ANALYSIS OF CLIMATE VARIABILITY AND ITS INFLUENCES ON SEASONAL RAINFALL PATTERNS ON THE BATÉKÉ PLATEAUS IN THE REPUBLIC OF CONGO-BRAZZAVILLE.

nine (09) variables, highlighted the impact of climate variability on seasonal rainfall patterns. The first principal component ( F1 ) is associated with the factors defining the climatic conditions of the Plateau Batéké. The second main component (F2) does not express clear correlations. Generally speaking, the temperature and relative humidity of the air are factors in the temporal variability of seasonal rainfall regimes on the Plateaux Batéké.

Climate variability, in Africa in general and in the Republic of Congo-Brazzaville in particular, has been widely documented over the past 20 years. However, variables related to rainfall, such as the nine-month rainy seasons, have been very little studied in the Batéké Plateaux zone, which play a major role in regional hydrology. The objective of this study is to investigate the influence of climatic events on seasonal rainfall patterns on the Plateaux Batéké. On the one hand, it involves monitoring the evolution of climatic parameters through the analysis of temperature, relative humidity and rainfall variation. On the other hand, it was a question of comparing the monthly rainfall normals over the period 1987-2016 in order to determine the behaviour of seasonal rainfall patterns in the context of climate variability. The principal component analysis, based on nine (09) variables, highlighted the impact of climate variability on seasonal rainfall patterns. The first principal component ( F1 ) is associated with the factors defining the climatic conditions of the Plateau Batéké. The second main component (F2) does not express clear correlations. Generally speaking, the temperature and relative humidity of the air are factors in the temporal variability of seasonal rainfall regimes on the Plateaux Batéké.

…………………………………………………………………………………………………….... Introduction:-
As the water cycle is one of the major components of climate, the implications of its changes on rainfall patterns are important (Kouassiet al., 2011;Faty, 2017). Precipitation is the most important climate factor for both populations and ecosystems. It is the climatic factor that is the most easily accessible along with temperature. For these reasons, most studies and analyses focus on precipitation much more than on other climate parameters. Characterising the impact of climate variability on seasonal rainfall patterns then becomes essential in order to propose solutions adapted to development projects. The variability of climatic conditions in West and Central Africa in general, and in the Republic of Congo in particular, is no longer in doubt (Mpounzaet al, 2003;Paturelet al, 1997;Servat et al, 1997;Servatet al, 1998;Servat et al, 1999;Ardoin et al, 2003, Ardoin, 2004Kouassi et al, 2008Kouassi et al, , 2010Kouassi et al, , 2011Laraque, 2001). Most of the climate studies carried out in Congo-Brazzaville have been limited to the analysis of annual or monthly rainfall data over the entire territory and in the southern part of the country. On the Batéké Plateaux ( Fig.1), the lack of meteorological stations does not facilitate access to data. This could be justified by the difficulty in acquiring daily measurement data that is reliable and has few gaps. Thus, in the course of this study, we ISSN: 2320-5407 Int. J. Adv. Res. 9(01), 327-337 328 will attempt, on the basis of two (02) stations, to analyse climatic parameters such as temperature or relative air humidity and study their impact on the variability of the precipitation regime.

Equipment:
The material is composed by:

Data available:
The data, at monthly time step, were obtained from the climate-data.org site and from the National Civil Aviation Agency (ANAC) in Brazzaville. They cover the periods 1982 -2012 and 1987-2016 respectively, i.e. two (02) periods of 30 years, which corresponds to the duration recommended by the World Meteorological Organization for climate analyses (WMO, 2011) on two (02) stations. Synoptic: Brazzaville and Djambala stations. Data quality control consisted of looking for outliers (too high or too low or even negative) using the XLSTAT Pro software (Addinsoft, 2016). But, no aberrant data has been identified. Some missing data has been completed with said software. This control result makes it possible to say that the data series used are homogeneous and of good quality for the analyses envisaged. The rainfall (Pt in mm) is measured using the rain gauge, the temperature (° C) is measured using a thermometer while the relative humidity of the air (RH in %) is measured at using the psychrometer and many other parameters such as wind speed on the ground. On the basis of monthly data, the study of the relative homogeneity of the series was carried out with the non-parametric test known as the Mann-Kendall trend test; this test makes it possible to detect the trend and the years of breakage of the series (Sneyers, 1975 andVandiepenbeeck, 1995). The temperature (° C) is measured using a thermometer while the relative humidity of the air (RH in %) is measured using the psychrometer and many other parameters such as wind speed on the ground. On the basis of monthly data, the study of the relative homogeneity of the series was carried out with the non-parametric 329 test known as the Mann-Kendall trend test; this test makes it possible to detect the trend and the years of breakage of the series (Sneyers, 1975 andVandiepenbeeck, 1995). The temperature (° C) is measured using a thermometer while the relative humidity of the air (RH in %) is measured using the psychrometer and many other parameters such as wind speed on the ground. On the basis of monthly data, the study of the relative homogeneity of the series was carried out with the non-parametric test known as the Mann-Kendall trend test; this test makes it possible to detect the trend and the years of breakage of the series (Sneyers, 1975 andVandiepenbeeck, 1995).

Methods:-
The methodological approach is based on acquisition techniques and data processing:

Data processing: a) Climatic parameters:
We calculated the monthly averages on the two (02) stations and on different localities of the study area.

b) Statistical study in Principal Component Analysis (PCA):
A statistical study (Max and min temperature; Max and min relative humidity; Evaporation; Insolation; Wind speed and Flow) using Principal Component Analysis would allow us to observe the impact of precipitation on flows at the Brazzaville station. On the other hand, at the Djambala station, it would make it possible to observe the correlation that exists between climatic factors.

Evolution of climatic parameters: a) Variation of average and annual precipitation levels:
The variation curve of the monthly means of precipitation ( Figure 2) period 1982-2012 of the four (04) localities, shows that the month of October is the rainiest on the whole, with an average of 251 mm of precipitation while the driest month is July, with 22 mm of precipitation. Individually, the locality of Lékana stands out from the others, with April 286 mm and November 277 mm the wettest. However, July remains the driest month.
For the period 1987-2016 (Fig. 2), the two stations present the same appearance as the localities studied. With 284.2 mm the month of November and the month of July 14.1mm, representing the rainiest and driest month respectively.
It can be noted that on the Djambala and Lékana stations the highest rainfall is recorded. These results confirm the proximity of the locality and the synoptic station.
330 It can be noted that the analysis of the pluviometric heights of the two (02) stations for the period from 1987-2016, shows a decreasing trend on the two stations.
This allows us to say that we are therefore witnessing a significant rainfall deficit which leads us to speak in terms of continuous drought (Bouka and al. Mpounza, 2009).
Inter-annual variability appears to be lower at the Djambala station than at the Brazzaville station, especially since 2000.  Temperature is a parameter very important in the evaluation of the flow deficit, which enters into the estimation of the water balance. The climate of the study area results in a long rainy season (September-May) and a dry season (June-August) as shown in Figure 2.There are growing trendlines in the period from 1987-2016 as well (Fig. 4).

c) Relative humidity variations (RH%) monthly interannual mean:
Relative humidity is expressed as a percentage and is defined as the ratio of the amount of water actually contained in the air and the absorption capacity at a given temperature.
Relative air humidity for the period 1990-2016 is high everywhere (Fig. 5); as an annual monthly average, it is equal to 95% on the two stations and presents an increasing curve on the two (02) stations, particularly in September and October (beginning of the wet season).

d) Changes in mean monthly interannual evaporation:
The interannual monthly average of evaporation measured under shelter (Fig. 6) on the synoptic station of Djambala (1990Djambala ( -2015, varies between 83.4 mm in 1992 and 29.8 mm in 1994. It is noted that during the months of June, July and August the evaporation undergoes a slight decrease, but it is always high whatever the month. We can see that the monthly variability has decreased since 1990. This is particularly visible for the month of October which shows a continuous decrease (around 15% over the period).

Statistical study in Principal Component Analysis (PCA):
In order to identify the climatic variables associated with the variability of precipitation, the application of the principal component analysis (PCA) method was undertaken on nine (09) annual average variables which are rain, maximum temperature and minimum, evaporation, maximum and minimum relative humidity, insolation, wind speed and flow at the Brazzaville station. The matrix results presented are the correlation matrices and the eigenvalue matrices.     The results obtained at the Brazzaville station (Fig. 8a) reveal that temperature, evaporation, precipitation and relative humidity of the air are factors that influence the temporal variability of rainfall variables. In fact, the influence of air temperature in all its components is more evident in the dynamics of atmospheric conditions acting on seasonal rainfall variability due to its very negative correlation with the factor F1. The air temperature is followed by the amplitude of variation of precipitation and relative humidity which, on the other hand, is positively correlated with F1. But the flow does not present a clear situation on these two axes. Finally, At the Djambala station (Fig. 9b), there is a correlation only around the F1 axis. The variables evaporation and minimum relative humidity are negatively correlated. Temperature and wind are positively correlated.

335
Regarding the years, we can note that the years 2005-2016 (Fig. 8b) are positively correlated with axis 1. This could allow us to say on the basis of the analysis of the results of figures 2 and 3 there is a general increase in temperature which has been accompanied by a drop in rainfall levels over the past 30 years. These results are generally consistent with those obtained by several authors (Samba 1999(Samba , 2008Servat, , 1999Laraque, 1998Laraque, , 2001;Nguimalet, 2017) The principal component analysis (PCA) with nine (09) variables made it possible to analyze the factors of temporal variability of seasonal climatic regimes on the Batéké Plateaux. It can be remembered that this statistical study highlights the impact of climate variability on seasonal rainfall patterns. The main factor (F1) is associated with the factors defining the atmospheric conditions. The second factor (F2) does not present any clear situation on the two stations. The factors which influence the variability of seasonal rainfall regimes are the rainfall variables which are influenced by the air temperature as well as the relative humidity of the air. We could say,