The study focuses on assessing the individual and combined impacts of climate variability and land use change on hydrological responses. The results indicate that the basin, urban area, cropland, degraded forest, and open forest shows an increasing trend, while gallery forest shows a decreasing trend over the period 1992–2015. Climatic variability is marked by two climatic periods (a wet one from 1980 to 1996 and a dry one from 1997 to 2016) with a 35% decrease in rainfall. Regardless of the state of the landcover used, the simulated mean annual runoff decreases by 67.92% between the wet and dry climate periods. Changes in land use between 1992 and 2015 reduce mean annual runoff by 4.71%. Analysis of the joint effect of climatic and LU variation shows a 68.96% reduction in runoff. In this catchment. The joint impact has a clearly greater effect on runoff than the climatic impact, which is greater than that of human activities. There is a need for policymakers to prioritise sustainable land use practices and integrated water resource management strategies in the area to mitigate the combined effects of climate variability and anthropogenic activities, ensuring the long-term resilience of the ecosystem and water availability for local communities.

  • The SWAT model was used to isolate the climate and land use change impacts on streamflow.

  • Impact of autocorrelation in the study of time series trends using modified Mann–Kendall test.

  • Hydrologic responses follow the direction of joint effects of climate and land use change.

  • Individual and combined impacts of climate and land use changes on hydrologic responses were evaluated.

Water is a vital resource for life and plays a crucial role in sustaining ecosystems and supporting human activities. However, its availability is limited, with freshwater accounting for only 3% of the world's resources, a significant portion of which is stored as ice (Baker et al. 2016; Karimidastenaei et al. 2021). The distribution of water resources across the globe is highly uneven, often not aligning with population distribution, leading to water scarcity issues (Tzanakakis et al. 2020). This scarcity is further exacerbated by the growing global population, making water an increasingly coveted and valuable commodity (Anouma 2020).

The availability of water is not only a global concern but also a regional issue, particularly in Africa. Africa has been facing significant water challenges, with limited access to safe drinking water and sanitation facilities. Water scarcity has become a source of tension between states in the region (Klare 2020; Salameh et al. 2021). Additionally, the temporal variability of water availability poses significant problems. Scarcity can lead to shortages, desertification, and population displacement, while excessive abundance can result in catastrophic floods (Soro 2011).

To address these challenges, it is crucial to develop effective decision-making and management tools that provide a better understanding of the functioning of natural hydrosystems. Numerous studies have been conducted to comprehend the processes and interactions within these systems (Bronstert et al. 2020; Yonaba et al. 2021). These are particularly focused on understanding the variability in climatic conditions of the hydrosystems.

For instance, a study by Arab & Mesgari (2018) analysed the spatial variability of rainfall extremes along geographical coordinates in northwest Iran, offering insights that could be relevant for assessing the potential impacts on water balance, particularly in regions susceptible to climate variability. Similarly, Arab & Gocic (2021a, 2021b) assessed the applicability of rainfall concentration indices over Serbia and provided fundamental insights into rainfall patterns, which are vital for understanding the input variables in hydrological models. In their follow-up study, Amiri & Gocic (2023) examined the temporal and spatial variations of drought over Serbia to understand the applicability of rainfall-based drought indices and provide valuable insights into the potential impacts of climate variability and drought on hydrological processes. Also, Gocic & Arab (2023) conducted an analysis of the spatial variability and patterns of drought, providing fundamental insights that are relevant for assessing the hydrological responses to climate variability in different regions.

However, despite these research efforts and advancements in measurement techniques such as remote sensing and geophysical methods, our knowledge of hydrosystems remains limited. The complexity of hydrosystems arises from the processes occurring within them and the continual transformations and changes in catchment surfaces driven by human activities and climatic conditions (Kouadio et al. 2015).

While researchers have often studied human activities and climatic conditions separately, it is crucial to understand their combined impact, which is significant and often predominant in the hydrological processes of a catchment (Khoi & Suetsugi 2014). Therefore, quantifying the impacts of climate variability and land use dynamics is essential for comprehending catchment hydrology and developing rational water resource management strategies.

Within the transboundary catchment area of the Cavally River at Toulepleu, located between Guinea and Côte d'Ivoire, both climatic disturbances and human activities, including agriculture and gold mining, have led to significant changes in surface conditions (Brou 2019; Yao & Soro 2021a, 2021b). Despite the socio-economic importance of this watershed, limited hydrological studies have been conducted in this region. Understanding the water resources in this area is crucial for rational management and adaptation to global changes.

Based on these observations, it is necessary to analyse recent fluctuations in rainfall and the dynamics of land use to determine their effects on the hydrological balance of the Cavally River catchment at Toulepleu. This study aims to (i) characterise climatic variability and land use dynamics in the basin and (ii) assess the individual and combined impacts of climatic variability and land use on the water balance of the Cavally catchment at Toulepleu. By achieving these objectives, we can enhance our understanding of the hydrological processes in the region and develop effective water resource management strategies to address the challenges posed by climate variability and land use dynamics. The novelty of the present study, therefore, lies in its integrated approach to examining the combined impact of climate variability and land use dynamics on the hydrological balance of the study area. The consideration of human-induced changes and the practical implications for water resource management further distinguish this study from existing studies in the field.

Study area

The Cavally is a transboundary river (Côte d'Ivoire, Guinea, and Liberia), rising in Guinea north of Mount Nimba at an altitude of over 650 m and flowing into the Atlantic Ocean at the border between Liberia and Côte d'Ivoire (Yao & Soro 2021a). For the purposes of this study, the selected outlet is the Toulepleu hydrometric station located Côte d'Ivoire. The watershed area of the River Cavally at Toulepleu straddles Côte d'Ivoire and Guinea and lies between longitudes 8°36′ and 7°75′ West and latitudes 7°57′22″ and 6°30′34″ North (Figure 1). This basin drains an area of 4,387 km2, two-thirds of which is in Côte d'Ivoire and one-third in Guinea. It has a uni-modal climate with two seasons: a dry season from November to March and a rainy season from April to October. The average annual temperature is 25.6 °C, and the average annual rainfall is around 1,120 mm (1980–2016). January is the driest month (average 15 mm), and September is the wettest (average 350 mm of rain).
Figure 1

Geographical location of the Cavally watershed at Toulepleu and hydro-climate stations.

Figure 1

Geographical location of the Cavally watershed at Toulepleu and hydro-climate stations.

Close modal

Model selection

To propose effective water resource management strategies to address the challenges posed by the collective effects of climate variability and land use dynamics, it is essential to employ a hydrological model that has been widely tested and applied in modelling hydrological processes. Evidence from previous studies has shown that the SWAT (Soil and Water Assessment Tool) model is a useful hydrological modelling tool for achieving this goal (Kibii et al. 2021). SWAT was therefore employed for this study because of its wide, successful application across several regions and extensive recognition in the scientific community, as well as its proven ability to address specific requirements of catchment-scale hydrological modelling, integrate land use data, simulate climate variability impacts, and handle data limitations. These factors collectively make SWAT a justified and appropriate hydrological modelling tool for examining the effects of climate variability and land use dynamics on the hydrological balance in the study area.

Data

In order to run the SWAT model, several temporal and spatial inputs were required, including time series of climate data (1980–2016) such as precipitation, maximum and minimum air temperatures, and spatial (grid) data such as land use (1992 and 2015), soil, and topography (e.g., DEM). In addition, flow time (1982–2016) series were required for model calibration and validation. These data were collected from several sources. Details of data types, data resolution, and data sources are summarised in Table 1.

Table 1

Summary of data and sources

Data typesDataResolutionSource
Spatial data DEM 30 × 30 m https://www.usgs.gov/products/maps/overview 
Soil 1: 5,000,000 http://www.fao.org/geonetwork/srv/en/metadata.show%3Fid=14116 
Land use (1992 and 2015) 300 m https://www.esa-landcovercci.org 
Temporal data Rainfall (1980–2016) Seven stations National Directions of Meteorology (Côte d'Ivoire and Guinea) 
Temperature Two stations 
Flow (1982–2016) One stations Hydrologic direction 
Data typesDataResolutionSource
Spatial data DEM 30 × 30 m https://www.usgs.gov/products/maps/overview 
Soil 1: 5,000,000 http://www.fao.org/geonetwork/srv/en/metadata.show%3Fid=14116 
Land use (1992 and 2015) 300 m https://www.esa-landcovercci.org 
Temporal data Rainfall (1980–2016) Seven stations National Directions of Meteorology (Côte d'Ivoire and Guinea) 
Temperature Two stations 
Flow (1982–2016) One stations Hydrologic direction 

Filling gaps

Figure 2 shows the percentage missing rainfall data by stations over 1980–2016. The missing daily rainfall values for Blolequin (8%), Zouan-Hounien (5%) and Danané (4%) stations were filled using ARC2 (Africa Rainfall Climatology Version 2) data (for stations in Côte d'Ivoire) and NASA (National Aeronautics and Space Administration) data (for stations in Guinea) based on regression equation. Before using ARC2 or NASA data in this work, the correlation between these data and the observed data was established over the common period of 1980–2000 for stations with missing data. The correlation coefficients were found to vary between 0.70 and 0.98.
Figure 2

Summary of missing daily rainfall data by station from 1980 to 2016.

Figure 2

Summary of missing daily rainfall data by station from 1980 to 2016.

Close modal

SWAT model set up: calibration and validation

The SWAT model uses the water balance equation for simulation, as follows:
(1)
where SWt is the final soil moisture content, mm; SW0 is the initial soil moisture content of the ith day, mm; t is the time, days; Rday is the precipitation of the ith day, mm; Qsurf is the surface runoff of day i, mm; Ea indicates the amount of evapotranspiration on day i, mm; Wseep indicates the amount of water entering the vadoze zone from the soil profile on day i, mm; and Qgw indicates the return flow amount on day i, mm.
Flow data collected at the gauging station of Toulepleu were used for the calibration (1983–1990) and validation (1998–2001). The sensitivity analysis of the basin was carried out by successive iterations by testing a total of 12 parameters. In this study, the Sequential Uncertainty Fitting version two (SUFI-2) algorithm set in SWAT-CUP 2012 was used during model sensitivity analysis, calibration and validation. The SWAT model was set up as shown in Figure 3.
Figure 3

A schematic diagram shows the SWAT model setup (Sahar et al. 2021).

Figure 3

A schematic diagram shows the SWAT model setup (Sahar et al. 2021).

Close modal
Figure 4

Assessment diagram of the climate variability impact on the water balance.

Figure 4

Assessment diagram of the climate variability impact on the water balance.

Close modal
Figure 5

Assessment diagram of the land use change impact on the water balance.

Figure 5

Assessment diagram of the land use change impact on the water balance.

Close modal
Figure 6

Assessment diagram of the climate variability and land use change impacts on the water balance.

Figure 6

Assessment diagram of the climate variability and land use change impacts on the water balance.

Close modal

The model evaluating decision factors are shown in Table 2.

Table 2

Model evaluation thresholds (Moriasi et al. 2007)

Performance ratingFlow
Very good 0.75 ≤ NSE ≤ 1 PBIAS ≤ ±10% R2 ≥ 0.80 
Good 0.65 ≤ NSE ≤ 0.75 PBIAS ≤ ±15% R2 ≥ 0.70 
Satisfactory 0.50 ≤ NSE ≤ 0.65 PBIAS ≤ ±25% R2 ≥ 0.5 
Unsatisfactory NSE ≤ 0.5 PBIAS ≥ ±25% R2 ≤ 0.5 
Performance ratingFlow
Very good 0.75 ≤ NSE ≤ 1 PBIAS ≤ ±10% R2 ≥ 0.80 
Good 0.65 ≤ NSE ≤ 0.75 PBIAS ≤ ±15% R2 ≥ 0.70 
Satisfactory 0.50 ≤ NSE ≤ 0.65 PBIAS ≤ ±25% R2 ≥ 0.5 
Unsatisfactory NSE ≤ 0.5 PBIAS ≥ ±25% R2 ≤ 0.5 

Impact of climate variability on the parameters of the hydrological balance of the Cavally catchment area at Toulepleu

Climate variability was characterised using the modified Mann–Kendall test (i.e., the Hamed test) and the standard homogeneity test (TSNH). The impacts of this variability were assessed using the 1992 landcover map and the climate period (1980–2016). In other words, the LU_1992 map was the same during the simulation of the water balance parameters before (1980–1996) and after (1997–2016) the breakpoint. Figure 4 summarises how this was done.

Modified Mann–Kendall test (Hamed test)

The Mann–Kendall Test is a widely used method for determining trends in hydrometeorological time series and is recommended by the World Meteorological Organization (WMO). The presence of autocorrelation in the data can seriously affect the power of statistical tests by overestimating the statistical significance of trends. A complementary approach to the classic Mann–Kendall test (Hamed test) is thus proposed in order to take into account this phenomenon of autocorrelation. The principle is based on a modification of Mann–Kendall's S test rather than the data itself:
(2)
is a corrective factor applied to the variance.

Two methods are noted in the literature to estimate this corrective factor.

  • Yue & Wang (2004) suggest correcting the Mann–Kendall test as follows:
    (3)
    denotes the autocorrelation of order 1
  • Hamed & Rao (1998) propose an empirical formula specifically calculated to correct Mann–Kendall's statistics:
    (4)

In this study, the corrective factor from Hamed & Rao (1998) was used. Because the tests carried out by Renard (2006) on the power of these methods show that the modification proposed by Hamed & Rao (1998) is slightly better under the AR (1) hypothesis than the formula of Yue & Wang (2004).

Standard Normal Homogeneity Test or Alexandersson method

The standard normal homogeneity test (SNHT) was introduced by Alexandersson (1986). The test statistic (Tk) is performed to compare the average of the first n year with the average of the last (nk) year with n data points (Vezzoli et al. 2012; Jaiswal et al. 2015. The Tk equation is written as follows:
(5)
Z1 and Z2 can be calculated as follows:
(6)
(7)

Here, and are mean and SD, respectively. The year in which reaches the maximum value is considered as the point of change.

Land use dynamics and impact assessment

For this study, the 16 initial LU classes were reclassified into six (6) major classes: crop or fallow, open forest, degraded forest, dense forest, habitats, and water. The initial and adapted classes are presented in Table 3. This reclassification was based on the ESA-CCI product user guide, version 2, available at the following address https://www.esa-landcover-cci.org/?q=webfm_send/84.

Table 3

LU reclassification summary table

Classes consideredInitial class codesInitial class meanings
Crops and fallow land 10; 11 Cultivated land 
130 Meadows, market gardens… 
Open forests 100 Mosaics of trees, shrubs, grasses 
120 Shrubs 
122 Evergreen or degraded shrubs 
Degraded forests 30 Mosaic of cultivated land (>50%)/natural vegetation (trees, shrubs, herbaceous cover) (<50%) 
40 Mosaic of natural vegetation (trees, shrubs, herbaceous cover) (>50%)/cropland (<50%) 
60; 62 Evergreen tree cover, (<15%) 
Dense forests 50 Evergreen tree cover, (>15%) 
160; 170 Tree cover, fresh, salty or brackish water 
180 Wetland 
Habitat 190 Urban 
Water 210 Water 
Classes consideredInitial class codesInitial class meanings
Crops and fallow land 10; 11 Cultivated land 
130 Meadows, market gardens… 
Open forests 100 Mosaics of trees, shrubs, grasses 
120 Shrubs 
122 Evergreen or degraded shrubs 
Degraded forests 30 Mosaic of cultivated land (>50%)/natural vegetation (trees, shrubs, herbaceous cover) (<50%) 
40 Mosaic of natural vegetation (trees, shrubs, herbaceous cover) (>50%)/cropland (<50%) 
60; 62 Evergreen tree cover, (<15%) 
Dense forests 50 Evergreen tree cover, (>15%) 
160; 170 Tree cover, fresh, salty or brackish water 
180 Wetland 
Habitat 190 Urban 
Water 210 Water 

The dynamics of landcover were assessed by the spatio-temporal variations in the areas of the landcover classes between 1987, 2000, and 2017. The rate of change (Tc) in landcover between these dates was calculated for each landcover class using the following formula (Toyi et al. 2013):
(8)
A1 and A2 are, respectively, the initial and final areas of the land use class.
The average annual rate of change for each land use class was calculated using the following formula:
(9)
r is the annual growth rate of class i; A1 is the area of class i at time t1; A2 is the area of class i at time t2.

The impact of landcover on the water balance was assessed using two LU maps and climate data for the period 1980–2016 (Figure 5). The water balance parameters of each landcover map (LU_1992 and LU_2016) were simulated using the SWAT hydrological model for the same climatic characteristics (1980–2016).

Joint impact of climate variability and variation in LU cover on the hydrological balance of the Cavally catchment area at Toulepleu

Two scenarios (P1 and P2) were considered.

P1 or natural or stable period: climatic period and LU cover prior to the year of disruption. That is the climate of 1980–1996 and the LU of 1992.

P2 or period of change: climatic period and LU cover after the year of disruption. That is the climate of 1997–2016 and the LU of 2015.

NB: Using the SWAT model, the water balance was simulated under both hypotheses. By comparing the results for the two hypotheses, it was possible to determine the impact of P2 on the water balance, taking P1 as the reference. Figure 6 summarises how this was done.

Analysis of rainfall stationarity in the Cavally catchment at Toulepleu

Figure 7 shows the results of the Hamed test for annual precipitation. Confidence levels are selected to be 10, 5, and 1%. The annual precipitation has significant trend at all station. 5/6 of stations are show downward trends and 1/6 (N'Zérékoré station) station shows an upward trend.
Figure 7

Significance of the statistics of the modified Mann–Kendall test (Hamed & Rao 1998) of the annual rainfall series of the Cavally watershed at Toulepleu (1980–2016). Note: Zéré: Zérégbo; Dan: Danané; Blo: Blolequin; Tou: Touleoleu; Zou: Zouan-hounien; Nzé: N'Zérékoré.

Figure 7

Significance of the statistics of the modified Mann–Kendall test (Hamed & Rao 1998) of the annual rainfall series of the Cavally watershed at Toulepleu (1980–2016). Note: Zéré: Zérégbo; Dan: Danané; Blo: Blolequin; Tou: Touleoleu; Zou: Zouan-hounien; Nzé: N'Zérékoré.

Close modal
Figure 8 shows the significance of abrupt changes in Cavally watershed at Toulepleu mean annual rainfall at the watershed from 1980 to 2016. The SNHT test is detected for annual total precipitation in 1996 at 99% confidence level.
Figure 8

Breakpoint in average annual rainfall in the Cavally catchment at Toulepleu.

Figure 8

Breakpoint in average annual rainfall in the Cavally catchment at Toulepleu.

Close modal

Analysis of the LU between 1992 and 2015

The main land use classes are degraded forest, dense forest, crops or fallow land, bare soil or habitats, and open forest. These different classes have seen a variation in their surface areas at the level of each sub-basin from 1992 to 2015 (Figure 9).
Figure 9

Land use map of the Cavally catchment at Toulepleu for the years 1992 and 2015.

Figure 9

Land use map of the Cavally catchment at Toulepleu for the years 1992 and 2015.

Close modal

The landcover maps (Figure 9) show that in 1992, gallery forest represented 17.18%, deciduous forest 35%, cropland 42%, urban 0.09%, and sparse forest 6.18% of the total area of the Cavally catchment at Toulepleu.

In 2015, gallery forest represented 12.76%, cropland increased to 42.61%, deciduous forest 35.18%, sparse forest 9.14% and urban 0.29% of the total area of the Toulepleu basin.

The rate of change (Tc) and average annual rate of change (r) show that in 23 years, the dense forest class has seen its area shrink by 25%, i.e., an annual loss of 1.3% (Table 4). As for the crop and fallow land, bare soil and housing, open forest, and degraded forest classes, their areas have increased by between 0.69 and 200%, i.e., annual losses of between 0.03 and 4.77% depending on the class (Table 4). Despite this, a few pockets of vegetation (dense forests) still exist in the basin, thanks to the presence of parks and reserves (Figure 8).

Table 4

Evolution of the different land use classes and overall changes in the Cavally sub-basins (from 1992 to 2015)

WatershedLand useArea (km2)
Rate (%)
Dynamic (%)
1992201519922015TCr
Touleupleu (4,396 km2Cropland 1,861.11 1,873.95 42.32 42.61 0.690 0.03 
Deciduous forest 1,504.10 1,546.93 34.20 35.18 2.84 0.12 
Galerie forest 755.62 561.04 17.18 12.76 −25.75 −1.29 
Urban 4.28 12.85 0.097 0.29 200.06 4.77 
Sparse forest 271.72 402.06 6.18 9.14 47.97 1.70 
WatershedLand useArea (km2)
Rate (%)
Dynamic (%)
1992201519922015TCr
Touleupleu (4,396 km2Cropland 1,861.11 1,873.95 42.32 42.61 0.690 0.03 
Deciduous forest 1,504.10 1,546.93 34.20 35.18 2.84 0.12 
Galerie forest 755.62 561.04 17.18 12.76 −25.75 −1.29 
Urban 4.28 12.85 0.097 0.29 200.06 4.77 
Sparse forest 271.72 402.06 6.18 9.14 47.97 1.70 

Sensitivity analysis of the hydrological model

The results of the sensitivity analysis of the Cavally basin at Toulepleu (Table 5), carried out by successive iterations, made it possible to identify the predominant influence of certain parameters (3/12) on the reproduction of flows. The most influential parameters have p-values < 0.1. The small number of sensitive parameters to be calibrated is a real challenge since each parameter change must have a physical meaning and requires knowledge of how the catchment functions. These parameters are: the runoff curve number to moisture condition II (CN2), wet bulk density in primary sol (SOL_BD (1)) and river water return coefficient (ALPHA_BNK).

Table 5

Lists of parameters by orders of sensitivity

Order of sensitivityParameter namet-Statp-Value
V__ALPHA_BNK.rte −3.3994 0.0016 
R__CN2.mgt 3.2660 0.0023 
R__SOL_BD(…).sol −2.7997 0.0081 
V__GWQMN.gw 0.6723 0.5055 
A__ESCO.hru 0.6063 0.5480 
V__CH_K2.rte 0.5966 0.5544 
V__CH_N2.rte −0.5374 0.5942 
V__GW_DELAY.gw 0.4415 0.6614 
R__SOL_K(…).sol 0.3982 0.6927 
10 R__SOL_AWC(..).sol −0.3687 0.7145 
11 R__HRU_SLP.hru −0.2975 0.7677 
12 V__ALPHA_BF.gw −0.1254 0.9008 
Order of sensitivityParameter namet-Statp-Value
V__ALPHA_BNK.rte −3.3994 0.0016 
R__CN2.mgt 3.2660 0.0023 
R__SOL_BD(…).sol −2.7997 0.0081 
V__GWQMN.gw 0.6723 0.5055 
A__ESCO.hru 0.6063 0.5480 
V__CH_K2.rte 0.5966 0.5544 
V__CH_N2.rte −0.5374 0.5942 
V__GW_DELAY.gw 0.4415 0.6614 
R__SOL_K(…).sol 0.3982 0.6927 
10 R__SOL_AWC(..).sol −0.3687 0.7145 
11 R__HRU_SLP.hru −0.2975 0.7677 
12 V__ALPHA_BF.gw −0.1254 0.9008 

Model calibration and validation

After calibration of the most sensitive parameters, analysis of the statistical indices NSE, R², and PBIAS shows that the flow simulations are good (Table 6). For the period used for calibration (1983–1990), the values for NSE, R2 and PBIAS are 0.75, 0.79, and −5%, respectively, for monthly flows at the Toulepelu station. The indices evaluated on a monthly scale for the validation period (1998–2001) are 0.52, 0.68, and −8.7% for NSE, R², and PBIAS respectively. It should be noted that there was a sharp deterioration in the performance indices on validation.

Table 6

Model performance at hydrometric stations in a changing climate

StationsCalibration
Validation
NSER2Pbias (%)NSER2Pbias (%)
Toulepleu 0.75 0.79 −5 0.52 0.68 −8.7 
StationsCalibration
Validation
NSER2Pbias (%)NSER2Pbias (%)
Toulepleu 0.75 0.79 −5 0.52 0.68 −8.7 

The analysis of the hydrographs (calibration and validation) confirms that they were carried out on the basis of the performance criteria, where good performance was obtained. Figure 10 shows perfect synchronisation between the simulated and observed flows during calibration and validation.
Figure 10

Hydrograph of observed and simulated flows at the Toulepleu station.

Figure 10

Hydrograph of observed and simulated flows at the Toulepleu station.

Close modal

Hydrological water balance impact of land use change and climate variability

Climate variability on the hydrological water balance

The average annual rainfall in the basin before the rupture was 1,425.66 mm. The surface runoff was estimated at 30.80% of precipitation, actual evapotranspiration was 39.74% of precipitation, and infiltration was 29.38% of precipitation. The average annual rainfall after the rupture was estimated at 919.55 mm. Over this period, the components of the balance were estimated at 15.35% of precipitation for surface runoff, 62.13% of precipitation for actual evapotranspiration, and 22.51% of precipitation for infiltration (Table 7).

Table 7

Evolution of water balance parameters in BVTC at Toulepleu before and after the 1996 rupture

ComponentsBefore (1980–1995)After (1997–2016)Deficit (After-Before in %)
Annual average (mm)Annual average (mm)
Rainfall 1,425.66 919.55 − 35.50 
Surface runoff 440.14 141.18 − 67.92 
Infiltration Eco_L 418.93 49.02 207.02 30.97 − 50.58 −36.82 
Eco_Sou 324.14 138.77 −57.19 
Rech_Aq 18.49 8.81 −52.35 
Eva_Sou 27.28 28.47 4.37 
Evapotranspiration (actual) 566.60 571.35 0.84 
ComponentsBefore (1980–1995)After (1997–2016)Deficit (After-Before in %)
Annual average (mm)Annual average (mm)
Rainfall 1,425.66 919.55 − 35.50 
Surface runoff 440.14 141.18 − 67.92 
Infiltration Eco_L 418.93 49.02 207.02 30.97 − 50.58 −36.82 
Eco_Sou 324.14 138.77 −57.19 
Rech_Aq 18.49 8.81 −52.35 
Eva_Sou 27.28 28.47 4.37 
Evapotranspiration (actual) 566.60 571.35 0.84 

The Cavally River basin at Toulepleu experienced a 35.50 mm reduction in rainfall (mean annual rainfall) between the two periods (before and after the rupture). Mean annual surface runoff from the basin after the rupture fell from 440.14 to 141.18 mm, a reduction of 67.92% compared with the period before the rupture (Figure 11). As with runoff, there was a decrease in parameters dependent on infiltration (lateral flow, groundwater flow, and aquifer recharge), with the exception of groundwater evaporation, which showed a marked increase.
Figure 11

Changes in water balance parameters before and after the break-up year.

Figure 11

Changes in water balance parameters before and after the break-up year.

Close modal

Land use change on the hydrological water balance

The dynamics of the LU experienced by the basin between 1992 and 2015 have slightly modified the components of its water balance (Figure 12).
Figure 12

Changes in balance parameters between LU_1992 and LU_2015.

Figure 12

Changes in balance parameters between LU_1992 and LU_2015.

Close modal

The average annual basin surface runoff in 2015 decreased from 268.29 to 255.66 mm, an 5% decrease from 1995. Conversely, a decrease in annual evaporation from 568.83 to 567.95 mm and a increase in annual infiltration from 280.31 to 293.82 mm were simulated during the two decades. Within these infiltrated quantities, an increase in annual lateral runoff, underground flow, aquifer recharge, and underground evaporation respectively from 37.82 to 38.05 mm, 201.91 to 215.01 mm and 12.60 to 15.78 mm in 2015. Finally, the underground evaporation was same (27.98 mm) in 2015.

Land use change and climate variability on the hydrological water balance

Scenarios P1 and P2 were developed to study the combined impacts of climate variability and land use change on the hydrologic response of the basin. In both scenarios, two LUs (1992 and 201) and two different periods' (after and before breakpoint) climate data were used to evaluate how climate variability and land use change affects the water balance of Toulepleu basin. The average annual precipitation and gallery forest cover decreased by 35.55 and 25.75%, respectively from P1 to P2. Resulting in a decrease in average annual surface runoff (68.96%), in lateral runoff (36.45%), in underground runoff (55.94%), in aquifer recharge (51.20%) and increase in average annual underground evaporation (4.37%) and actual evapotranspiration by less than 1% (Figure 13).
Figure 13

Evolution of balance parameters under natural period (P1) and change period (P2).

Figure 13

Evolution of balance parameters under natural period (P1) and change period (P2).

Close modal

Hydrological model performance

The model's ability to simulate the flow of the Cavally River at Toulepleu was analysed over the period 1980–2016.

The hydrographs showed a good correlation between the observed and simulated monthly mean flows of the river at the Toulepleu station during the calibration period and the validation period. The statistical indices of the SWAT model's performance indicated that the mean monthly flows of the river at Toulepleu were well reproduced by the model. These indices refer to the criteria ‘good’ in calibration and ‘satisfactory’ in validation according to the evaluation threshold of Moriasi et al. (2007). Specifically, for the calibration periods, the values of NSE and R2 are greater than 0.70 and Pbias is less than ±10, and for the validation periods, NSE and R2 are greater than 0.5 and Pbias is less than ±10. In addition, the negative values of PBIAS show that the SWAT model slightly underestimated the flows of the Cavally River at Toulepelu. In general, the SWAT model is reliable for use as a tool for simulating hydrological processes in the Cavally River catchment at Toulepleu. These results given by the model are approved by several authors throughout the world who have tested it for catchments of various sizes and in different geological contexts (Roukia et al. 2019; Hassan et al. 2021; Khoi et al. 2021). In the Ivorian context, N'Dri et al. (2019a, 2019b); and Koua et al. (2019) have demonstrated through their studies the robustness of the model in simulating flows and other balance parameters.

The balance shows an inhomogeneous redistribution of precipitation, with actual evapotranspiration dominating (50%). The importance of this parameter is due to the high vegetation cover combined with an average temperature of 25 °C over the whole basin. It could also be linked to the exposure of soils due to agriculture and the destruction of forest cover (Yao 2015; Roukia et al. 2019).

Runoff in the basin was 268.29 mm/year, i.e., 24% of rainfall. This low proportion of runoff could be linked either to the drop in rainfall in the basin, the high vegetation cover or the morphometric characteristics. According to FAO (2009), the transformation of precipitation into groundwater and watercourses is reduced because it is intercepted by forests and evaporates from the very high forest cover.

Impacts of climate variability on the hydrological functioning of the Cavally catchment at Toulepleu

Analysis of climatic variability shows that annual rainfall has varied over the period 1980–2016, with a downward trend although the N'Zérékoré station is affected by an upward trend. These results also indicate a break year in 1996 linked to the transition from a wet to a dry period, with a rainfall deficit of 35%. All these results show that the Cavally catchment area experienced a variation in rainfall between 1980 and 2016, with a general downward trend. this decrease may be linked to the effects of climate change or to the destruction of landcover. Almost all the studies on climate variability in West Africa (Ilori & Ajayi 2020; Klassou & Komi 2021), particularly in Côte d'Ivoire (Anouma 2020; Yao & Soro 2021a), confirm the general trend towards lower rainfall, which is thought to be due to climate change.

The average annual hydrological balances before and after the disruption show that the drop in rainfall associated with the change from a wet to a dry period led to a reduction in runoff and infiltration. These results indicate that climate variability modifies the hydrological response of the basin. According to N'Guessan et al. (2015) and Anouma (2020), fluctuations in climatic periods could lead to changes in the hydrological response of basins. The work of Setyorini et al. (2017) on the Brantas basin in Indonesia has shown that major changes in climate have had an impact on surface runoff and other hydrological characteristics of the catchment.

Changes in actual evapotranspiration remained practically constant despite the change in climate. This result could mean that this parameter does not depend on rainfall fluctuations but rather on another factor. It could be linked to the conservation of the internal characteristics of the sub-basins, precisely the vegetation cover, during modelling. In fact, during the various simulations (before and after breaks), only the LU_1992 map was used, so there was no variation in the surface condition even though rainfall changed. However, the density of the vegetation cover and the LU mode have a strong influence on the quantity of evapotranspired water, particularly through interception. According to Kouadio et al. (2015), the low variation in evapotranspired water heights over the study period in the face of rainfall variability could be explained by saturation of the evaporative capacity of the atmosphere, which is reached earlier.

Impacts of land use on the hydrological functioning of the Cavally catchment at Toulepleu

Analysis of the statistics for the different LU classes between 1992 and 2015 indicates a regression (≈ 25%) of dense forest in favour of bare soil and dwellings (≈ 200%), open forest (≈ 47.97%), degraded forest (≈ 2.84%), and crops and fallow land (≈ 0.69%). This predominantly forest dynamic is thought to be due to climatic conditions, but above all to the human pressures exerted on the basin. Côte d'Ivoire's economy is based on cocoa and coffee farming. This extensive agriculture is entirely dependent on the good fertility of the forest soils, which has led to massive destruction of the forest. This finding is in line with the work of some researchers in West Africa (Deguy 2021) who each reveal, a regression of vegetation cover in favour of mosaics of crops, fallow land and built-up areas.

The average annual hydrological balances obtained with the LU maps show that in 23 years of LU changes, the modifications observed in the hydrological process of the basin are slight. In fact, the two decades of LU changes have resulted in a slight decrease in average annual runoff and therefore in water supply and a slight increase in infiltration and groundwater flow. It can be argued that over 23 years, changes in LU may not be significant enough to create a distinct change in the basin's water balance. This result could also indicate that the variation in water balance parameters on Cavally at Toulepelu would be dependent on other factors other than the change in LU. Setyorini et al. (2017) also studied both LU change (1987–2006) and climate impacts in the Brantas basin. The results of their study showed that significant changes in LU and climate, in particular temperature, have impacted surface runoff and other hydrological characteristics of the catchment.

Joint impact of climate and plant cover on the functioning of the Cavally catchment at Toulepleu

The transition from P1 to P2 results in a reduction in runoff and infiltration and a clear increase in actual evapotranspiration. This indicates that the parameters of the balance, and therefore the hydrological response of the basin, depend on the combined evolution of external factors (climate) and internal factors (land use). In fact, this basin has experienced a 35% drop in rainfall and a 25% decline in dense forest. These results indicate that the combined effect of climate change and land use is considerably modifying the hydrological balance of the Cavally at Toulepleu. Seguis et al. (2004) indicated that analysis of the joint effect of climate change and human activity resulted in an increase in runoff of the order of 30–70% in a small Sahelian basin. Studies carried out in various catchment areas concur with these results. The study by Yan et al. (2019) reports that the combined effect of LU and climate increased surface runoff, water yield, and evapotranspiration. Similarly, Berihun et al. (2019) reported that the effects of climate and land use variation significantly increased surface runoff and evapotranspiration.

In this study, the hydrological components caused by the combined effects of climate and land use variations were higher than the individual variations. This finding was confirmed by Seguis et al. (2004), Marhaento et al. (2018), and Kuma et al. (2021), who found that variations in climate and land use individually cause water balance but that greater changes are likely if the factors are combined.

In conclusion, this research assessed the effects of climate variability and land use dynamics on the hydrological balance of the Cavally River catchment at Toulepleu, West Africa. The study utilised the SWAT hydrological model to simulate runoff in the catchment based on climate data from 1980 to 2016 and land use maps from 1992 to2015.

The analysis of hydrograph simulation and statistical indices demonstrated that the SWAT model performed well in reproducing the flow of the Cavally River at Toulepleu, with good correlations between observed and simulated monthly mean flows. However, discrepancies in extreme flows were attributed to factors such as uneven spatial distribution of rain gauges, anthropogenic activities, or limitations in the SWAT model for tropical regions.

The research found that the Cavally catchment experienced a drop in rainfall, leading to a reduction in runoff and infiltration during the dry climate period from 1997 to 2016 compared to the wet period from 1980 to 1996. Climate variability had a significant impact on the hydrological response of the basin, affecting the water balance parameters.

Land use changes between 1992 and 2015 showed a regression of dense forest and an increase in open forest, degraded forest, crops, fallow land, and dwellings. However, these changes had only a slight effect on the average annual runoff and infiltration of the basin over the 23-year period.

The combined effect of climate variability and land use changes resulted in a substantial modification of the hydrological balance of the Cavally catchment. The transition from wet to dry periods, coupled with changes in land use, led to a 68.96% reduction in runoff and a 49.52% reduction in infiltration, with a slight increase in evapotranspiration.

Importantly, the study revealed that the combined impact of climate variation and LU dynamics had a more significant effect on surface runoff and other hydrological parameters than the individual impacts of either climate or human activities. This highlights the need for policymakers to prioritise sustainable land use practices and integrated water resource management strategies to mitigate the combined effects of climate variability and anthropogenic activities. Implementing such measures would contribute to the long-term resilience of the ecosystem and ensure water availability for local communities in the Cavally River catchment at Toulepleu, West Africa.

The authors acknowledge the African-German Network of Excellence in Science (AGNES) for a mobility grant to Ghana to deepen and complete this research. Furthermore, we would acknowledge the National Directions of Meteorology (Côte d'Ivoire and Guinea) for making the data available to them. They also thank the anonymous reviewers for their critical comments and advice during the preparation of the manuscript.

This work was carried out in collaboration among all authors. All authors read and approved the final manuscript.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Alexandersson
H.
1986
A homogeneity test applied to precipitation data
.
Journal of Climatology
6
(
6
),
661
675
.
https://doi.org/10.1002/joc. 3370060607.
Anouma
D. G. L.
2020
Etude Hydroclimatic and Morphostructural Conditions in the Sassandra Catchment Area: Influence on the Risk of Flooding and Drying Up of Watercourses
.
PhD
,
Université Nangui-Abrogoua
,
Abidjan (Côte d'Ivoire)
, p.
202
.
Arab
A. M.
&
Gocic
M.
2021a
Analysis of temporal and spatial variations of drought over Serbia by investigating the applicability of precipitation-based drought indices
.
Theoretical and Applied Climatology
.
https://doi.org/10.1007/s00704-023-04554-6.
Arab
A. M.
&
Gocic
M.
2021b
Analyzing the applicability of some precipitation concentration indices over Serbia
.
Theoretical and Applied Climatology
.
https://doi.org/10.1007/s00704-021-03743-5
.
Arab
A. M.
&
Mesgari
M. S.
2018
Analyzing the spatial variability of precipitation extremes along longitude and latitude, Northwest Iran, Kuwait
.
Journal of Science
45
(
1
),
121
127
.
Baker
B. H.
,
Aldridge
C. A.
&
Omer
A. R.
2016
Water: Availability and Use
.
Mississippi State University Extension
, Starkville, MS, USA, p.
3011
.
Berihun
M. L.
,
Tsunekawa
A.
,
Haregeweyn
N.
,
Meshesha
D. T.
,
Adgo
E.
,
Tsubo
M.
,
Masunaga
T.
,
Fenta
A. A.
,
Sultan
D.
,
Yibeltal
M.
&
Ebabu
K.
2019
Hydrological responses to land use/land cover change and climate variability in contrasting agro-ecological environments of the Upper Blue Nile Basin, Ethiopia
.
Science of the Total Environment
689
,
347
365
.
https://doi.10.1016/j.scitotenv.2019.06.338
.
Bronstert
A.
,
Crisologo
I.
,
Heistermann
M.
,
Ozturk
U.
,
Vogel
K.
&
Wendi
D.
2020
Flash-Floods: More Often, More Severe, More Damaging? An Analysis of Hydro-geo-environmental Conditions and Anthropogenic Impacts
. In:
Climate Change, Hazards and Adaptation Options (Leal Filho, W., Nagy, G., Borga, M., Chávez Muñoz, P., Magnuszewski, A., eds). Climate Change Management. Springer, Cham
, pp.
225
244
.
https://doi.org/10.1007/978-3-030-37425-9_12.
Brou
L. A.
2019
Modélisation de la dynamique hydrologique du fleuve Cavally sous l'influence de fortes pressions anthropiques dans la zone de Zouan – Hounien (Côte d'Ivoire)
.
Thèse de Doctorat
,
Université Jean Lorognon Guédé
, p.
209
.
Deguy
A. J.-P.
2021
Evaluation de la vulnérabilité à l’érosion hydrique des sols du bassin versant de la Lobo (Côte d'Ivoire) : Approche par modélisations
.
Thèse unique de Doctorat
,
Université Nangui-Abrogoua
,
Abidjan (Côte d'Ivoire)
, p.
238
.
FAO 2009 Les forêts et l'eau: étude thématique préparée dans la carde de l’évaluation des ressources forestière mondiale 2005. ISBN978-92-5206090-1. 108p.
Gocic
M.
&
Arab
A. M.
2023
Analysis of spatial variability and patterns of drought
. In:
book: Integrated Drought Management
, Vol.
2
.
CRC Press
, p.
12
,
doi:10.1201/9781003276548-3
.
Hamed
K. H.
&
Rao
R.
1998
A modified Mann-Kendall trend test for autocorrelated data
.
Journal of Hydrology
204
,
182
196
.
Hassan
E.-G.
,
Morteza
L.
,
Masoud
S. D. S.
&
Afshin
A.
2021
Improving the performance of rainfall-runoff models using the gene expression programming approach
.
Journal of Water and Climate Change
00
(
0
),
1
22
.
doi:10.2166/wcc.2021.064
.
Jaiswal
R. K.
,
Lohani
A. K.
&
Tiwari
H. L.
2015
Statistical analysis for change detection and trend assessment in climatological parameters
.
Environmental Processes
2
(
4
),
729
749
.
https://doi.org/10.1007/s40710-015-0105-3.
Karimidastenaei
Z.
,
Klöve
B.
,
Sadegh
M.
&
Haghighi
A. T.
2021
Polar ice as an unconventional water resource: Opportunities and challenges
.
Water
13
(
22
),
3220
.
Khoi
D. N.
&
Suetsugi
T.
2014
The responses of hydrological processes and sediment yield to land-use and climate change in the Be River Catchment, Vietnam
.
Hydrological Processes
28
(
3
),
640
652
.
http://doi.org/10.1002/hyp.9620.
Khoi
D. N.
,
Nguyen
V. T.
,
Sam
T. T.
,
Mai
N. T. H.
,
Vuong
N. D.
&
Cuong
H. V.
2021
Assessment of climate change impact on water availability in the upper Dong Nai River Basin, Vietnam
.
Journal of Water and Climate Change
00
(
0
),
1
14
.
doi:10.2166/wcc.2021.255
.
Klassou
K. S.
&
Komi
K.
2021
Analysis of extreme rainfall in Oti River Basin (West Africa)
.
Journal of Water and Climate Change
.
doi:10.2166/wcc.2021.154
.
Koua
T. J.
,
Anoh
K. A.
,
Soro
T. D.
,
Kouame
K. J.
&
Jourda
J. P.
2019
Evaluation of Agricultural practices scenarios for reducing erosion in Buyo Lake Catchment (Sassandra; Côte d'Ivoire) by Use of GIS
.
Journal of Geoscience and Environment Protection
7
,
154
171
.
Kouadio
Z. A.
,
Kouakou
K. E.
,
Konan
W. A. B.
,
Goula
B. T. A.
&
Savane
I.
2015
Modelling the hydrological behaviour of the Boubo catchment in the humid tropical environment of Côte d'Ivoire using the distributed hydrological model CEQUEAU
.
Afrique Science
11
(
3
),
82
100
.
Kuma
H. G.
,
Feyessa
F. F.
&
Demissie
A. T.
2021
Hydrologic responses to climate and land-use/land-cover changes in the Bilate catchment, Southern Ethiopia
.
Journal of Water and Climate Change
00
(
0
),
1
.
doi:10.2166/wcc.2021.281
.
Marhaento
H.
,
Boodij
M. J.
&
Hoekstra
A. Y.
2018
Hydrological changes to future land use change and climate change in a tropical catchment
.
Hydrological Sciences Journal
63
(
9
),
1368
1385
.
https://doi.org/10.1080/02626667.2018.1511054.
Moriasi
D. N
,
Arnold
J. G.
,
Van Liew
M. W.
,
Bingner
R. L.
,
Harmel
R. D.
&
Veith
T. L.
2007
Model evaluation guidelines for systematic quantification of accuracy in watershed simulations
.
Transactions of the Asabe
50
,
885
900
.
Journal of Hydrology 204: 182–196.
N'Dri
W. K. C.
,
Séverin
P.
,
Jourda
J. P.
&
Kouamé
K. J.
2019a
Application of SWAT to estimate water balance in the Aghien Lagoon Basin, South-East of Côte d'Ivoire
.
International Journal of Science and Research
8
,
10
.
N'guessan
A. G.
,
Yoboue
K. B.
&
N'daoule
R.
2015
La variabilité pluviométrique et ses impacts sur l'alimentation en eau des milieux ruraux du département de Dimbokro (centre-est de la cote)
.
European Scientific Journal
11
,
1857
7881
.
Renard
B.
2006
Detecting and Taking Into Account the Possible Impacts of Climate Change on Hydrological Extremes in France
.
PhD Thesis
,
Institut National Polytechnique de Grenoble (France)
, p.
269
.
Roukia
B. T.
,
Mustapha
N.
,
Mohamed
C.
,
Damien
R.
&
Mohamed
S.
2019
Evaluation Des Performances Du Modele AgroHydrologique SWAT à Reproduire Le Fonctionnement Hydrologique Du Bassin Versant Nakhla (Rif occidental, Maroc)
.
Scientific Journal February 2019 edition
15
,
1857
7881
.
Salameh
M. T. B.
,
Alraggad
M.
&
Harahsheh
S. T.
2021
The water crisis and the conflict in the Middle East
.
Sustainable Water Resources Management
7
,
69
.
https://doi.org/10.1007/s40899-021-00549-1
.
Seguis
L.
,
Cappelaere
B.
,
Milesi
G.
,
Peugeot
C.
,
Massuel
S.
&
Favreau
G.
2004
Simulated impacts of climate change and land-clearing on runoff from a small Sahelian catchment
.
Hydrological Processes
18
(
17
),
3401
3413
.
ff10.1002/hyp.1503ff. ffird-00187211
.
Sahar
A. A.
,
Hassan
M. A.
&
Jasim
A. A.
2021
Estimating the Volume of Sediments and Assessing the Water Balance of the Badra Basin, Eastern Iraq, Using Swat Model and Remote Sensing Data
.
Iraqi Geological Journal
.
Soro
G. E.
2011
Statistical Modelling of Extreme Rainfall in Côte d'Ivoire
.
Unique PhD
,
Université Abobo-Adjamé
,
Côte d'Ivoire
, p.
172
.
Toyi
M. S.
,
Barima
S.
,
Mama
A.
,
Andre
M.
,
Bastin
J. F.
,
De Cannière
C.
&
Bogaert
J.
2013
Tree plantation will not compensate natural woody vegetation cover loss in the Atlantic Department of Southern Benin
.
Tropicultura
31
(
1
),
62
70
.
Tzanakakis
V. A.
,
Paranychianakis
N. V.
&
Angelakis
A. N.
2020
Water supply and water scarcity
.
Water
12
,
2347
.
https://doi.org/10.3390/w12092347
.
Vezzoli
R.
,
Pecora
S.
,
Zenoni
E.
&
Tonelli
F.
2012
Data analysis to detect inhomogeneity, change points, trends in observations: An application to Po river discharge extremes (SSRN Scholarly Paper ID 2195345)
.
Social Science Research Network
.
https://doi.org/10.2139/ssrn.2195345
.
Yan
R.
,
Cai
Y.
,
Li
C.
,
Wang
X.
&
Liu
Q.
2019
Hydrological responses to climate and land use changes in a watershed of the Loess Plateau, China
.
Sustainability
11
(
5
),
1
9
.
https://doi.org/10.3390/su11051443
.
Yao
A. B.
,
2015
Evaluation of the Water Potential of the Lobo Catchment Area with A View to Rational Management (West-Central Ivory Coast)
.
PhD Thesis
,
Université Nangui Abrogoua
,
Côte d'Ivoire
, p.
330
.
Yao
B. A. F.
&
Soro
G. E.
2021a
Detection of hydrologic trends and variability in transboundary Cavally Basin (West Africa)
.
American Journal of Water Resources
9
(
2
),
92
102
.
doi: 10.12691/ajwr-9-2-6
.
Yao
B. A. F.
&
Soro
G. E.
2021b
Recent rainfall trend and abrupt changes over cavally basin (West Africa)
.
International Journal of Environment and Climate Change
11
(
12
),
52
66
.
doi:10.9734/IJECC/2021/v11i1230556
.
Yonaba
R.
,
Biaou
A. C.
,
Koïta
M.
,
Tazen
F.
,
Mounirou
L. A.
,
Zouré
C. O.
,
Queloz
P.
,
Karambiri
H.
&
Yacouba
H.
2021
A dynamic land use/land cover input helps in picturing the Sahelian paradox: Assessing variability and attribution of changes in surface runoff in a Sahelian watershed
.
Science of the Total Environment
757
,
143792
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).