Impact of Meteorological Drought on Streamflows in the Lobo River Catchment at Nibéhibé, Côte d’Ivoire

The management of water resources in watersheds has become increasingly difficult in recent years due to the frequency and intensity of drought sequences. The Lobo River catchment, like most tropical regions, has experienced alternating wet and dry periods. These drought periods have a significant impact on the availability of water resources in the basin. The objective of this study is to analyse the impact of meteorological drought on flows in the Lobo River catchment. Therefore, using the Normalized Precipitation and Evapotranspiration Index (SPEI) and the Drought Flow Index (SDI), the characteristics of droughts were studied. The results of this study show that meteorological droughts were more frequent than hydrological droughts in the Lobo River watershed. However, the hydrological drought was longer and more intense than the meteorological drought. The greater relationship between meteorological and hydrological drought was observed at the Daloa and Vavoua station (0.43 < r < 0.50) compared to the Zuenoula station (r < 0.5). In addition, there was a resumption of precipitation and runoff between 2007 and 2013 in the basin. The study of these climatic trends would be very useful in the choice of management and adaptation policies for water resources management.

For the past decade, the city of Daloa has been facing a crucial problem of water accessibility in terms of availability and quality. Surface water, treated by Water Distribution Company of Côte d'Ivoire, does not serve all the populations, especially those living in rural areas. The rate of access to drinking water for populations is 73% in urban areas and 50% in rural areas [7]. In addition to the decrease in rainfall, the water reservoir built to supply water to the city is invaded by macrophytes, resulting in a degradation of water quality [8]. On the other hand, the massive destruction of forest cover for agriculture and the decline in rainfall after the break-up of the 1970s have had an impact on the availability of water resources, creating water shortages in recent years in the Lobo River catchment area.
Predicting the onset of drought sequences would then be very valuable for water resource managers. Indeed, the complexity of the physical factors involved in their onset is evidence of the difficulties in recognizing and planning for them [9]. According to [10], the current accepted management approach is based on statistical indices. Of all the drought indices, the Standardised Precipitation Index (SPI) and Standardised Drought Index (SDI) are the most widely used to characterise them [11] [12]. They are applied to characterize the intensity, duration, frequency and impacts of meteorological drought on runoff. The SPI index [13] is powerful and simple to calculate. It can be applied on a variety of time scales, and allows for early detection of drought situations. However, the main limitation is that it considers only one parameter during drought index calculations. In response to this limitation of SPIs, [14] has developed a new index (SPI) that takes temperature into account in the calculations. The Standardized Precipitation and Evapotranspiration Index (SPEI) is virtually similar to the SPI, however it only takes into account temperature data for the calculation of potential evapotranspiration [15]. According to [16], the SPEI is the ideal approach to assess meteorological drought.
Thus, in a climatic context marked by a possible increase in temperature and the impact of droughts in the coming years [17], it is essential to analyse drought sequences. It is within this framework that the present study was initiated to analyse the impact of meteorological drought on the availability of surface water in the Lobo catchment area using the SPEI and SDI indices.
B. Koffi et al.

Study Area
The watershed of the Lobo River in Nibéhibé is located between the Longitude 6˚2' and 6˚8' West and between the latitudes 6˚8' and 7˚9' North, with an area of 7280 km 2 in Nibéhibé ( Figure 1). This river is the main tributary of the left bank of the Sassandra River [18]. The Lobo River is a very important resource, especially as a source of drinking water for the city of Daloa, the third largest city in Côte d'Ivoire. Anthropogenic activities in the basin are very diverse. However, agriculture remains the main activity of the people.
The climate of the Lobo watershed is of a transitional equatorial type with a rainy season from March to October and a dry season from November to February ( Figure 2).
where, ET 0 HG = reference evapotranspiration (mm/day); T, T max and T min = mean, maximum and minimum temperature (˚C) respectively. For computing SPEI, we used method of [14]. This index is accumulated at different time scales.
The SPEI is based on a monthly climatic water balance (precipitation minus ET 0 ), which is adjusted using a three-parameter log-logistic distribution. SPEI was calculated using a script on R.3.6.1 software (https://cran.r-project.org/bin/windows/base/).
The deficit or surplus accumulation of a climate water balance (D) at different time scales is determined by the difference between the precipitation (P) and ET 0 for the day i (Equation (2)): The probability density function of a three-parameter log-logistic distributed variable is expressed as Equation (3): where , α β and γ are scale, shape and origin parameters for the values of D, in the rang ( D γ > < ∞ ). The parameters of the Pearson III distribution can be obtained by Equations (4)-(6): where ( ) β Γ is the gamma function of β . The probability distribution function of the D series is presented by Equation (7): In the last step, with the value of F(x), the SPEI can be estimated as the standardized values of F(x). The SPEI equation is calculated by Equation (8): , where P is the probability of exceeding a given D value, , P is replaced by 1-P and the sign of the resulting SPEI is reversed. The constants are C 0 = 2.515517, C 1 = 0.802853, C 2 = 0.010328, d 1 = 1.432788, d 2 = 0.189269 and d 3 = 0.001308.
The mean value of the SPEI is 0, and the standard deviation is 1. The SPEI is a normalized variable, so it can be compared with other SPEI values over time and space. An SPEI of 0 indicates a value corresponding to 50% of the cumulative probability of D, according to a log-logistic distribution.
The classification of drought severity based on SPEI values is presented in Table 1. Negative value of SPEI is an indication of dry condition while the positive value signifies wet condition.

Streamflow Drought Index (SDI)
The Drought Flow Index (SDI) is used to determine the hydrological response of the Lobo River to flow deficits [20]. A monthly volume of flow flowing in a basin over a given period of time Q i,j . With Q i,j , i indicating the hydrological year and j the rank of the month in the hydrological year in which it occurred (Equation 1, 2, ,12; 1, 2 3, 4 , , where V i,k is the cumulative elapsed volume of the hydrological year of the reference period. Based on the cumulative elapsed volume V i,k , the hydrological drought index SDI is defined for each reference period k of the ith water year of the following series by the Equation (11): where k V and k s are respectively the mean and the standard deviation of the cumulative volume passed during the reference period k. This type of index had already been used [22] but it was impossible to solve the problem of non-stationarity of the series because it worked on an annual time scale. The two-parameter log-normal distribution was used. Thus, the SDI index was calculated using Drinc 1.5 software (https://drinc.software.informer.com/1.2/) according to Equation (11).
The logarithm of the flows cumulated with k y the mean and , y k S the standard deviation estimated over a long period of time.
Five (5) states are considered starting from 0 (no drought) to 4 (extreme drought) and are defined through Table 2.

Maximum Duration of Meteorological and Hydrological Drought
Duration is one of the important characteristics of drought. Indeed, if a drought starts quickly under certain climatic regimes, it usually takes at least two to three months before it moves to other regions. It can then persist for months or even years. The duration is calculated as follows: • analyzing a series of data over a long period of time to determine the driest period; • subtract the year of the end of the drought from the year of the initial; the result obtained is added by one (Equation (13)).
With: end A : Year of end of dry period.
initial A : Year of the beginning of the dry period.

Meteorological and Hydrological Drought Intensity
Drought intensity can be defined as the magnitude and severity of the impact of  Figure 3).

Frequency of Meteorological and Hydrological Droughts
The cumulative frequency (F) of drought gives an idea of the occurrence of dry sequences over a period of time. It is obtained by reporting the cumulative number of dry sequences by the total number of rainfall and flow data.
With: n ∫ : Cumulative dry sequence size; N: total size of data.

Analyse Correlation
In order to analyze the relationship between meteorological and hydrological droughts in the Lobo river watershed, the Pearson correlation coefficient between the SPEI and SDI indices was calculated at the stations of Daloa, Vavoua and Zuénoula. According to [23] [24], the Pearson correlation coefficient is a very effective method for the analysis of potential relationships between two independent variables. This coefficient was calculated using software R.

Meteorological and Hydrological Drought in the Lobo Watershed at Nibehibe
The

Hydrological Drought in the Catchment Area of the Lobo River at Nibéhibé
The analysis of hydrological drought shows an alternation of dry and wet pe-

Intensity and Maximum Duration of Meteorological and Hydrological Drought Sequences in the Catchment Area
The

Frequency of Meteorological and Hydrological Drought Intensity Classes in the Catchment Area
Analysis of the frequency of drought episodes observed at the Daloa, Vavoua and Zuénoula and Nibéhibé stations shows that these are between 1.6% and 13.3% ( Figure 6). Periods close to normal (attenuated drought) are higher in the basin. They are between 52.7% and 72.9%. Severe to extreme droughts were rare during the observation period. However, these extreme episodes were most observed at the Zuénoula (5%) and Nibéhibé (3.8%) stations and, compared to the Daloa (1.7) and Vavoua (1.8%) stations.

Relationship between Meteorological (SPEI) and Hydrological (SDI) Drought in the Lobo Watershed at Nibéhibé
The

Discussion
The catchment area of the Lobo River in Nibéhibé has experienced climatic variability, materialised by an alternation of wet and dry years after the break observed around the 1970s with regard to the evolution of the SPEI and SDI indic-  [34] showed that the decades 1970-1979; 1980-1989 and 1990-1999 were dry periods. This decline in rainfall intensified during the 1980s and 1990s before rising slightly between 2007 and 2013.
With regard to hydrological drought, the SDI has shown that the Lobo catchment area is characterized by an attenuated type of drought. The dry period observed during 1983 could be due to an accumulated rainfall deficit in the late 1970s [33] [35]. Similarly, below-average rainfall during the 1980s probably resulted in an attenuated to severe drought in the early 1990s, as shown by the SDI with very low values (≤−2), synonymous with severe to extreme drought. This decline in runoff during the 1980s could be due to high water demand pressure or very low water conditions in the basin. However, there is a recovery in runoff from 1995 to 2013, suggesting that the decrease in precipitation has had little effect on the water available in the basin. According to [36], from the decade 2000 onwards, more or less favourable rainfall conditions are at the origin of the increase in runoff in West African catchment areas. This recovery has also been supported by the work of [18] on the Lobo watershed. According to the results of this work, this increase in flows could be due to the strong degradation of surface conditions, through the intensification of agricultural activities. Indeed, like the forest areas of Côte d'Ivoire, the Lobo watershed is undergoing rapid deforestation due to agricultural activities [37]. Indeed, the work of [38] and [39] Journal of Water Resource and Protection on the Boubo, Agneby and Davo basins clearly showed that a reduction in vegetation cover contributes to an increase in flows.
In order to assess the impact of meteorological drought on flows, a correlative analysis was carried out between meteorological (SPEI) and hydrological (SDI) It is also found that the flow deficit marked by a long drought period (58 months) during the period 1983-1986 is at a faster rate than the rainfall deficit

Conclusion
The objective of this study is to assess the impact of meteorological drought on runoff in the Lobo to Nibéhibé catchment area. The methodological approach Journal of Water Resource and Protection was based on meteorological (SPEI) and hydrological (SDI) drought indices.
These methods enabled us to assess the duration, intensity and frequency of drought sequences in the Lobo to Nibehibe catchment area. The results of this study show that meteorological drought is more frequent than hydrological drought in the basin. However, the hydrological drought has been longer and