2019) :
following formula:
𝑉𝑎𝑟 𝑒𝑛 % =
𝑚𝑒𝑑𝑖𝑎𝑛 𝑠𝑙𝑜𝑝𝑒 (𝑄𝑖 )∗𝑙𝑒𝑛𝑔𝑡ℎ 𝑜𝑓 𝑝𝑒𝑟𝑖𝑜𝑑 (𝑛)
Equation (5) 𝑆 = ∑𝑛−1 𝑛
𝑖=1 ∑𝑗=𝑖+1 𝑠𝑖𝑔𝑛 (𝑥𝑗 − 𝑥𝑖 ) Equation (7)
𝑚𝑒𝑎𝑛
The trend test is carried out on a time series 𝑥𝑖 ranked
Trend analysis from 𝑖 = 1, 2, ... n-1 and 𝑥𝑗 , which is ranked from 𝑗 = i + 1,
2, ... n. Each of the 𝑥𝑖 data points is taken as a reference
Different statistical test methods are used to analyze point that is compared to the rest of the 𝑥𝑗 data points so
trends in the fields of hydrology and hydrometeorology. In that (Mann, 1945; Kendall, 1975; Sonali and Kumar, 2016;
general, these methods are classified into two categories: Asfaw et al., 2018; Aawar et al., 2019):
parametric and non-parametric tests. Due to the rare +1 𝑖𝑓(𝑥𝑗 − 𝑥𝑖 ) > 0
probability of normal distribution in climate time series
data, non-parametric tests are primarily used 𝑠𝑖𝑔𝑛 (𝑥𝑗 − 𝑥𝑖 ) = {0 𝑖𝑓(𝑥𝑗 − 𝑥𝑖 ) = 0, Equation (8)
(Chattopadhyay et Edwards, 2016; Asfaw et al., 2018; −1 𝑖𝑓(𝑥𝑗 − 𝑥𝑖 ) < 0
Aawar et al., 2019). The Mann-Kendall (MK) test and the with 𝑥𝑖 and 𝑥𝑗 the annual/seasonal values of
rho test of Spearman are the most important and common years/seasons 𝑖 and 𝑗 (𝑗 > 𝑖) respectively. The (𝑆) statistic
(Aawar et al., 2019). However, the rho test of Spearman is is closely related to Kendal's 𝜏 which is represented by
less used in research than the Mann-Kendall test for Aawar et al. (2019) and El-Nesr (2010) respectively as
analyzing monotonic trends in weather data (Ahmad et al., follows:
2015). In this study, precipitation and temperature indices 𝑠
𝜏= ,
𝐷
are calculated. Then, the Mann-Kendall test and the slope
With
estimation of Sen (1968) are done. 1 1
1 1 𝑞 2 1 2
Standardized indices of meteorological variables 𝐷 = [ 𝑛(𝑛 − 1) − ∑𝑝=1 𝑡𝑝 (𝑡𝑝 − 1)] [ 𝑛(𝑛 − 1)]
2 2 2
(Equation 9)
𝑠
These indices are calculated to examine the nature of the 𝜏= ,
𝑆𝑚𝑎𝑥
trends. This is used to determine dry and wet years and to
With
assess the frequency and severity of droughts. The 1
general shape of these indices is as follows (Asfaw et al., 𝑆𝑚𝑎𝑥 = 𝑛(𝑛 − 1)
2018) : 2
(𝑋𝑖 −𝑋̅)
𝑍= Equation (6) 𝑆𝑚𝑎𝑥 is the maximum value of 𝑆. A positive (negative)
𝜎
where 𝑍 is a normalized anomaly of meteorological value of 𝑆 or 𝜏 indicates an upward (downward) trend. It
variables; 𝑋𝑖 is the annual meteorological variable for a has been established that for a sample size more than 10
given year; 𝑋̅ is the mean annual meteorological variable (n≥10), the 𝑆 statistic of the MK test is assumed to be
over the observation period and 𝜎 is the standard deviation normally distributed. Thus, the mean is 𝐸[𝑆] = 0 (Kendall,
of the annual meteorological variable over the observation 1975). In this case, the variance of the S statistic is defined
period. For precipitation, the drought severity classes as as follows (Mann, 1945; Kendall, 1975; El-Nesr, 2010;
defined by Agnew et Chappell (1999) are: extreme drought Chattopadhyay and Edwards, 2016; Sonali and Kumar,
(𝑍 < -1.65), severe drought (-1.65< 𝑍 <-1.28), moderate 2016 Asfaw et al., 2018; Aawar et al., 2019):
1 𝑞
drought (-1.28 < 𝑍 < -0.84) and no drought (𝑍 > -0.84). 𝜎𝑆2 = [𝑛(𝑛 − 1)(2𝑛 + 5) − ∑𝑝=1 𝑡𝑝 (𝑡𝑝 − 1)(2𝑡𝑝 + 5)]
18
Mann-Kendall test Equation (10)

This test is a non-parametric test developed by Mann where 𝑛 is the number of observations, 𝑞 is the number
(1945) for the identification of monotonic trends in of linked groups in the data set (when certain data values
environmental, climatic or hydrological data series and by appear more than once in the database, these groups of
Kendall (1975) for the magnitude of the statistical values are called linked groups) and 𝑡𝑝 is the number of
distribution of the test. The MK test is used in this study to data points in the pth linked group. In this equation, the
detect the presence of monotonic (increasing or second part supports linked data and sets the effects of
decreasing) trends of meteorological variables and the linked groups in the data set. Thus, the (𝑍) statistics of
determine whether or not the trend is statistically Mann-Kendall, which is approximately normally
significant. This choice is justified by the fact that outliers distributed, is formalized as follows (Mann, 1945; Kendall,
are likely to be present in the data set. The non-parametric 1975; El-Nesr, 2010; Chattopadhyay et Edwards, 2016;
test of MK is useful because its statistics is based on the Sonali et Kumar, 2016 Asfaw et al., 2018; Aawar et al.,
signs (+ or -), rather than on the values of the random 2019) :

Spatio-Temporal Dynamics of Climatic Parameters in Togo

 𝑠−1
, 𝑖𝑓 𝑆 > 0 (2004) have been proposed. The modified variance of the
𝜎𝑆
𝑆 statistic of the MK test is given by:
𝑍𝑠 = 0 , 𝑖𝑓 𝑆 = 0 Equation (11) 2
𝜎𝑚𝑜𝑑𝑆 = 𝐶𝐹 ∗ 𝜎𝑆2 Equation (15)
𝑠+1
{ 𝜎𝑆 , 𝑖𝑓 𝑆 < 0 where CF is the correction factor. The CF proposed by
where 𝑍𝑠 follows a normal distribution. A positive or Hamed et Rao (1998) and Yue and Wang (2004) are
negative 𝑍𝑠 represent an upward and downward trend for designated by CF1 and CF2 respectively and defined as
the period, respectively. follows (Sonali and Kumar, 2016) :
2
𝐶𝐹1 = 1 + ∑𝑛−1
𝑘=1 (𝑛 − 𝑘)(𝑛 − 𝑘 − 1) ∗ (𝑛 − 𝑘 −
Sen’s slope estimators 𝑛(𝑛−1)(𝑛−2)
2)𝑟𝑘𝑅 Equation (16)
Sen's (1968) slope estimator is a non-parametric method
𝑘
complementary to the MK test for calculating the 𝐶𝐹2 = 1 + 2 ∑𝑛−1
𝑘=1 (1 − ) 𝑟𝑘 Equation (17)
𝑛
magnitude of trends. The non-parametric approach to the
slope of Sen (1968) shows both the direction and where 𝑟𝑘 and 𝑟𝑘𝑅 are the coefficients of the lag-k serial
magnitude of the trend in a time series. The linear 𝑓(𝑖) correlation of the data and the data ranks respectively, 𝑛
model can be described as follows (Aawar et al., 2019): is the size of the series. In this study, the series of weather
variables were examined for the presence of serial
𝑓(𝑖) = 𝑄𝑖 + 𝐵 Equation (12) correlation in order to judge the need for "Pre-Whitening".
where 𝑄𝑖 is the slope and 𝐵 a constant. The magnitude The serial correlation coefficient 𝑟 is calculated as follows
of the trend is predicted by the slope estimator methods of (Chattopadhyay and Edwards, 2016) :
(Sen, 1968; Theil, 1992). Here, the slope 𝑄𝑖 of all data 1
∑𝑛−1
𝑖=1 (𝑥𝑖 −𝑥̅ )(𝑥𝑖+1 −𝑥̅ )
pairs is calculated as (Sen, 1968). To calculate the slope 𝑟 = 𝑛−1 1 𝑛 Equation (18)
∑ (𝑥 −𝑥̅ )2
𝑄𝑖 , it is necessary to calculate the slopes of all data pairs 𝑛 𝑖=1 𝑖
No significant serial correlation is present if
as follows (Aawar et al., 2019):
𝑋𝑗 −𝑋𝑘 (Chattopadhyay and Edwards, 2016) :
𝑄𝑖 = , 𝑖=1,2,…,N, 𝑗> 𝑘 Equation (13) −1−1,645√𝑛−2 1+1,645√𝑛−2
𝑗−𝑘 ≤𝑟≤ Equation (19)
where 𝑥𝑗 and 𝑥𝑘 are considered as data values at the 𝑛−1 𝑛−1
Thus, if the value of 𝑟 lies between the two bounds, the
times 𝑗 and 𝑘 (𝑗 > 𝑘) accordingly. The number of slopes MK test and the Sen's slope test are applied directly to the
belongs to the number of 𝑋𝑗 values, so if there are n 𝑋𝑗 series of weather variables. Otherwise, the "Pre-
𝑛(𝑛−1)
values in a time series, then 𝑁 = corresponds to the Whitening" method is applied before the estimates.
2
𝑄𝑖 estimates of the slope. The median of the N values of
𝑄𝑖 indicates the slope estimator of (Sen, 1968). The N In this study, several softwares and tools such as R,
values of 𝑄𝑖 are sorted from smallest to largest and the MAKESENS 1.0 Excel template and XLSTAT are used to
estimator of Sen (1968) given by Aawar et al. (2019) is: analyze the meteorological data.
𝑄𝑁+1 𝑖𝑓 𝑁 𝑖𝑠 𝑜𝑑𝑑
2
𝑄𝑚𝑒𝑑 = {1 Equation (14) Study area and data collection
(𝑄𝑁 + 𝑄𝑁+2 ) 𝑖𝑓 𝑁 𝑖𝑠 𝑒𝑣𝑒𝑛
2 2 2
Togo is a French-speaking country in West Africa. It is
A positive value of 𝑄𝑖 indicates an upward trend and located between 6° and 11° N and 0° and 1°40 E. Togo is
a negative value of 𝑄𝑖 indicates a downward trend in the subdivided into five (05) administrative regions (Maritime,
time series. The existence of a positive serial correlation in Plateaux, Central, Kara and Savanes). The climatic and
a time series increases the probability that the MK test will rainfall conditions in the five regions determine the
detect a significant trend (serial correlation increases the cropping systems and cropping calendars. The latters are
probability of committing Type I error) (Von Storch, 1999). influenced by the rainy season regimes. The country is
Hamed and Rao (1998) reported that the presence of a characterized by a bimodal rainfall regime in the South to
serial correlation in a time series does not falsify neither a monomodal regime in the North. The country has two
the asymptotic normality nor the mean of the 𝑆 statistics in climatic zones: the Sudanian climate zone (kara and
the MK test, but the variance does change. Hence, savannah regions) in the north and the Guinean climate
correction factors are proposed to correct the variance of zone (maritime regions and plateaus) in the south. The
𝑆 using only the uncorrelated samples. Yue et al. (2002) boundary between the two climatic zones is actually a
found that "Pre-Whitening" changes the true slope present transition zone that oscillates around 7°N (Adewi et al.,
in the time series and suggested the TFPW-MK (Trend 2010).
Free Pre-Whitening Mann Kendall) approach that takes
into account the serial correlation effect in trend detection The hydrographic system is made up of three main basins:
approaches. In addition, the presence of positive the Volta basin, to the north, drains over 26,700 km2 the
(negative) serial correlation leads to an increase main rivers (Oti, Kara, Mô) and represents 47.3% of the
(decrease) in the variance of the 𝑆 statistic of the MK test. territory; the Mono basin drains over 21,300 km2 the main
Therefore, the variance correction approaches MMK-CF1 rivers (Mono, Anié, Amou), or 37.5% of the territory
(Modified Mann Kendall) by Hamed and Rao (1998) and towards the south-east and the coastal basin (Lake Togo
MMK-CF2 (Modified Mann Kendall) by Yue and Wang and rivers Zio and Haho) which represents 14.3% of the
territory (MERF, 2015).

Spatio-Temporal Dynamics of Climatic Parameters in Togo

 J. Agric. Econs. Rural Dev. 982

The data analyzed are from the meteorological directorate- Kara and Dapaong stations. The sunshine data are from
general (MDG). The data on temperature, Lomé, Atakpamé, Sokodé, Kara and Mango stations due
evapotranspiration, relative humidity, wind speed and to lack of sunshine data at Dapaong station and in order to
precipitation are taken from the Lomé, Atakpamé, Sokodé, cover all regions of Togo.

Table 1: Geographical details of synoptic weather stations in Togo

Regions Stations Longitude (East) Latitude (North) Altitude (m)
Maritime Lomé 01°15’ 06°10’ 19.6
Plateaux Atakpamé 01°07’ 07°35’ 399.66
Centrale Sokodé 01°09’10’’ 08°59’50’’ 386.48
Kara Kara 01°10’ 09°33’ 341.73
Savanes Mango 00°22’ 10°22’ 144.7
Dapaong 00°15’ 10°52’ 329.01

the minimum is 00 mm and the maximum is 263.70 mm

RESULTS AND DISCUSSIONS
(Atakpamé). The distribution of rainfall according to
months, years, rainy and dry seasons is very varied. The
Primary statistical parameters such as mean, standard
analysis of the coefficients of annual variation and rainy
deviation (σ), skewness coefficient (𝐶𝑠 ), coefficient of
seasons shows low rainfall variations in all stations except
kurtosis (𝐶𝑘 ), coefficient of variation (𝐶𝑣 ), rainfall
that of Lomé where the variations are moderate. In dry
concentration index (𝑅𝐶𝐼) and the percentage of change seasons, rainfall variations are very high (Sokodé, Kara
over time series are calculated using Equations 1, 2, 3, 4 and Dapaong) and extremely high (Lomé and Atakpamé).
and 5. Then, the standardized precipitation and The distribution of annual rainfall and rainy seasons varies
temperature (maximal and minimal) index, the normalized from fairly symmetrical to moderately biased. In the dry
statistics (𝑍) of the Mann-Kendall test and the median seasons, the distribution varies from moderately biased to
slope estimator of Sen (1968) are calculated through highly biased. Annual and rainy season rainfall is generally
Equations 6 to 19 for all meteorological variables. less flattened. In dry seasons, they are more flattened. At
Analysis of precipitation dynamics the different weather stations, the months with average
rainfall above 40 mm correspond to the months of the rainy
As indicated in Table 2, annual and rainy season rainfall seasons. Analysis of the coefficients of variation of
varied from a minimum of 423.90 mm and 378.80 mm monthly rainfall shows, according to the classification of
respectively (Lomé) to a maximum of 1850.10 mm and Hare (2003), very high and extremely high variations for
1837.60 mm respectively (Atakpamé). In the dry season, several months of the year at all stations.

Table 2: Values of primary statistical parameters of precipitation (mm)

Stations Periods Min Max Mean SD Cv Cs Ck % chan. n
Lomé Year 423.90 1416.70 806.79 205.39 0.26 0.18 3.41 0.07 46
RS 378.80 1250.20 733.72 184.04 0.25 0.27 3.26 0.13
Atakpamé Year 767.30 1850.10 1332.99 247.39 0.19 0.02 2.50 0.03 46
RS 758.70 1837.60 1326.71 246.10 0.19 -0.02 2.43 0.02
Sokodé Year 964.50 1646.10 1272.19 152.81 0.12 0.19 2.78 -0.05 46
RS 918.50 1581.40 1194.35 157.96 0.13 0.41 2.59 -0.04
Kara Year 1003.70 1771.10 1305.57 199.44 0.15 0.67 2.67 -0.05 46
RS 945.00 1695.90 1243.88 190.90 0.15 0.50 2.43 -0.05
Dapaong Year 765.30 1383.60 1040.36 168.66 0.16 0.06 2.11 0.22 46
RS 686.40 1256.60 959.69 155.20 0.16 0.03 2.15 0.20
Year: annual, RS: rainy season, Min: minimum, Max: maximum, SD: standard deviation, Cv: coefficient of variance, Cs:
coefficient of skewness, Ck: coefficient of kurtosis, % chan.: percentage change over the period, n: length of period

Spatio-Temporal Dynamics of Climatic Parameters in Togo

Balaka et al. 983

Coefficients of variation of monthly precipitation at almost all stations are start and end dates of potential rainfall to be useful. These spatial and temporal
greater than 30%. This implies that the intra-annual variation in precipitation is rainfall variations negatively affect the agricultural production system.
high. These results corroborate those of Hounnou and Dedehouanou (2018) in Analysis of the standardized precipitation indices reveals remarkable inter-
Benin, Hayelom et al. (2017) in south Ethiopia, Asfaw et al. (2018) in north- annual fluctuations with a succession of wet and dry periods at all weather
central Ethiopia. Maléki et al. (2014) detected well-contrasting spatial and inter- stations analyzed (See Figure 1).
annual rainfall variability in Togo. Adewi et al. (2010) found variability in the

Figure 1: Illustration for some weather stations of the inter-annual variability of precipitation, maximal and minimal temperatures.
3 3

Standardized indexs

Standardized indexs
Standardized indexs

of precipitation
2

of precipitation
of precipitation

1 1 4
0 2
-1 0
-2 -1 0
-3 -2 -2

2007
1971
1975
1979
1983
1987
1991
1995
1999
2003
2007
2011
2015

1971
1975
1979
1983
1987
1991
1995
1999
2003

2011
2015

1971
1975
1979
1983
1987
1991
1995
1999
2003
2007
2011
2015
Years Years Years
Atakpamé Kara Dapaong

2 3 2
Maximal temperature
standardized indexs

Maximal temperature

Maximal temperature
2

standardized indexs

standardized indexs
0
0 1
0 -2
-2 -1
-4
-2
-4 -3 -6

1997
1971
1975
1979
1983
1987
1991
1995
1999
2003
2007
2011
2015

1977
1981
1985
1989
1993

2001
2005
2009
2013

1971
1975
1979
1983
1987
1991
1995
1999
2003
2007
2011
2015
Years Years Yeare
Atakpamé Kara Dapaong

4 2 2
Minimal temperature

Minimal temperature

Minimal temperature
standardized indexs

standardized indexs

standardized indexs
2 1
0 0
0
-2 -1
-2 -2
-4 -4 -3
1997
1971
1975
1979
1983
1987
1991
1995
1999
2003
2007
2011
2015

1977
1981
1985
1989
1993

2001
2005
2009
2013

1971
1975
1979
1983
1987
1991
1995
1999
2003
2007
2011
2015
Years Years Years
Kara Dapaong
Atakpamé
Spatio-Temporal Dynamics of Climatic Parameters in Togo
 J. Agric. Econs. Rural Dev. 984

The indices of annual and seasonal rainfall concentration Bewket (2017) in the central highlands of Ethiopia and
range from moderate to very high (Table 3). They are more Asfaw et al. (2018) north-central Ethiopia. Zamani et al.
often high at all stations. Similar high rainfall (2018) found a highly irregular and non-uniform distribution
concentrations have been reported by Alemayehu and of annual and seasonal rainfall.

Table 3: Concentration index of annual and in rainy season precipitation from 1971 to 2016
Indices Concentration Lomé Atakpamé Sokodé Kara Dapaong
Annual precipitation (number of years from 1971-2016)
<10 Weak (Uniform) 0 0 0 0 0
11&15 Moderate 8 33 13 13 0
16&20 high 29 13 33 32 38
≥21 Very high 9 0 0 1 8
Indice Concentration Rainfall in rainy seasons (Number of years from 1971-2016)
<10 Weak (Uniform) 0 0 0 0 0
11&15 moderate 1 33 0 0 0
16&20 high 27 13 46 44 22
≥21 Very high 18 0 0 2 24
The detection of rainfall trends athwart Mann-kendall's test and Sen (1968) slope estimator reveals a general upward trend
in rainy and dry seasons as shown in Table 4.

Table 4: Mann-kendall’s and Sen’s tests applied to precipitation (mm)

Stations Lomé Atakpamé Sokodé Kara Dapaong
Statistics MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi
Jan 0.32 0.00 -0.07 0.00 0.01 0.00 3.45 0.00 1.91+ 0.00+
Feb 0.04 0.00 -0.07 0.00 0.86 0.00 0.89 0.00 2.51 * 0.00*
Mar 1.15 0.53 -1.14 -1.02 -1.02 -0.37 -1.51 -0.35 -1.06 -0.08
Apr 0.3 0.14 -0.59 -0.45 -0.45 -0.26 -0.42 -0.2 2.11* 0.53*
May 0.28 0.23 0.66 0.45 -1.49 -0.79 -2.42 ** -1.09** 0.08 0.04
June -0.04 -0.06 -0.17 -0.22 -1.24 -0.98 -0.04 -0.06 1.99* 1.23*
July 0.29 0.16 0.76 0.75 -0.98 -0.86 -1.31 -1.09 -0.74 -0.48
Aug 1.34 0.32 0.62 0.6 0.88 0.85 -0.18 -0.15 0.93 1.06
Sep 0.61 0.3 0.4 0.37 2.02* 1.62* 0.57 0.51 1.86+ 1.38+
Oct 2.01 * 1.11 * 0.45 0.38 0.28 0.19 1.23 0.92 1.25 6.1
Nov 0.29 0.04 -0.7 -0.04 -0.03 0.00 0.48 0.00 0.9 0.00
Dec -1.2 0.00 -0.47 0.00 -1.02 0.00 -1.37 0.00 -2.84** 0.00**
Year 0.59 1.3 0.27 0.76 -0.79 -1.41 -0.55 -1.33 2.50 * 4.94*
RS 0.83 2.05 0.25 0.69 -0.62 -1.12 -0.64 -1.4 2.56 * 4.14*
DS 0.68 0.36 -0.07 0.00 0.01 0.01 -0.4 -0.18 1.25 0.71
RS: rainfall season, DR: dry season, Z: normalized statistic of the Mann-Kendall test, Qi: the median slope of Sen, +, *, **
and ***: statistically significant at 10%, 5%, 1% and 0.1% respectively.

The maximum rates of change in annual and wet season is detected. A significant upward trend is detected in
rainfall values are 4.94 mm/year and 4.14 mm/season at October (Lomé); September (Sokodé); April, June and
Dapaong, respectively. Indeed, a significant upward trend September (Dapaong). A significant downward trend is
in annual and rainy season rainfall is detected at Dapaong. detected in May in Kara. The results reveal intra- and
In addition, a general upward trend in intra-annual rainfall interannual variability in rainfall.

Spatio-Temporal Dynamics of Climatic Parameters in Togo

 Balaka et al. 985

Figure 2: Mann-Kendall standard (Z) statistic and Sen slope magnitude (Qi) for all stations (Precipitation:
mm/year, rainy and dry seasons)

MK (Z) statistic Sen's slope (Qi)

6.00
4.00
2.00
0.00
-2.00
Year RS DS Year RS DS Year RS DS Year RS DS Year RS DS
Lomé Atakpamé Sokodé Kara Dapaong

The general upward trend in annual and seasonal rainfall yields and cultivated areas (Allé et al., 2013; Hounnou et
obtained is similar to the results obtained by al., 2019).
Chattopadhyay and Edwards (2016) in Kentucky, USA,
Analysis of temperature dynamics (maximal and
Aawar et al. (2019) in Afghanistan, Phuong et al. (2019)
minimal)
in Vietnam. Indeed, Phuong et al. (2019) find, like this
study, an upward trend in annual and seasonal rainfall. The primary statistical parameters of the maximum
temperature at the different stations are presented in Table
This study also finds combinations of increasing and 5. The annual and rainy season maximum temperature
decreasing trends at some stations as found by Zamani et ranges from a minimum of 29.07°C and 28.69°C
al. (2018) in Jharkhand, India and Jones et al. (2015) in respectively at Atakpamé to a maximum of 34.90°C and
the United States. The results obtained corroborate those 33.65°C respectively at Dapaong. In dry seasons, the
of Maléki et al. (2014) and Adewi et al. (2010) in Togo. minimum is 29.21 (Dapaong) and the maximum is 41.76
Maléki et al. (2014) found spatio-temporal variability in (Atakpamé). At all stations, August has the lowest average
rainfall in northern Togo from 1960 to 2010 with a general temperatures, ranging from 28.22°C at Atakpamé to
decrease in rainfall. Indeed, considering the five stations 29.66°C at Kara. Slight variations in monthly, annual, rainy
from Sokodé to Dapaong, three (Mango, Kara and and dry season temperatures are observed. The
Sokodé) show a declining trend. Adewi et al. (2010) distribution of maximum annual, wet season and dry
showed a disorganization of rainy seasons with a late season temperatures varies from fairly symmetrical to
onset and early end of useful potential rainfall. This high highly skewed. This distribution is less flattened at most
variability of intra and inter-annual rainfall has negative stations.
effects on agriculture in Togo considering that it is
essentially rainfed. Indeed, rainfall variations affect crop

Table 5: Values of primary statistical parameters of maximum temperatures (°C)

Station Period Min Max Mean SD Cv Cs Ck % chan. n
Lomé Year 30.08 32.63 31.33 0.63 0.02 0.15 2.17 0.03 46
RS 29.90 32.54 31.20 0.61 0.02 0.01 2.14 0.03
Atakpamé Year 29.07 32.50 31.45 0.67 0.02 -1.12 4.89 0.05 46
RS 28.69 32.36 31.28 0.70 0.02 -1.24 5.51 0.06
Sokodé Year 31.85 34.12 32.77 0.52 0..02 0.40 2.52 0.03 46
RS 30.34 32.41 31.22 0.50 0.02 0.54 2.73 0.03
Kara Year 32.65 34.80 33.78 0.53 0.02 -0.08 2.10 0.04 40
RS 30.89 33.52 32.20 0.52 0.02 0.05 3.28 0.03
Dapaong Year 29.69 34.90 33.55 0.85 0.03 -1.96 10.62 -0.02 46
RS 29.56 33.65 31.55 0.91 0.03 -0.10 2.79 -0.02
Year: annual, RS: rainy season, Min: minimum, Max: maximum, SD : standard deviation, Cv: coefficient of variance, Cs:
coefficient of skewness, Ck: coefficient of kurtosis, % chan.: percentage change over the period, n: length of period

Trend analysis (Table 6) shows an increasing trend in maximum temperatures is recorded at all stations except
annual and seasonal maximum temperatures. Indeed, a at Dapaong where a significant downward trend in annual
significant upward trend in annual, wet and dry seasons of maximum temperatures is detected. A significant

Spatio-Temporal Dynamics of Climatic Parameters in Togo

increasing trend in maximum temperatures is also Dapaong where a significant decreasing trend is detected
detected for all months of the year at all stations except in July and December.

Table 6: Mann-kendall's and Sen's tests applied to maximum temperatures (°C)

Stations Lomé Atakpamé Sokodé Kara Dapaong
Statistics MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi
Jan 3.20** 0.03** 2.64** 0.03** 2.59** 0.02** 2.28* 0.03* 0.13 0.00
Fev 3.84*** 0.03*** 2.16* 0.03* 2.48* 0.03* 2.46* 0.03* -0.26 -0.01
Mar 3.73*** 002*** 2.94** 0.03** 3.89*** 0.03*** 2.06* 0.03* 0.6 0.01
Avr 4.39*** 0.04*** 3.50*** 0.03*** 2.39* 0.03* 3.03** 0.04** -0.89 -0.01
May 3.98*** 0.03*** 3.18** 0.03** 3.30*** 0.03*** 2.06* 0.02* -5.97 -0.01
June 3.32*** 0.02*** 4.67*** 0.04*** 3.20** 0.03** 3.97*** 0.04*** -1.36 -0.02
Jlt 3.81*** 0.03*** 4.72*** 0.04*** 4.43*** 0.03*** 4.55*** 0.05*** -1.98* -0.02*
Aug 3.55*** 0.03*** 4.21*** 0.03*** 2.26* 0.01* 3.09** 0.02** -1.38 -0.01
Sep 3.98*** 0.02*** 5.65*** 0.04*** 3.46*** 0.02*** 1.42 0.01 -1.52 -0.01
Oct 5.16*** 0.03*** 4.20*** 0.03*** 2.64** 0.02** 2.67** 0.03** -9.29 -0.02
Nov 4.65*** 0.03*** 3.22** 0.03** 4.31*** 0.03*** 2.91** 0.04** 1.03 0.01
Dec 5.72*** 0.04*** 4.88*** 0.05*** 3.91*** 0.04*** 4.71*** 0.07*** 2.37* 0.03*
Year 4.72*** 0.02*** 5.84*** 0.04*** 4.65*** 0.02*** 4.84*** 0.03*** -1.67+ -0.01+
RS 4.69*** 0.02*** 6.11*** 0.04*** 4.18*** 0.02*** 4.45*** 0.03*** -1.34 -0.01
DS 4.41*** 0.02*** 2.64** 0.03** 4.74*** 0.03*** 5.18*** 0.05*** 0.4 0.00
RS: rainfall season, DR: dry season, Z: normalized statistic of the Mann-Kendall test, Qi: the median slope of Sen, +, *, **
and ***: statistically significant at 10%, 5%, 1% and 0.1% respectively.

Analysis of standardized maximum temperature indices a maximum of 25.20°C and 25.19°C respectively (Lomé).
reveals remarkable interannual fluctuations with a Monthly, annual, rainy season and dry season variations
succession of periods of decreases and increases. A in minimum temperature are small at all weather stations.
warming trend is observed at all weather stations (Figure The annual, wet season and dry season minimum
1). temperature is fairly symmetrical for most stations. The
distribution of annual minimum temperature and dry
With respect to the minimum temperature (Table 7), the season minimum temperature is less flattened for most
minimum annual and rainy season temperatures vary from stations. It is more flattened in the wet seasons.
a minimum of 19.30°C and 20.19°C respectively (Kara) to

Table 7: Values of primary statistical parameters of minimum temperatures (°C)

Stations Periods Min Max Mean SD Cv Cs Ck % chan. n
Lomé Year 22.30 25.20 23.93 0.79 0.03 -0.44 2.23 0.05 46
RS 22.52 25.19 24.05 0.75 0.03 -0.49 2.07 0.05
Atakpamé Year 20.39 22.95 21.59 0.48 0.02 -0.27 3.81 0.04 46
RS 20.33 22.95 21.55 0.49 0.02 -0.26 3.90 0.04
Sokodé Year 19.73 21.89 20.81 0.50 0.02 -0.09 2.63 0.05 46
RS 20.72 22.59 21.52 0.44 0.02 0.36 2.72 0.05
Kara Year 19.30 22.14 21.25 0.55 0.03 -1.26 5.55 0.02 40
RS 20.19 22.78 22.05 0.49 0.02 -1.67 7.09 0.02
Dapaong Year 20.73 23.83 22.53 0.94 0.04 -0.71 2.08 0.02 46
RS 20.84 23.82 22.45 0.65 0.03 -0.43 3.32 0.01
Year: annual, RS: rainy season, Min: minimum, Max: maximum, SD : standard deviation, Cv: coefficient of variance, Cs:
coefficient of skewness, Ck: coefficient of kurtosis, % chan.: percentage change over the period, n: length of period

Trend analysis (Table 8) reveals a general upward and intra-annual level, the results are significant for all months
significant trend in minimum monthly, annual, wet season of the year in Lomé, Atakpamé and Sokodé. They are also
and dry season minimum temperatures at all stations with significant in Dapaong except in May and October. In Kara,
the exception of Dapaong in the wet season and Kara in significant results are recorded only in February, May,
the dry season where the results are not significant. At the July, August and November.

Spatio-Temporal Dynamics of Climatic Parameters in Togo

Balaka et al. 987

Table 8: Mann-kendall’s and Sen’s tests applied to minimum temperatures (°C)

Stations Lomé Atakpamé Sokodé Kara Dapaong
Statistics MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi
Jan 3.20** 0.05** 4.27*** 0.03*** 4.07*** 0.04*** 0.99 0.01 2.91** 0.06**
Fev 3.32*** 0.04*** 3.43*** 0.03*** 4.22*** 0.04*** 1.66+ 0.03+ 2.87** 0.05**
Mar 4.35*** 0.04*** 4.33*** 0.03*** 3.96*** 0.03*** 1.26 0.02 3.71*** 0.05***
Avr 4.82*** 0.04*** 3.59*** 0.02*** 4.33*** 0.02*** 0.34 0.00 2.42* 0.02*
May 5.16*** 0.04*** 5.85*** 0.03*** 5.47*** 0.03*** 2.21* 0.02* 1.36 0.01
June 3.40*** 0.02*** 5.38*** 0.03*** 5.01*** 0.02*** 1.4 0.01 2.19* 0.01*
Jlt 3.77*** 0.02*** 5.08*** 0.03*** 4.54*** 0.02*** 1.94+ 0.02+ 3.01** 0.02**
Aug 4.10*** 0.03*** 4.82*** 0.02*** 4.03*** 0.02*** 2.08* 0.02* 2.20* 0.01*
Sep 4.57*** 0.03*** 4.95*** 0.02*** 5.99*** 0.03*** 1.44 0.01 2.06* 0.01*
Oct 4.45*** 0.03*** 5.07*** 0.02*** 5.69*** 0.02*** 0.87 0.7 7.73 0.01
Nov 6.13*** 0.06*** 5.41*** 0.03*** 3.01** 0.03** 1.88+ 0.03+ 2.53* 0.04*
Dec 4.86*** 0.06*** 5.55*** 0.03*** 330*** 0.03*** 1.53 0.02 2.98** 0.05**
Year 4.67*** 0.03*** 5.47*** 0.02*** 6.04*** 0.02*** 1.96+ 0.01+ 1.77+ 0.01+
RS 5.06*** 0.03*** 5.47*** 0.02*** 5.47*** 0.02*** 2.71** 0.01** 1.18 0.01
DS 377*** 0.03*** 4.27*** 0.03*** 5.35*** 0.03*** 1.45 0.01 2.77** 0.03**
RS: rainfall season, DR: dry season, Z: normalized statistic of the Mann-Kendall test, Qi: the median slope of Sen, +, *, **
and ***: statistically significant at 10%, 5%, 1% and 0.1% respectively.

Analysis of standardized minimum temperature indices increasing trend in maximum and minimum temperatures
reveals remarkable interannual fluctuations with a in India. Mansour et al. (2017) also found an increasing
succession of periods of decreases and increases. An trend in maximum temperature. At Çanakkale city in north-
upward trend has been observed in recent years at all western Turkey, (Kale, 2017a) detected an increasing
stations (Figure 1). trend in annual, seasonal and monthly temperature from
1970 to 2012. Those results of Kale (2017a) confirm those
The results obtained in this study on the spatio-temporal of Ejder et al. (2016a) and Ejder et al. (2016b) in the same
variability of maximum and minimum temperatures are city. The general upward trend in maximum and minimum
consistent with those obtained by Maléki et al. (2014) in temperatures is likely to cause significant damage to
northern Togo, Darshana et Pandey (2013) and Ahmadi et agricultural production and hence food security (Adewi et
al. (2018) in India, Chattopadhyay et Edwards (2016), al., 2010; Maléki et al., 2014; Ropo and Ibraheem, 2017).
Haggag et al. (2016) in Rwanda, Cui et al. (2017) in China,
Analysis of the dynamics of relative humidity
Hounnou et Dedehouanou (2018) in Benin, Asfaw et al.
(maximal and minimal)
(2018) in Ethiopia. The results of this research are also
similar to those obtained in Turkey by several studies;
As shown in Table 9, the maximum annual and rainy
especially those of: Sönmez et Kale (2020) in Filyos river
season relative humidity varies from a minimum of 82.20
basin in the northern part, Kale et al. (2018) on the coast
per cent (Sokodé) and 78.29 per cent (Kara) respectively
of the aegean sea, Kale et al. (2016) in Karamenderes
to a maximum of 97.53 per cent (Lomé) and 98.57 per cent
river basin, Ejder et al. (2016a), Ejder et al. (2016b) and
(Sokodé) respectively. In dry seasons, the minimum is
(Kale, 2017a) at Çanakkale province.
35.52% (Atakpamé and Dapaong) and the maximum is
Indeed, Maléki et al. (2014) found an upward trend in
97.89% (Lomé). A small variation in maximum relative
temperature at all the weather stations in northern Togo
humidity is observed in all stations. Its annual and dry
that they studied. Hayelom et al. (2017), in the case of
season distribution is fairly symmetrical for most stations.
Ethiopia, also found an increasing trend in maximum
In rainy seasons, it is moderately biased and strongly
temperatures. On the other hand, these authors found a
biased. In addition, the distribution of annual and dry
decreasing trend in minimum temperatures. Aawar et al.
season maximum relative humidity is less flattened for
(2019) found that minimum temperatures increased
most stations. In wet seasons it is more flattened.
significantly in the Kabul River sub-basin in Afghanistan
from 2008 to 2018. Sonali and Kumar (2016) found an

Table 9: Values of primary statistical parameters of maximum relative humidity (%)

Stations Période Min Max Mean SD Cv Cs Ck % chan, n
Lomé Year 92.33 97.53 95.48 1.36 0.01 -0.56 2.49 -0.02 36
RS 92.75 97.64 95.67 1.39 0.02 -0.55 2.35 -0.02
Atakpamé Year 88.17 95.00 91.79 1.66 0.02 0.04 2.41 -0.02 36
RS 90.91 95.91 93.50 1.12 0.01 0.11 2.61 -0.02

Spatio-Temporal Dynamics of Climatic Parameters in Togo

Sokodé Year 82.80 91.33 87.41 2.24 0.03 -0.18 2.41 0.01 36
RS 94.74 98.57 97.19 0.81 0.01 -0.76 3.56 0.00
Kara Year 66.08 86.69 81.21 3.36 0.04 -2.36 12.55 0.00 36
RS 78.29 97.00 93.97 3.11 0.03 -3.65 19.24 -0.03
Dapaong Year 88.17 95.00 91.79 1.66 0.02 -0.12 2.21 -0.01 28
RS 90.91 95.91 93.50 1.12 0.01 -1.40 4.95 -0.10
Year: annual, RS: rainy season, Min: minimum, Max: maximum, SD : standard deviation, Cv: coefficient of variance, Cs:
coefficient of skewness, Ck: coefficient of kurtosis, % chan.: percentage change over the period, n: length of period

The results of the Mann-Kendall’s and Sen (1968) tests relative humidity in the rainy seasons is -0.33% in
are presented in Table 10. They show a general downward Dapaong. A general downward trend in maximum relative
trend in annual maximum relative humidity in both wet and humidity is also detected at the interannual level. Indeed,
dry seasons. Indeed, a significant downward trend in a significant downward trend is recorded in Lomé (January
maximum annual relative humidity is detected in Lomé and to May and October to December), Atakpamé (April and
Atakpamé. In the rainy seasons, the trend is also October), Sokodé (May), Kara (May, June, August and
downward in all stations and significant except in Sokodé. September) and Dapaong (May to August).
The maximal rate of change in the values of maximum

Table 10: Mann-kendall’s and Sen’s tests applied to maximum relative humidity (%)
Stations Lomé Atakpamé Sokodé Kara Dapaong
Statistiques MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi
Jan -2.53* -0.12* -0.57 -0.13 0.22 0.06 1.08 0.17 -0.04 -0.03
Fev -3.06** -0.12** 0.01 0.00 0.52 0.13 0.41 0.06 0.29 0.06
Mar -3.84*** -0.12*** -0.08 0.00 0.48 0.04 -0.26 -0.06 0.84 0.29
Avr -3.69*** -0.13*** -1.90+ -0.04+ -0.78 -002 -0.86 -0.07 -1.59 -0.17
May -2.68** -0.07** -3.06** -0.03** -1.80+ -0.03+ -1.79+ -0.08+ -2.14* -0.38*
June -1.48 -0.03 -2.86** -0.03** -1.32 -0.01 -1.99* -0.04* -2.06* -0.24*
Jlt 0.98 -0.03 -3.41*** 0.00*** -1.21 -0.01 -1.63 -0.02 -1.73+ -0.13+
Aug -0.59 -0.02 -3.83*** -0.04*** -024 0.00 -2.73** -0.05** -2.76** -0.43**
Sep -1.64 -0.04 -3.82*** -0.03*** -6.71 -0.01 -3.39*** -0.05*** -1.4 -0.1
Oct -2.51* -0.05* -1.89+ -0.04+ -0.61 0.00 -1.61 -0.02 -1.51 -0.17
Nov -2.79** -0.10** -1.15 -0.03 -0.12 0.00 1.41 0.12 0.55 0.17
Dec -4.08*** -0.13*** -1 -0.1 0.27 0.04 1.35 0.19 -0,17 -0.01
Year -2.73** -0.05** -1.88+ -0.06+ 2.84 0.01 -0.15 -0.01 -0.17 -0.02
RS -2.36* -0.04* -2.83** -0.05** -1.07 -0.01 -2.60** -0.07** -3.07** -0.33**
DS -2.7 -0.06 -0.57 -0.13 0.74 0.09 1.1 0.09 -0.17 -0.01
RS: rainfall season, DR: dry season, Z: normalized statistic of the Mann-Kendall test, Qi: the median slope of Sen, +, *, **
and ***: statistically significant at 10%, 5%, 1% and 0.1% respectively.

With regard to minimum relative humidity (Table 11), the (Dapaong) to a maximum of 70.58% and 72.72%
annual and rainy season minimum relative humidity varies respectively (Lomé). In dry seasons, the minimum is
from a minimum of 33.57% and 47.39% respectively 15.57% (Dapaong) and the maximum is 68.25% (Lomé).

Table 11: Values of primary statistical parameters of minimum relative humidity (%)
Station Période Min Max Mean SD Cv Cs Ck % chan, n
Lomé Year 60.03 70.58 65.81 2.32 0.04 -0.40 3.27 0.00 36
RS 63.20 72.72 68.02 2.05 0.03 0.07 3.10 0.00
Atakpamé Year 47.33 57.25 51.18 1.85 0.04 0.87 4.77 -0.01 36
RS 48.36 58.64 53.32 1.91 0.04 0.29 3.90 0.00
Sokodé Year 43.00 52.83 46.65 2.68 0.06 0.86 2.89 0.01 36
RS 54.84 66.14 60.05 2.40 0.04 0.32 3.12 0.03
Kara Year 37.92 49.00 44.02 2.56 0.06 -0.20 2.73 -0.01 36
RS 50.14 62.29 57.79 2.67 0.05 -0.54 3.23 -0.02
Dapaong Year 33.57 52.67 42.88 4.42 0.10 0.47 3.17 0.03 28
RS 47.39 65.00 59.95 3.83 0.06 -1.14 5.22 -0.06
Year: annual, RS: rainy season, Min: minimum, Max: maximum, SD : standard deviation, Cv: coefficient of variance, Cs:
coefficient of skewness, Ck: coefficient of kurtosis, % chan.: percentage change over the period, n: length of period

Spatio-Temporal Dynamics of Climatic Parameters in Togo

Balaka et al. 989

The variation in minimum relative humidity is generally the other stations in dry seasons. The trend tests (Table
small at all stations. Its annual distribution is fairly 12) do not show an annual trend in both the wet and dry
symmetrical at most stations. It varies from fairly seasons. At the intra-annual level, the results show a
symmetrical to moderately skewed at most stations during general downward trend. Indeed, a significant downward
dry and wet seasons. In addition, its annual distribution is trend is detected in Atakpamé (July, September and
less flattened at most stations. In the wet season, it is more December) and Kara (June). A significant upward trend is
flattened at all stations. It is normal in Kara and less flat in recorded in Sokodé (March, April and October).

Table 12: Mann-Kendall and Sen’s tests applied at minimum relative humidity (%)
Stations Lomé Atakpamé Sokodé Kara Dapaong
Statistiques MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi
Jan 0.37 0.08 -1.1 -0.13 0.24 0.02 0.14 0.00 -0.1 -0.03
Fev 0.71 0.07 -0.01 0.00 1.22 0.16 -1.07 -0.08 0.27 0.06
Mar 0.71 0.03 0.34 0.01 1.99+ 0.23+ 0.53 0.06 0.88 0.15
Avr -0.37 -0.02 -0.26 -0.01 1.83* 0.10* -1.21 -0.13 -0.42 -0.08
May 0.25 0.00 0.51 0.03 0.6 0.04 -0.95 -0.09 -0.62 -0.1
June 1.19 0.06 -0.4 -0.01 1.16 0.07 -2.07* -017* -142 -0.19
Jlt -0.55 -0.03 -1.73+ -0.08+ 1.3 0.07 -0.89 -0.04 -1.2 -0.13
Aug 0.74 0.06 -0.93 -0.04 1.46 0.09 02 0.00 -0.91 -0.11
Sep 0.1 0.00 -1.73+ -0.08+ 0.94 0.03 -0.18 -0.01 0.11 0.00
Oct 0.89 0.02 1.12 0.05 2.44* 0.20* 0.9 0.07 0.13 0.00
Nov 2.04 0.09 0.79 0.05 1.68 0.12 -0.05 0.1 1.62 0.4
Dec -0.16 0.00 -2.35* -0.22* 0.26 0.01 -0.19 0.00 0.46 0.11
Year 0.11 0.00 -0.43 -0.01 0.28 0.01 -1.04 -0.01 0.83 0.05
RS -0.14 0.00 0.23 0.01 136 0.04 0.53 -0.04 -1.26 -0.14
DS 0.91 0.06 -1.1 -0.13 0.31 0.02 053 0.03 0.42 0.04
RS: rainfall season, DR: dry season, Z: normalized statistic of the Mann-Kendall test, Qi: the median slope of Sen, +, *, **
and ***: statistically significant at 10%, 5%, 1% and 0.1% respectively.

The results show strong spatio-temporal variability in humidity in Spain. This decrease in maximum and
maximum and minimum relative humidity. The general minimum relative humidity affects agricultural crop yields
downward trend in maximum and minimum relative (Ali et al., 2017). In fact, low humidity level increases
humidity could explain the general upward trend in transpiration and leads to water deficits in the plant, which
evapotranspiration observed in this study. Indeed, affects photosynthesis. However, it can promote
Espadafor et al. (2011) linked it to a decrease in relative pollination of some crops (Zhang et al., 2017).
Analysis of insolation dynamics

As indicated on Table 13, annual and rainy season (Lomé) and the maximum is 9.81 h (Mango). The
insolation varies from a minimum of 5.39 h (Atakpamé) and distribution of annual sunshine, in the rainy and dry
4.98 h (Sokodé) respectively to a maximum of 8.25 h and seasons is fairly symmetrical in most stations. It is less
7.11 h (Mango). In the dry season, the minimum is 4.88 h flattened in most stations.
Table 13: Values of the primary statistical parameters of insolation (h)
Station Periods Min Max Mean SD Cv Cs Ck % chan, n
Lomé Year 5.71 6.95 6.30 0.31 0.05 0.10 2.40 -0.08 36
RS 5.79 6.98 6.36 0.31 0.05 0.42 2.40 -0.05
Atakpamé Year 5.39 6.87 6.28 0.31 0.05 -0.45 3.41 -0.05 36
RS 5.32 6.72 6.18 0.30 0.05 -0.35 3.23 -0.06
Sokodé Year 6.12 7.10 6.63 0.26 0.04 -0.08 1.96 -0.07 36
RS 4.98 6.33 5.57 0.31 0.06 0.15 2.81 -0.01
Kara Year 5.70 7.56 7.12 0.33 0.05 -2.31 10.14 0.04 36
RS 5.67 6.91 6.33 0.22 0.04 -0.50 5.26 0.01
Mango Year 5.85 8.25 7.20 0.59 0.08 -0.40 2.19 -0.17 36
RS 5.52 7.11 6.40 0.39 0.06 -0.37 2.37 -0.10
Year: annual, RS: rainy season, Min: minimum, Max: maximum, SD : standard deviation, Cv: coefficient of variance, Cs:
coefficient of skewness, Ck: coefficient of kurtosis, % chan.: percentage change over the period, n: length of period

Spatio-Temporal Dynamics of Climatic Parameters in Togo

 J. Agric. Econs. Rural Dev. 990

The Mann-Kendall’s and Sen’s tests (Table 14) show a significant general trend of a decrease in annual, the wet and dry
season’s insolation. Only the Kara station registers a significant upward trend.
Table 14: Mann-Kendall’s and Sen’s tests applied to insolation (h)
Stations Lomé Atakpamé Sokodé Kara Mango
Statistics MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi
Jan -1.53 -0.02 -3.10** -0.04** -3.58*** -0.04*** -1.19 -0.02 -4.37*** -0.09***
Feb -2.69** -0.04** -4.21*** -0.05*** -3.42*** -0.03*** 2.61** 0.00** -3.95*** -0.07***
Mar -2.06* -0.02* -3.06** -0.04** -3.85*** -0.04*** 3.30*** 0.00*** -4.24*** -0.08***
Apr -1.83+ -0.02+ 1.21 0.01 -0.34 0.00 -1.99* 0.00* -2.30* -0.04*
May -1.54 -002 -0.48 0.00 1.32 0.01 3.28*** 0.00*** -3.75*** -0.06***
June -2.81** -0.04** 1.35 0.02 -0.46 -0.01 6.85*** 0.09*** -3.35*** -0.04***
July 0.87 0.02 1.38 0.02 0.00 0.00 -6.67*** -0.04*** -0.48 -0.01
Aug -1.88+ -0.02+ 0.64 0.01 -0.07 0.00 -3.85*** -0.03*** 1.61 0.02
Sep -1.15 -0.01 2.49* 0.02* 0.01 0.00 -4.94*** -0.01*** 1.98* 0.02*
Oct -0.23 0.00 -0.91 -0.01 -1.92+ -0.02+ -3.83*** -0.01*** -3.09** -0.04**
Nov 0.82 0.00 -2.47* -0.02* -1.16 -0.01 -6.11*** -0.01*** -3.72*** -0.04***
Dec 1.05 0.01 -3.55*** -0.05*** -2.73 -0.03 6.65*** 0.18*** -3.24** -0.06**
Year -2.48* -0.01* -2.13* -0.01* -2.79** -0.01** 2.53* 0.01* -4.97*** -0.03***
RS -1.99* -0.01* -2.03* -0.01* -0.34** 0.00** 3.38*** 0.00*** -2.76+ -0.02+
DS -2.72** -0.02** -3.49*** -0.05*** -4.43*** -0.03*** 5.23*** 0.03*** -5.00*** -0.05***
RS: rainfall season, DR: dry season, Z: normalized statistic of the Mann-Kendall test, Qi: the median slope of Sen, +, *, **
and ***: statistically significant at 10%, 5%, 1% and 0.1% respectively.

There is also a general downward trend in insolation at the The results of this study corroborate those of Padma
intra-year level. Indeed, a significant downward trend is Kumari et al. (2007) who observed a downward trend in
detected in Lomé (February to April, June, August), surface solar radiation in India between 1981 and 2004.
Atakpamé (January to March, September, November and The downward trend in insolation affects agricultural yields
December), Sokodé (January to March, October), Kara because light allows carbon fixation and plant life (Godoy
(July to November) and Mango (January to June and Herz et al., 2019).
October to December).

Analysis of the dynamics of mean wind speed

The primary statistical parameters of wind speed are m/s (Lomé, Atakpamé and Dapaong). The wind speed has
presented in Table 15. The average annual and rainy low and moderate annual and rainy season variations.
season wind speed varies from a minimum of 01 m/s They are moderate and high in dry seasons. These
(Atakpamé and Sokodé) to a maximum of 4.08 m/s and variations are high and very high for most months in all
4.13 m/s respectively (Lomé). In the dry season, the stations.
minimum is 00 m/s (Atakpamé) and the maximum is 04
Table 15: Values of primary statistical parameters of wind speed (m/s)
Station Periods Min Max Mean SD Cv Cs Ck n
Lomé Year 2.25 4.083 2.92 0.38 0.13 0.67 3.80 38
RS 2.25 4.125 2.94 0.39 0.13 0.76 3.74
Atakpamé Year 1 3.5 1.94 0.54 0.28 0.75 3.97 39
RS 1 3.455 1.94 0.55 0.28 0.77 3.91
Sokodé Year 1 2 1.30 0.33 0.25 0.73 1.97 39
RS 1 2 1.22 0.30 0.24 1.06 2.84
Kara Year 1.417 2.667 1.93 0.33 0.17 0.64 2.42 39
RS 1.143 2.429 1.72 0.35 0.20 0.40 2.44
Dapaong Year 1.917 2.917 2.32 0.27 0.12 0.18 2.10 36
RS 1.333 2.167 1.77 0.24 0.14 0.14 1.92
Year: annual, RS: rainy season, Min: minimum, Max: maximum, SD : standard deviation, Cv: coefficient of variance, Cs:
coefficient of skewness, Ck: coefficient of kurtosis, % chan.: percentage change over the period, n: length of period

Spatio-Temporal Dynamics of Climatic Parameters in Togo

Balaka et al. 991

The distribution of annual and dry season mean wind Analysis of evapotranspiration dynamics
speeds is moderately skewed at most stations. In the wet
season, it is predominantly fairly symmetrical and The results of the primary statistical parameters of
moderately skewed. In addition, the annual and wet evapotranspiration are compiled in Table 16. Annual and
season distribution is less flattened at most stations. In the rainy season evapotranspiration varies from a minimum of
dry season it is more flattened. Trend analysis is not 1440.00 mm (Lomé) and 819.80 mm (Dapaong)
necessary because the slope of Sen (1968) is zero at all respectively to a maximum of 2096.90 mm (Dapaong) and
weather stations (Darshana and Pandey, 2013) except at 1649.36 mm (Atakpamé). In dry seasons, the minimum is
Sokodé (dry seasons), Niamtougou (June and in dry 133.45 mm (Atakpamé) and the maximum is 1065.20 mm
seasons) and Mango (rainy seasons) where non- (Dapaong). A slight variation of evapotranspiration is
significant downward trends are detected. The low and recorded at all stations. Annual and rainy season
moderate variation in wind speed in the rainy season is evapotranspiration is moderately biased at most stations.
favorable to crops and may be conducive to yield In dry seasons, it is predominantly fairly symmetrical. In
improvement. addition, annual, dry and wet seasons evapotranspiration
is less flattened at most stations.

Table 16: Values of primary statistical parameters of evapotranspiration (mm)

Station Periods Min Max Mean SD Cv Cs Ck % chan n
Lomé Year 1440.00 1689.60 1544.99 54.23 0.04 0.85 3.71 0.00 46
RS 972.70 1128.20 1039.89 38.05 0.04 0.53 3.10 0.00
Atakpamé Year 1676.32 1803.39 1754.58 30.69 0.02 -0.62 2.70 0.03 46
RS 1535.14 1649.37 1609.02 29.65 0.02 -0.71 2.70 0.03
Sokodé Year 1838.70 1989.70 1912.12 38.86 0.02 0.50 2.34 0.02 46
RS 1006.40 1115.50 1066.14 25.68 0.02 0.18 2.66 0.02
Kara Year 1831.00 2070.00 1987.08 46.25 0.02 -1.21 4.94 0.03 37
RS 1048.60 1157.60 1114.25 26.57 0.02 -0.47 2.78 0.01
Dapaong Year 1770.80 2096.90 1884.23 83.99 0.05 0.70 2.96 -0.05 46
RS 819.80 1039.80 901.96 49.27 0.06 0.72 3.35 -0.11
Year: annual, RS: rainy season, Min: minimum, Max: maximum, SD : standard deviation, Cv: coefficient of variance, Cs:
coefficient of skewness, Ck: coefficient of kurtosis, % chan.: percentage change over the period, n: length of period

The results of the Mann-Kendall and Sen (1968) tests, in On the other hand, a significant downward trend (annual
Table 17, indicate a general upward trend in and in rainy seasons) is recorded in Dapaong.
evapotranspiration. The maximum rates of change in
values of annual, in dry and wet seasons There is also a general upward Intra-annual trend in
evapotranspiration are -1.86 mm/year (Dapaong), 1.10 evapotranspiration. Indeed, a significant upward trend is
mm/season (Kara) and -2.18 mm/season (Dapaong), recorded in Atakpamé (March, April, and September to
respectively. Indeed, an annual (Atakpamé, Sokodé and December), Sokodé (January, March, June and
Kara), rainy seasons (Atakpamé) and dry seasons December), Kara (January to March, June, and
(Sokodé and Kara) significant upward trend is detected. November). On the other hand, in Dapaong, a significant
downward trend is detected from January to November.
Table 17: Mann-kendall’s and Sen’s test applied to evapotranspiration (mm)
Stations Lomé Atakpamé Sokodé Kara Dapaong
Statistics MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi MK (Z) Qi
Jan 0.98 0.1 0.81 0.04 1.73+ 0.12+ 2.63** 0.22** -2.44* -0.23*
Fev 0.01 0.00 1.36 0.07 1.42 0.09 1.66+ 0.14+ -3.30*** -0.27***
Mar -1.05 -0.06 2.16* 0.13* 3.26** 0.21** 1.73+ 0.18+ -2.14* -0.17*
Avr 0.07 0.01 2.03* 0.14* 0.57 0.04 -0.01 -0.01 -2.96** -0.30**
May 0.24 0.02 1.34 0.08 1.27 0.08 0.6 0.1 -2.16* -0.22*
June -1.07 -0.06 1.16 0.08 1.65+ 0.09+ 1.67+ 0.14+ -3.26** -0.27**
Jlt 0.15 0.01 1.59 0.12 1.41 0.08 1.62 0.09 -4.74*** -0.49***
Aug -0.6 -0.04 0.74 0.04 0.24 0.02 -0.9 -0.08 -5.31*** -0.45***
Sep -0.01 0.00 2.84* 0.14* 1.18 0.06 -0.59 -0.07 -4.25*** -0.30***
Oct 0.44 0.03 2.10** 0.10** 1.44 0.09 -0.1 -0.02 -4.83*** -0.40***
Nov 0.01 0.00 1.87+ 0.09+ 1.59 0.11 1.88+ 0.22+ -2.42* -0.18*
Dec 0.62 0.05 4.16*** 0.22*** 2.94** 0.17** 3.2 0.3 -0.81 -0.05
Spatio-Temporal Dynamics of Climatic Parameters in Togo
Year 0.3 0.1 2.89** 0.96** 1.95+ 0.75+ 2.50* 1.40* -3.37*** -1.86***
RS 0.07 0.03 2.79** 0.88** 1.48 0.41 0.54 0.29 -5.25*** -2.18***
DS 0.11 0.03 0.81 0.04 2.87** 0.62** 3.60*** 1.10*** -1.59 -0.32
RS: rainfall season, DR: dry season, Z: normalized statistic of the Mann-Kendall test, Qi: the median slope of Sen, +, *, **
and ***: statistically significant at 10%, 5%, 1% and 0.1% respectively.

Figure 3: Mann-Kendall standard (Z) statistic and Sen slope magnitude (Qi) for all stations (Evapotranspiration :
mm/year, rainy and dry seasons)

MK (Z) statistic Sen's slope (Qi)

6.00
4.00
2.00
0.00
-2.00
-4.00
-6.00
Year RS DS Year RS DS Year RS DS Year RS DS Year RS DS
Lomé Atakpamé Sokodé Kara Dapaong

The general upward trend in evapotranspiration obtained evapotranspiration, relative humidity (maximum and
in this study is similar to the results obtained by Chaouche minimum), insolation and wind speed in six (06) of the ten
et al. (2010) in Pyrenees-Orientales and Aude in France, (10) meteorological stations in Togo were examined at
El-Nesr (2010) in the Arabian Peninsula, Ibrahim (2012) in intra and inter-annual level. The non-parametric methods
Burkina Faso, Onyutha (2016) in the Nile basin, Jhajharia of Mann-Kendall (MK) and Sen (1968) were used. The
et al. (2015) and Sonali et Kumar (2016) in India. The "Pre-Whitening" approach prior to estimation was used
results of this research corroborate those obtained in where appropriate. The results revealed a high spatio-
Turkey by several studies; especially those of: Kale et al. temporal variability of all meteorological variables. General
(2018) on the coast of the Aegean sea, Kale et al. (2016) upward trends in monthly, annual, wet and dry season’s
at Karamenderes river basin, Ejder et al. (2016a), Ejder et precipitation were detected. Precipitation is moderate and
al. (2016b) and Kale (2017b) at Çanakkale province. high or even very high in some stations. A general upward
Inded, Kale (2017b) found an increasing upward trend in trend in monthly, annual, wet and dry seasons temperature
annual, seasonal and monthly evaporation. Kale et al. (maximum and minimum) is recorded at all stations except
(2018) results are similar to those obtained by Kale Dapaong. Remarkable inter-annual temperature
(2017b) and confirm those obtained by Ejder et al. (2016a) fluctuations (maximum and minimum) and a warming trend
et de Ejder et al. (2016b). This study shows, in addition to are observed at all stations. Evapotranspiration also
the general upward trend in evapotranspiration, spatial shows a general intra-annual, annual, in wet and dry
and temporal heterogeneities over the whole country as seasons upward trend. On the other hand, a general
Mansour et al. (2017) in Tunisia, Obada et al. (2017) in downward trend in maximum and minimum relative
Benin. This spatio-temporal variability of humidity is detected at the intra-annual, annual, wet
evapotranspiration is explained by its complex relationship season and dry season levels. Also, a general downward
with other meteorological variables such as temperature, trend in insolation is recorded at the intra-annually,
relative humidity, insolation and wind speed (Yin et al., annually, in the rainy and dry seasons levels. Wind speed,
2010). Thus, the general upward trend in on the other hand, shows high and very high intra-annual
evapotranspiration observed in this study could be variations in all weather stations. These variations are low
explained by the general upward trend in maximum and and moderate at the annual level and in the rainy seasons.
minimum temperature as found by Obada et al. (2017) and They are moderate and high in dry seasons.
Sonali and Kumar (2016). Increased evapotranspiration
reduces the availability of water resources with Increases in temperature, precipitation and
consequences for agricultural crops. evapotranspiration could have negative effects on
agricultural production and ecosystems. It is more likely to
lead to floods, periods of drought, rising sea levels and
CONCLUSION high winds. Heavy heat and abundant rainfall are
supposed to increase the occurrence of plant diseases,
In this study, the spatio-temporal variability and trend of pests, vector-borne and infectious diseases with
rainfall, temperature (maximum and minimum), consequences for food security. This study has the merit

Spatio-Temporal Dynamics of Climatic Parameters in Togo

Balaka et al. 993

of analyzing at the local level, the spatio-temporal variation Birsan, M.-V., Molnar, P., Burlando, P., & Pfaundler, M.
of meteorological variables. These variables influence (2005). Streamflow trends in Switzerland. Journal
plant growth and agricultural yields with consequences on of hydrology, 314(1–4), 312–329.
food security. Chaouche, K., Neppel, L., Dieulin, C., Pujol, N., Ladouche,
B., Martin, E., Salas, D., & Caballero, Y. (2010).
Analyses of precipitation, temperature and
REFERENCES evapotranspiration in a French Mediterranean
region in the context of climate change. Comptes
Aawar, T., Khare, D., & Singh, L. (2019). Identification of Rendus Geoscience, 342(3), 234–243.
the trend in precipitation and temperature over the Chattopadhyay, S., & Edwards, D. R. (2016). Long-term
Kabul River sub-basin : A case study of trend analysis of precipitation and air temperature
Afghanistan. Modeling Earth Systems and for Kentucky, United States. Climate, 4(1), 10.
Environment, 5(4), 1377–1394. Coulibaly, N., Coulibaly, T. J. H., Mpakama, Z., & Savané,
Adewi, E., Badameli, K., & Dubreuil, V. (2010). Évolution I. (2018). The impact of climate change on water
des saisons des pluies potentiellement utiles au resource availability in a trans-boundary basin in
Togo de 1950 à 2000. Climatologie, Volume 7, West Africa : The case of Sassandra. Hydrology,
89–107. 5(1), 12. https://doi.org/https ://doi.org/10.3390/
Agnew, C. T., & Chappell, A. (1999). Drought in the Sahel. hydro logy5 01001 2
GeoJournal, 48(4), 299–311. Cui, L., Wang, L., Lai, Z., Tian, Q., Liu, W., & Li, J. (2017).
Ahmad, I., Tang, D., Wang, T., Wang, M., & Wagan, B. Innovative trend analysis of annual and seasonal
(2015). Precipitation trends over time using Mann- air temperature and rainfall in the Yangtze River
Kendall and spearman’s rho tests in swat river Basin, China during 1960–2015. Journal of
basin, Pakistan. Advances in Meteorology, Atmospheric and Solar-Terrestrial Physics, 164,
2015(ID 431860). 48–59.
http://dx.doi.org/10.1155/2015/431860 Darshana, P., Ashish, & Pandey, R. P. (2013). Analysing
Ahmadi, F., Nazeri Tahroudi, M., Mirabbasi, R., Khalili, K., trends in reference evapotranspiration and
& Jhajharia, D. (2018). Spatiotemporal trend and weather variables in the Tons River Basin in
abrupt change analysis of temperature in Iran. Central India. Stochastic Environmental Research
Meteorological Applications, 25(2), 314–321. and Risk Assessment, 27(6), 1407–1421.
Alemayehu, A., & Bewket, W. (2017). Local spatiotemporal https://doi.org/10.1007/s00477-012-0677-7
variability and trends in rainfall and temperature in Easterling, D. R., Kunkel, K. E., Wehner, M. F., & Sun, L.
the central highlands of Ethiopia. Geografiska (2016). Detection and attribution of climate
Annaler: Series A, Physical Geography, 99(2), extremes in the observed record. Weather and
85–101. Climate Extremes, 11, 17‑27.
Ali, R. O., & Abubaker, S. R. (2019). Trend analysis using https://doi.org/10.1016/j.wace.2016.01.001
mann-kendall, sen’s slope estimator test and Ejder, T., Kale, S., Acar, S., Hisar, O., & Mutlu, F. (2016a).
innovative trend analysis method in Yangtze river Effects of climate change on annual streamflow of
basin, china. International Journal of Engineering Kocabaş Stream (Çanakkale, Turkey). Journal of
&Technology, 8(2), 110‑119. https://doi.org/DOI: Scientific Research and Reports, 11(4), 1–11.
10.14419/ijet.v7i4.29591 Ejder, T., Kale, S., Acar, S., Hisar, O., & Mutlu, F. (2016b).
Ali, S., Liu, Y., Ishaq, M., Shah, T., Ilyas, A., & Din, I. U. Restricted effects of climate change on annual
(2017). Climate change and its impact on the yield streamflow of Sarıçay stream (Çanakkale,
of major food crops : Evidence from Pakistan. Turkey). Marine Science and Technology Bulletin,
Foods, 6(6), 39. 5(1), 7–11.
Allé, C. S. U. Y., Afouda, A. A., Agbossou, K. E., & Guibert, El-Nesr. (2010). Analysis of Evapotranspiration Variability
H. (2013). Evolution des descripteurs and Trends in the Arabian Peninsula. American
intrasaisonniers des saisons pluvieuses au sud- Journal of Environmental Sciences, 6(6), 535–
Bénin entre 1951 et 2010. American Journal of 547. https://doi.org/10.3844/ajessp.2010.535.547
Scientific Research, 94, 55–68. Espadafor, M., Lorite, I. J., Gavilán, P., & Berengena, J.
Asfaw, A., Simane, B., Hassen, A., & Bantider, A. (2018). (2011). An analysis of the tendency of reference
Variability and time series trend analysis of rainfall evapotranspiration estimates and other climate
and temperature in northcentral Ethiopia : A case variables during the last 45 years in Southern
study in Woleka sub-basin. Weather and Climate Spain. Agricultural Water Management, 98(6),
Extremes, 19, 29–41. 1045‑1061.
https://doi.org/10.1016/j.wace.2017.12.002 https://doi.org/10.1016/j.agwat.2011.01.015
Gebre, H., Kindie, T., Girma, M., & Belay, K. (2013). Trend
and variability of rainfall in Tigray, northern
Spatio-Temporal Dynamics of Climatic Parameters in Togo
 J. Agric. Econs. Rural Dev. 994

Ethiopia : Analysis of meteorological data and hydrologiques sur le bassin du Nakanbé [PhD
farmers’ perception. Academia Journal of Thesis]. Université Pierre et Marie Curie.
Agricultural Research, 1(6), 088–100. Jhajharia, D., Pandey, P. K., Dabral, P. P., Kumar, R., &
Godoy Herz, M. A., Kubaczka, M. G., Brzyżek, G., Servi, Choudhary, R. (2015). Variability in Temperature
L., Krzyszton, M., Simpson, C., Brown, J., and Potential Evapotranspiration over West Siang
Swiezewski, S., Petrillo, E., & Kornblihtt, A. R. in Arunachal Pradesh, India. Jour. Ind. Geol.
(2019). Light Regulates Plant Alternative Splicing Cong, 7(2), 37–43.
through the Control of Transcriptional Elongation. Jones, J. R., Schwartz, J. S., Ellis, K. N., Hathaway, J. M.,
Molecular Cell, 73(5), 1066-1074.e3. & Jawdy, C. M. (2015). Temporal variability of
https://doi.org/10.1016/j.molcel.2018.12.005 precipitation in the Upper Tennessee Valley.
Haggag, M., Kalisa, J. C., & Abdeldayem, A. W. (2016). Journal of Hydrology: Regional Studies, 3, 125–
Projections of precipitation, air temperature and 138. https://doi.org/10.1016/j.ejrh.2014.10.006
potential evapotranspiration in Rwanda under Kale, S. (2017a). Climatic Trends in the Temperature of
changing climate conditions. African Journal of Çanakkale City, Turkey. Natural and Engineering
Environmental Science and Technology, 10(1), Sciences, 2(3), 14–27.
18–33. https://doi.org/10.28978/nesciences.348449
Hamed, K. H., & Rao, A. R. (1998). A modified Mann- Kale, S. (2017b). Analysis of Climatic Trends in
Kendall trend test for autocorrelated data. Journal Evaporation for Çanakkale (Turkey). Middle East
of hydrology, 204(1–4), 182–196. Journal of Science, 3(2), 1–14.
Hare, W. (2003). Assessment of knowledge on impacts of https://doi.org/10.23884/mejs.2017.3.2.01
climate change-contribution to the specification of Kale, S., Ejder, T., Hisar, O., & Mutlu, F. (2016). Climate
art. 2 of the UNFCCC : Impacts on ecosystems, Change Impacts On Streamflow of Karamenderes
food production, water and socio-economic River (Çanakkale, Turkey). Marine Science and
systems. Technology Bulletin, 5(2), 1–6.
Hartmann, D. L., Tank, A. M. K., Rusticucci, M., Alexander, Kale, S., Hisar, O., Sönmez, A. Y., Mutlu, F., & Filho, W. L.
L. V., Brönnimann, S., Charabi, Y. A. R., Dentener, (2018). An assessment of the effects of climate
F. J., Dlugokencky, E. J., Easterling, D. R., & change on annual streamflow in rivers in Western
Kaplan, A. (2013). Observations : Atmosphere Turkey. International Journal of Global Warming,
and surface. In Climate change 2013 the physical 15(2), 190–211.
science basis : Working group I contribution to the https://doi.org/10.1504/IJGW.2018.092901
fifth assessment report of the intergovernmental Kate, S., Hounmenou, C. G., Amagnide, A., & Sinsin, B.
panel on climate change (p. 159–254). Cambridge (2015). Effets des changements climatiques sur
University Press. les activitees agricoles dans la commune de
Hayelom, B., Chen, Y., Marsie, Z., & Negash, M. (2017). banikoara (nord benin). e-Journal of Science &
Temperature and precipitation trend analysis over Technology, 10(2).
the last 30 years in Southern Tigray Regional Kendall, M. G. (1975). Rank Correlation Methods, Book
State, Ethiopia. Series, Charles Griffin. Oxford University Press,
https://doi.org/doi:10.20944/preprints201702.001 USA, London.
4.v1 Maléki, B. H., Kpérkouma, W., Komlan, B., Koffi, A., &
Hobbins, M. T., Ramírez, J. A., & Brown, T. C. (2004). Sidéra, E. K. (2014). Analyse de la variabilité
Trends in pan evaporation and actual temporelle et spatiale des séries climatiques du
evapotranspiration across the conterminous U.S. : nord du togo entre 1960 et 2010. European
Paradoxical or complementary? Geophysical Scientific Journal, 10(11).
Research Letters, 31(13), n/a-n/a. Mann, H. B. (1945). Nonparametric tests against trend.
https://doi.org/10.1029/2004GL019846 Econometrica: Journal of the Econometric
Hounnou, E. F., & Dedehouanou, H. (2018). Variability of Society, 245–259.
temperature, precipitation and potential Mansour, M., Hachicha, M., & Mougou, A. (2017). Trend
evapotranspiration time series analysis in analysis of potential evapotranspiration case of
Republic of Benin. IJAER, 4, 991–1019. Chott-Meriem Region (The Sahel of Tunisia).
Hounnou, F. E., Dedehouanou, H., Zannou, A., Bakary, S., International Journal of Agriculture Innovations
& Mahoussi, E. F. (2019). Influence of Climate and Research, 5(5), 703–708.
Change on Food Crop Yield in Benin Republic. Meena, P. K., Khare, D., Shukla, R., & Mishra, P. K.
Journal of Agricultural Science, 11(5). (2015). Long term trend analysis of mega cities in
Ibrahim, B. (2012). Caractérisation des saisons de pluies northern India using rainfall data. Indian Journal of
au Burkina Faso dans un contexte de changement Science and Technology, 8(3), 247.
climatique et évaluation des impacts MERF. (2015). Troisième communication nationale sur le
changement climatique au Togo (p. 160).
Spatio-Temporal Dynamics of Climatic Parameters in Togo
Balaka et al. 995

Obada, E., Alamou, E. A., Chabi, A., Zandagba, J., & contributions to economics and econometrics (p.
Afouda, A. (2017). Trends and Changes in Recent 345–381). Springer.
and Future Penman-Monteith Potential Von Storch, H. (1999). Misuses of statistical analysis in
Evapotranspiration in Benin (West Africa). climate research. In Analysis of climate variability :
Hydrology, 4(3), 38. Applications of Statistical Techniques (p. 11–26).
Oliver, J. E. (1980). Monthly precipitation distribution : A Springer.
comparative index. The Professional Geographer, Wang, W., Xing, W., Shao, Q., Yu, Z., Peng, S., Yang, T.,
32(3), 300–309. Yong, B., Taylor, J., & Singh, V. P. (2013).
Onyutha, C. (2016). Statistical analyses of potential Changes in reference evapotranspiration across
evapotranspiration changes over the period 1930– the Tibetan Plateau : Observations and future
2012 in the Nile River riparian countries. projections based on statistical downscaling.
Agricultural and forest meteorology, 226, 80–95. Journal of Geophysical Research: Atmospheres,
Padma Kumari, B., Londhe, A. L., Daniel, S., & Jadhav, D. 118(10), 4049–4068.
B. (2007). Observational evidence of solar Yin, Y., Wu, S., Chen, G., & Dai, E. (2010). Attribution
dimming : Offsetting surface warming over India. analyses of potential evapotranspiration changes
Geophysical Research Letters, 34(21). in China since the 1960s. Theoretical and Applied
Pandey, B. K., & Khare, D. (2018). Identification of trend in Climatology, 101(1–2), 19–28.
long term precipitation and reference Yong-fang, Z., Jun-li, D., De-xin, G., Chang-jie, J., An-zhi,
evapotranspiration over Narmada river basin W., Jia-bing, W., & Feng-hui, Y. (2011).
(India). Global and Planetary Change, 161, 172– Spatiotemporal changes of potential
182. https://doi.org/https :// doi.org/10.1016/j.glopl evapotranspiration in Songnen Plain of Northeast
acha.2017.12.017 China. Yingyong Shengtai Xuebao, 22(7).
Parry, M. L., Canziani, O. F., Palutikof, J. P., Van Der Yue, S., Pilon, P., Phinney, B., & Cavadias, G. (2002). The
Linden, P. J., & Hanson, C. E. (2007). IPCC, influence of autocorrelation on the ability to detect
2007 : Climate change 2007: impacts, adaptation trend in hydrological series. Hydrological
and vulnerability. Contribution of working group II processes, 16(9), 1807–1829.
to the fourth assessment report of the Yue, S., & Wang, C. (2004). The Mann-Kendall test
intergovernmental panel on climate change. modified by effective sample size to detect trend
Cambridge Uni-versity Press, Cambridge, UK. in serially correlated hydrological series. Water
Phuong, D. N. D., Linh, V. T., Nhat, T. T., Dung, H. M., & resources management, 18(3), 201–218.
Loi, N. K. (2019). Spatiotemporal variability of Zamani, R., Mirabbasi, R., Nazeri, M., Meshram, S. G., &
annual and seasonal rainfall time series in Ho Chi Ahmadi, F. (2018). Spatio-temporal analysis of
Minh city, Vietnam. Journal of Water and Climate daily, seasonal and annual precipitation
Change, 10(3), 658–670. concentration in Jharkhand state, India.
https://doi.org/10.2166/wcc.2018.115 Stochastic Environmental Research and Risk
Ropo, O. I., & Ibraheem, A. A. (2017). Response of Assessment, 32(4), 1085–1097.
cassava and maize yield to varying spatial scales https://doi.org/10.1007/s00477-017-1447-3
of rainfall and temperature scenarios in Port Zhang, P., Zhang, J., & Chen, M. (2017). Economic
Harcourt. Res. J. Environ. Sci, 11, 137–142. impacts of climate change on agriculture : The
Sen, P. K. (1968). Estimates of the regression coefficient importance of additional climatic variables other
based on Kendall’s tau. Journal of the American than temperature and precipitation. Journal of
statistical association, 63(324), 1379–1389. Environmental Economics and Management, 83,
Sonali, P., & Kumar, D. N. (2016). Spatio-temporal 8–31.
variability of temperature and potential Zhang, X., Chen, S., Sun, H., Shao, L., & Wang, Y. (2011).
evapotranspiration over India. Journal of Water Changes in evapotranspiration over irrigated
and Climate change, 7(4), 810–822. winter wheat and maize in North China Plain over
Sönmez, A. Y., & Kale, S. (2020). Climate change effects three decades. Agricultural Water Management,
on annual streamflow of Filyos River (Turkey). 98(6), 1097–1104.
Journal of Water and Climate Change, 11(2), 420–
433. https://doi.org/10.2166/wcc.2018.060
Sun, S., Chen, H., Wang, G., Li, J., Mu, M., Yan, G., Xu,
B., Huang, J., Wang, J., & Zhang, F. (2016). Shift
in potential evapotranspiration and its implications
for dryness/wetness over Southwest China.
Journal of Geophysical Research: Atmospheres,
121(16), 9342–9355.
Theil, H. (1992). A rank-invariant method of linear and
polynomial regression analysis. In Henri Theil’s
Spatio-Temporal Dynamics of Climatic Parameters in Togo
J. Agric. Econs. Rural Dev. 996

Accepted 15 December 2020

Citation: Balaka M.M, Yovo K, Egbendewe A (2021). Spatio-temporal dynamics of climatic parameters in Togo. Journal
of Agricultural Economics and Rural Development, 7(1): 978-996.

Copyright: © 2021: Balaka 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.

Spatio-Temporal Dynamics of Climatic Parameters in Togo