278
I. Thiaw et al.
1 Introduction
The 1990s and 2000s are marked by a slight resumption of rainfall in the Sahelian
zone (Lebel and Ali 2009; Panthou et al. 2014; Fall 2014; Marega 2016) after a
long period of drought that not only weakened the functioning of hydrosystems but
also affected human activities such as agriculture, fisheries and livestock (Dacosta
1989; Servat et al. 1998). This return of rainfall intensity has been observed across
several West African watersheds, including the Senegal (Cissé et al. 2014), Gambia
(Faye 2018), and Diarha (Thiaw 2017) rivers, putting their hydrological functioning
at risk. It has been suggested that the return to humid conditions observed in some
West African watersheds over the past few decades could be caused by anthropogenic
contributions to climate change (Paeth and Hence 2004). However, the monitoring
and assessment of these hydrosystems are hampered by a low density of monitoring networks associated with a lack of financial resources for national hydrological
services (Bodian et al. 2015), so the attribution of hydrological impacts of climate
change to rainfall or any other cause or causes would not be based on robust data.
Spot data for temperature, precipitation and potential evapotranspiration available at several monitoring stations are key inputs to existing water balance models,
compensating for a lack of flow data (Bop et al. 2014); in our study, a rainflow model
transforms time series describing the climatic conditions of a watershed into a series
of flows. This indirect assessment is an essential tool to understand the future of rainwater and its dynamics in a watershed, the preferred scale for hydrological studies.
Given the complexity of this process, often several mathematical models are used
in parallel: global conceptual models, semi-distributed and distributed models. The
accuracy of these models to simulate flows is improved by calibrating parameters
(Gutpa et al. 2003; Vrugt et al. 2006)—modifying them until the output of the model
corresponds to an observed data set (Liu 2009)—which has enabled automatic optimization techniques for hydroclimatic models in recent decades (Duan et al. 1992;
Vrugt et al. 2003).
Hydrological impacts of climate change include an increase in evaporation
coupled with rainfall variability, which can significantly affect surface runoff,
frequency and intensity of floods and droughts, soil moisture and water available for
irrigation and hydroelectric production (Setegn et al. 2011). In West Africa, the work
of Roudier et al. (2014) shows that previous studies focused mainly on individual
watersheds and were based on Global- (GCM) or Regional Climate Model (RCM)
projections. More recently, the CORDEX (Regional Climate Reduction Experience
Coordinated, Giorgi et al. 2009; Famien et al. 2018) initiative indicates an overall
decrease in rainfall of 20%-30% in Africa, resulting in a decrease in river flow ranging
from 40 to 60% (Biao 2017).
In Senegal, studies of the Gambia and Senegal river basins (Ardoin-Bardin et al.
2009; Bodian et al. 2013; Roudier et al. 2014; Mbaye et al. 2015; Bodian et al. 2018;
Stanzel et al. 2018) suggest a decrease in the flow of its rivers by the end of the
twenty-first century, correlating with a decrease in precipitation. Future impacts of
climate change have not been projected for the Diarha sub-basin. The few studies of
I. Thiaw et al.
1 Introduction
The 1990s and 2000s are marked by a slight resumption of rainfall in the Sahelian
zone (Lebel and Ali 2009; Panthou et al. 2014; Fall 2014; Marega 2016) after a
long period of drought that not only weakened the functioning of hydrosystems but
also affected human activities such as agriculture, fisheries and livestock (Dacosta
1989; Servat et al. 1998). This return of rainfall intensity has been observed across
several West African watersheds, including the Senegal (Cissé et al. 2014), Gambia
(Faye 2018), and Diarha (Thiaw 2017) rivers, putting their hydrological functioning
at risk. It has been suggested that the return to humid conditions observed in some
West African watersheds over the past few decades could be caused by anthropogenic
contributions to climate change (Paeth and Hence 2004). However, the monitoring
and assessment of these hydrosystems are hampered by a low density of monitoring networks associated with a lack of financial resources for national hydrological
services (Bodian et al. 2015), so the attribution of hydrological impacts of climate
change to rainfall or any other cause or causes would not be based on robust data.
Spot data for temperature, precipitation and potential evapotranspiration available at several monitoring stations are key inputs to existing water balance models,
compensating for a lack of flow data (Bop et al. 2014); in our study, a rainflow model
transforms time series describing the climatic conditions of a watershed into a series
of flows. This indirect assessment is an essential tool to understand the future of rainwater and its dynamics in a watershed, the preferred scale for hydrological studies.
Given the complexity of this process, often several mathematical models are used
in parallel: global conceptual models, semi-distributed and distributed models. The
accuracy of these models to simulate flows is improved by calibrating parameters
(Gutpa et al. 2003; Vrugt et al. 2006)—modifying them until the output of the model
corresponds to an observed data set (Liu 2009)—which has enabled automatic optimization techniques for hydroclimatic models in recent decades (Duan et al. 1992;
Vrugt et al. 2003).
Hydrological impacts of climate change include an increase in evaporation
coupled with rainfall variability, which can significantly affect surface runoff,
frequency and intensity of floods and droughts, soil moisture and water available for
irrigation and hydroelectric production (Setegn et al. 2011). In West Africa, the work
of Roudier et al. (2014) shows that previous studies focused mainly on individual
watersheds and were based on Global- (GCM) or Regional Climate Model (RCM)
projections. More recently, the CORDEX (Regional Climate Reduction Experience
Coordinated, Giorgi et al. 2009; Famien et al. 2018) initiative indicates an overall
decrease in rainfall of 20%-30% in Africa, resulting in a decrease in river flow ranging
from 40 to 60% (Biao 2017).
In Senegal, studies of the Gambia and Senegal river basins (Ardoin-Bardin et al.
2009; Bodian et al. 2013; Roudier et al. 2014; Mbaye et al. 2015; Bodian et al. 2018;
Stanzel et al. 2018) suggest a decrease in the flow of its rivers by the end of the
twenty-first century, correlating with a decrease in precipitation. Future impacts of
climate change have not been projected for the Diarha sub-basin. The few studies of