Skip to Content
MDPIMDPI
Search
WaterWater
PDF
  • Article
  • Open Access

29 January 2026

Applicability of the Lumped GR4J Model for Modeling the Hydrology of the Inland Valleys of the Sudanian Zones of Benin

Order Reprints
,
,
,
and
1
Earth and Life Institute (ELI), Environmental Sciences, Université Catholique de Louvain, Croix du Sud 2, L7.05.02, BE-1348 Louvain-la-Neuve, Belgium
2
Laboratory of Hydraulics and Environmental Modeling (HydroModE-Lab), Faculty of Agronomy, University of Parakou, Parakou, Benin
3
Institut des Sciences et Technologies pour l’Innovation en Afrique (ISTI-Africa), Parakou, Benin
*
Author to whom correspondence should be addressed.
Water2026, 18(3), 340;https://doi.org/10.3390/w18030340
This article belongs to the Special Issue Advances in Ecohydrology in Arid Inland River Basins, 2nd Edition
Version Notes
Review Reports

Abstract

Achieving sustainable agricultural intensification in inland valleys while limiting the adverse environmental impacts and uncertainties related to water availability requires an analysis of the long-term hydrological behavior of the catchment. Such a task is particularly challenging in West Africa and Benin due to the limited availability of climate and hydrological data. This study evaluates the applicability of the lumped GR4J model for simulating streamflow in three inland valleys of the Sudanian zone of Benin (Lower-Sowé, Bahounkpo and Nalohou). Additionally, we test the reliability of satellite-based rainfall data (GPM-IMERG, CHIRPS or GSMAP) in modeling hydrological dynamics in these small catchments. The results demonstrate that the GR4J model is effective in simulating daily discharge in the three inland valleys (KGE > 0.5 during both calibration and validation periods), with particularly interesting performance in mean-flow conditions. The modeling using GPM-IMERG and GSMAP rainfall data shows mitigated results with acceptable performance at Nalohou and less accurate results at Bahounkpo and Lower-Sowé. CHIRPS emerged as the most consistent among the evaluated products, providing a sound basis for reconstructing general trends and seasonal variations in historical streamflow time series. The approach of combining historical CHIRPS data and the GR4J model provides insights and can support decision-making related to water resource management in terms of resource capacity and volume in the study area. Except for Nalohou (KGE = 0.19 with GPM-IMERG data), we observe limitations in predicting high flows with satellite-based climatic data at Bahounkpo (KGE = 0.02 with GPM-IR) and Lower-Sowé (KGE = −0.01 with CHIRPS), where the near-zero KGE scores indicate marginal improvement over a mean-flow benchmark. Future work should explore how hybrid or flexible modeling approaches can improve the accuracy of runoff simulations in inland valleys, particularly for extreme (low- and high-) flow conditions. Additionally, the analysis of the trends of indicators of hydrological alteration (IHA) must be deepened in these important ecosystems, especially under climate and land-use change scenarios.

1. Introduction

In Sub-Saharan Africa, inland valleys are an important asset of local agriculture due to favorable water availability and soil conditions in their bottom valleys [1,2]. Water resource management is a challenge in these areas, where agricultural productivity and ecological stability are heavily dependent on the adequate management of water [3,4]. The pressures of climate variability, land use change, increasing population, and economic activities further constrain water resources in inland valley ecosystems, making sustainable management more challenging, yet imperative [5,6].
Streamflow dynamics are particularly important in water management schemes in the inland valleys due to their impact on flood and drought risks, agricultural productivity and water supply reliability [3,7]. Understanding streamflow patterns in inland valley ecosystems is essential for appropriate water management in Sub-Saharan Africa. Hydrological modeling emerges as an indispensable tool in this context, providing insights into the behavior of various water flow components and helping in decisions related to their sustainable use [8]. Consequently, numerous studies have employed a range of modeling approaches to simulate streamflow in inland valley hydrological systems [9,10,11,12,13,14,15,16,17,18].
Previous streamflow modeling studies of inland valleys are mainly targeted to two objectives: (1) to assess the effectiveness of models to reproduce the system’s functioning, and (2) to predict the impact of global changes on water availability and water-related ecosystem services. Concerning the first objective, the analysis of the rainfall–runoff modeling studies in the inland valleys of Benin reveals a mismatch between the model types applied in the literature and the practical needs of operational stakeholders. The models used are mainly distributed models (Table 1), which are often complex and require numerous input parameters at fine temporal and spatial resolutions [19]. Most of the input parameters required by this type of model are not easily available at sites of interest in Benin [20], creating a gap in their operational use as a decision-support tool. Selecting an appropriate modeling approach that captures the essential dynamics of the hydrological system, while considering data availability and computational resources, has always been a major topic of discussion in the hydrologic community [21,22,23,24,25]. In regions with limited data availability, there is a practical need for models that can produce satisfactory simulation results and offer user-friendly platforms for stakeholders [9,26]. In the context of hydrological modeling of inland valleys of Benin, the reliability of lumped hydrological models, which can represent interesting alternatives to deal with limited hydrological information, has been underexplored. The challenges related to the complexity of water flow dynamics in these ecosystems and operational needs for their sustainable management underscore the importance of evaluating the predictive capability of this type of model in inland valleys.
Regarding the objective to predict the impact of global changes on water availability and water-related ecosystem services, inland valleys have received limited attention in comparison to large fluvial catchments in Benin due to the high costs associated with hydrological experimentation and the scarce resources allocated by public services to this sector. The water resource monitoring system in the country is characterized by a lack of continuous monitoring infrastructure for inland valleys. Examining data published in peer-reviewed journals and focusing on runoff at the outlets of basins with an area of less than 200 km2, thirteen (13) inland valleys have been subject to documented monitoring in the country. Sites studied in the literature, include the inland valleys of Upper Niaou, Upper Aguima, Lower Aguima [9,10,27], Bamé [28], Nalohou, Ara, Kolokondé [11,15,29,30,31,32,33,34], Bokpérou [35], Tossahou, Kounga, Kpandouga [36,37], Koupendri [38] and Ouriyori [16,39]. There remains an important need for research to understand long-term climate dynamics, how it impacts each component of water flows and what the implications are for flood and drought prediction processes in inland valleys.
In the Sudanian zone of Benin, research is expected to adapt engineering applications in water resources management. The applicability of emergent satellite-based rainfall products in modeling studies appears in this context as an important aspect that needs attention in the field of inland valley hydrology. Similar observations have been made in the literature for numerous tropical catchments [40]. GPM-IMERG, CHIRPS or GSMAP, which emerge as interesting alternative rainfall data in West Africa [41,42,43,44,45], have not been tested to determine their effectiveness in modeling hydrological dynamics in inland valleys. As a result, there is currently a lack of comprehensive understanding regarding the applicability of these products for supporting hydrological modeling and, finally, decision-making, particularly in the Sudanian zones of Benin.
As part of this study, we explored the possibilities of using the GR4J model [46] for the reconstruction of long-term discharge at the outlet of three inland valleys with the aim of better understanding the occurrence of extreme hydrological events in these catchments. We seek to bridge the gap between theoretical hydrology and practical application by adapting a lumped hydrological model to the needs of water management purposes in the inland valleys of the Sudanian zones of Benin. The outcomes are expected to inform better water resource management practices, support decision-making processes in agricultural management and aid in the development of strategies to mitigate the impacts of climate variability and change on local and regional scales.
Table 1. Overview of models applied to simulate discharge in inland valleys of Sudanian zones of Benin: model type, spatial and temporal resolution.

2. Methodology

2.1. Study Sites

The experimental sites used in this study are the inland valleys of Lower-Sowé, Bahounkpo and Nalohou. According to the Köppen classification [49], the climate of the 3 study sites is tropical savanna. Vegetation cover is composed of a mosaic of woody savannah, forest and grasslands with scattered trees. The inland valley of Lower-Sowé (15.22 km2) belongs to the Sowé watershed (49 km2) and is located West of the city of Glazoué in the center of Benin. The Bahounkpo inland valley (5.38 km2) is located approximately 10 km from N’dali in the North East of Benin. The Nalohou inland valley (0.16 km2) is located near the town of Djougou in the North West of Benin.
The inland valleys of Lower-Sowé and Bahounkpo are monitored as part of the technical and scientific support project for the valorization and sustainable integrated management of water resources in the inland valleys in Benin, the BAFONBE project. The Nalohou inland valley, which has been the focus of previous studies [15,32], is monitored in the framework of the AMMA-CATCH program [33]. Figure 1 provides an overview of Beninese inland valley sites that have been documented on the subject of outlet runoff and displays site locations in this study.
Figure 1. Documented hydrological monitoring of inland valleys in Benin (a) and study sites: Bahounkpo (b), Itchedjiro—Lower-Sowé (c), and Nalohou—Ara (d). The climatic zone categorization in this study is based on information from the World Bank Climate Knowledge Portal for Benin (https://climateknowledgeportal.worldbank.org/country/benin (accessed on 31 July 2025)). Additionally, the classification of tropical savanna climate types follows [50] for the Atacorian type and [51] for the sub-equatorial type.

2.2. GR4J Model Overview and Input Data Used

GR4J is a four-parameter, daily lumped rainfall–runoff model. It was chosen here due to its performance across numerous studies as well as its parsimony and concise model structure [24,46,52,53,54,55]. The GR4J model adopts two nonlinear reservoirs for rainfall–runoff modeling, of which the first reservoir represents the production process, and the second reservoir is employed to conceptualize the confluent process. Figure 2 provides a schematic diagram of the GR4J catchment conceptualization and model process.
Figure 2. Structural diagram of the GR4J model [46]. Terms mean the following: P: precipitation input (mm/day), E: evapotranspiration (mm/day), S: water in the production (soil) store (mm/day), X1: maximum capacity of the production store (mm), Pn: net rainfall reaching soil (mm/day), En: net evapotranspiration demand applied to the store (mm/day), Ps: infiltrated part of Pn into the store (mm/day), Es: actual evaporation from the store (mm/day), Perc: percolation from production to routing (mm/day), Pr: effective rainfall to routing (mm/day), UH1, UH2: unit hydrographs distributing Pr in time (1/day), X4: time base of UH1 (day); R: water in the routing store (mm), X3: maximum capacity of the routing store (mm), F(X2) denotes the groundwater exchange flux computed from the parameter X2 (groundwater exchange coefficient, mm/day), Qr: routed flow from the routing store (mm/day), Qd: direct flow from UH2 branch (mm/day), and Q: total discharge at outlet (mm/day).
Input data used include rain and reference evapotranspiration for the study period of each site, corresponding, respectively, to the periods 2021–2022 for Lower-Sowé, 2021–2023 for Bahounkpo and 2015–2019 for Nalohou. Data for the Lower-Sowé and Bahounkpo inland valleys originate from experimental monitoring carried out as part of the BAFONBE project. Data for Nalohou originates from the AMMA CATCH database.
GPM-IMERG, CHIRPS and GSMAP data were chosen for evaluation based on previous analyses demonstrating their local accuracy when compared to in situ rainfall used as a reference. Reference evapotranspiration was estimated using the Hargreaves method with ERA5 temperature data, which also provides reliable estimates in the study region [56].

2.3. Data Analysis

The objectives of this study were twofold. Firstly, we evaluated the performance of the GR4J model in simulating rainfall–runoff processes in the study sites. Secondly, the research interest has been extended to include the reconstruction of hydroclimatic conditions to provide a historical context for trend and extreme hydrological signature analysis. Runoff simulations were performed using the GR4J parameterization from the HydroGR package for hydrological modeling, which is inspired by the Irstea R language package airGR [52].

2.3.1. Performance of the GR4J Model

The first specific objective of this study involves calibration and validation of the model with different sets of rainfall data (in situ measured, GPM-IMERG, CHIRPS and GSMAP). We evaluated the performance of the GR4J model in simulating rainfall–runoff processes in the study sites using in situ measured data (Approach 0). Additionally, the study evaluates the applicability of satellite-derived precipitation datasets for streamflow modeling. Two methodological approaches were used to assess the effectiveness of satellite-derived precipitation data in streamflow modeling: (1) calibration and validation of the GR4J hydrological model using each satellite precipitation dataset independently, and (2) calibration of the GR4J hydrological model using in situ precipitation data followed by validation using each of the satellite precipitation datasets. The employment of these two distinct methodological approaches is designed to provide a robust and comprehensive assessment of the strengths and limitations of using satellite precipitation data in hydrological modeling under different scenarios of data availability and model parameterization. Calibrating and validating the model with each satellite precipitation product (Approach 1) allows for a direct assessment of the standalone capability of these products to drive hydrological simulations, particularly relevant in data-scarce regions where in situ hydro-climatic data for calibration may be limited or unavailable. This approach helps to identify the inherent biases and uncertainties associated with using satellite precipitation as the primary input for hydrological models. It allows for exploring the dependence of the results on reference data. Conversely, calibrating the model with in situ precipitation data and then validating with each satellite precipitation product (Approach 2) assesses the ability of these satellite products to accurately simulate streamflow when the model parameters have been optimized using potentially more reliable ground observations [57]. This approach explores whether satellite data can effectively capture the precipitation patterns needed for reliable streamflow simulation when the model’s internal processes are reasonably well represented. The split-sample validation technique was used to split each discharge time series into two equal periods to ensure robust calibration and reliable evaluation of the hydrological model. Table 2 presents data used in the calibration and validation step according to each approach.
Table 2. Data used in the calibration and validation step according to the explored approach.
The best parameter set obtained from each inland valley (according to the explored approaches) was applied to analyze the capacity of the model to simulate streamflow outside the calibration period, thereby demonstrating its validity. The model performance was evaluated on different goodness-of-fit indicators, including visual and comparative time series metrics-based analysis. We also incorporated the analysis of hydrological signature indicators, specifically considering high flows (Q10), low flows (Q90), and the runoff coefficient. The Q10 and Q90 have been computed only for discharge values above 0.1 mm to avoid biasing the detection of high and low flows, since intermittent rivers frequently have zero flow values. A similar approach has been applied by [58]. The corresponding subsets of the hydrograph (all values ≥ Q10 and all values ≤ Q90) were retained for analysis. The metrics used include the coefficient of correlation of Pearson, the Root Mean Square Error (RMSE), the Nash–Sutcliffe Efficiency (NSE) and the Kling–Gupta Efficiency (KGE). The Pearson correlation coefficient indicates the linear relationship between observed and simulated discharge values. The RMSE quantifies the average magnitude of errors, emphasizing larger errors and providing insight into the model’s accuracy. The NSE assesses model predictive power by comparing the observed variance against the variance of the residuals. The KGE assesses both the correlation, bias and variability between observed and simulated data, offering a comprehensive assessment of model performance. KGE was used as the main evaluation criterion due to its balanced consideration of various aspects of model performance, making it more robust for hydrological applications. Other metrics were used to gain a deeper understanding of specific discrepancies between observed and simulated discharge, helping to identify potential areas for model improvement. For results interpretation purposes, we adopted a threshold of model performance in the range −0.41 < KGE ≤  1 as reasonable, following [59], −0.41 being the KGE value corresponding to a mean-flow benchmark. We considered that a KGE value greater than or equal to 0.75 indicates a very good similarity between the 2-time series, and values between 0.75 and 0.5 represent satisfactory similarity as commonly used in hydrology [60]. Sensitivity analysis of the parameters of the model was performed using the variance-based sensitivity analysis of Sobol [61], whose usefulness was demonstrated in numerous studies [62,63].

2.3.2. Long-Term Climatic and Hydrological Analysis

For the second specific objective, the model that demonstrated the highest performance for each inland valley was applied to simulate historical runoff time series. Subsequently, climatic and hydrological trend analyses were conducted, along with an assessment of hydrological extremes.
Climatic and Hydrological Trend Analysis
Climatic and hydrological signatures were analyzed using the Mann–Kendall and Pettit test on multi-annual daily time series of rain, reference evapotranspiration, aridity index, total runoff, high flow and runoff coefficient. These characteristics are particularly interesting either because they can be used for decision-making in terms of water management or because they provide practical information about the state of alteration of hydrological regimes. In addition, to explore the interannual relationship between rainfall and discharge, daily rainfall and discharge data were aggregated to obtain annual totals. The year-to-year variability was assessed by calculating the percentage change in annual rainfall and discharge using the following formula:
C h a n g e   % = X i − X i − 1 ∗ 100 X i − 1
where Xi and Xi−1 represent the annual total values of rainfall or discharge for years i and i − 1, respectively. The computed changes were then used to assess the sensitivity of discharge to variations in rainfall on an annual scale.
Hydrological Extreme Analysis
Extreme value analysis was performed on reconstructed historical daily discharge data by fitting the Generalized Extreme Value (GEV) distribution to the high-flow events and by generating return value plots. High-flow event identification was based on the above-threshold method. This method offers the advantage of using a rational selection of events. It results in a high number of observations and a more homogeneous sample [64]. The 99-th percentile of the daily discharge time series was used as the threshold to identify extreme high-flow events in this study. A 95% confidence interval was applied to the estimates, providing a robust statistical framework for predicting the likelihood and magnitude of extreme flow events over various return periods.

3. Results

3.1. Model Performance Assessment

3.1.1. Model Performance with Measured Rainfall

The model performed well in simulating discharge in the three inland valleys. Visual comparisons clearly illustrate a good fit between observed and simulated discharge (Figure 3, Figure 4 and Figure 5). The performance of the model in simulating discharge is supported by the statistical metrics, which indicate an overall satisfactory accuracy (KGE > 0.57). We observed consistent accuracy in the validation phase, suggesting that the model is robust. The agreement between observed and simulated discharge highlights the model’s effectiveness in capturing the study area’s hydrological processes.
Figure 3. Results of discharge simulation with in situ rainfall data at Bahounkpo.
Figure 4. Results of discharge simulation with in situ rainfall data at Lower-Sowé.
Figure 5. Results of discharge simulation with in situ rainfall data at Nalohou.
Despite overall satisfying simulation results, the comparative analysis of observed and simulated hydrological signatures shows some weaknesses. While the model performs acceptably in high flow and runoff coefficient simulations, there are consistent challenges in accurately capturing low and peak flow conditions across all sites (Figure 6, Figure 7 and Figure 8). Specific simulation biases are observed at the beginning of the rainy season and in reproducing flood peaks during high-flow events.
Figure 6. Observed and simulated hydrological signatures with in situ rainfall data in the inland valley of Bahounkpo: (a) high flows, (b) low flows, and (c) runoff coefficient.
Figure 7. Observed and simulated hydrological signatures with in situ rainfall data in the inland valley of Lower-Sowé: (a) high flows, (b) low flows, (c) runoff coefficient.
Figure 8. Observed and simulated hydrological signatures with in situ rainfall data in the inland valley of Nalohou: (a) high flows, (b) low flows, (c) runoff coefficient.
The sensitivity analysis performed during simulations with in situ rainfall data indicates that parameter X2 (Groundwater exchange coefficient) is consistently critical across all catchments, while parameters X3 and X4 exhibit varying degrees of sensitivity depending on the catchment (Figure 9). These parameters provide insights into the hydrological behavior of the basin, highlighting that surface–groundwater exchange and catchment response time are crucial areas where refinement may be needed to enhance model performance.
Figure 9. Sensitivity analysis of the parameters of the model in the study sites (a) Bahounkpo, (b) Lower-Sowé, and (c) Nalohou.

3.1.2. Model Performance with Satellite-Based Rainfall Data

Table 3 summarizes the results of the model performance using satellite-based rainfall data at calibration and validation across the three catchments. The modeling using GPM-IMERG rainfall data shows mixed results, with acceptable performance at Nalohou and Bahounkpo (KGE > 0.44) and less accurate results at Lower-Sowé. GSMAP yields poor performance at Bahounkpo and Lower-Sowé but provides acceptable results at Nalohou. During the validation period, the decreased performance in terms of KGE is particularly marked for GSMAP at Bahounkpo and Lower-Sowé. Among the evaluated products, CHIRPS emerges as the most consistent, providing a good capture of general trends and seasonal variations in discharge prediction (KGE > 0.47). It is interesting to note the acceptable estimation of hydrological signatures like runoff coefficient and median flows observed with CHIRPS data, making the modeling approach with this satellite rainfall data useful for operational water management and planning in data-scarce inland valleys of the region. Except for Nalohou (KGE = 0.19 with GPM-IMERG data), we observe limitations in predicting high flows when using the satellite-based rainfall approach for both calibration and validation.
Table 3. Results of daily streamflow simulations at three sites using satellite-based rainfall data at calibration and validation.
Calibration with in situ rainfall data performs better than with satellite data. During validation of the hybrid approach (rainfall data at calibration, combined with satellite rainfall for validation), model performance varied depending on the site and the satellite product used (Table 4). Validation metrics indicated less satisfactory performance at the Nalohou inland valley across all simulation runs, particularly concerning the reproduction of temporal flow dynamics. Conversely, at the Bahounkpo and Lower-Sowé inland valleys, an inverse pattern was observed, with interesting model performance achieved when in situ rainfall data were used for calibration and satellite-based rainfall data for validation. Regardless of the calibration data source, the CHIRPS product consistently demonstrated the highest performance in simulating temporal flow dynamics within the studied inland valley systems. Furthermore, CHIRPS provided the most interesting estimations of high flow events at the Lower-Sowé station, while GPM-IMERG exhibited the best performance for this specific variable at the Bahounkpo and Nalohou sites. While it is above the 0.41 threshold, the near-zero KGE scores recorded for the performing high-flow simulations at Bahounkpo and Lower-Sowé (based on satellite-derived rainfall) indicate marginal improvement over a mean-flow benchmark, necessitating cautious interpretation. Despite limitations related to the prediction of high flows, the approach of combining historical CHIRPS data and the GR4J model provides insights and can support decision-making related to water resource management in the study area, particularly in estimating mean water balance, resource capacity and volume.
Table 4. Results of daily streamflow simulations at three sites using in situ rainfall data at calibration and satellite-based rainfall data at validation.

3.2. Long-Term Discharge Occurrence Analysis

Following the results obtained in high flow simulation with satellite rainfall data (approaches 1 and 2), long-term climatic and hydrological analysis was performed with the following:
  • CHIRPS data historical simulation (1981–2023) following approach 1 at Lower-Sowé;
  • GPM-IMERG data historical simulation (2000–2023) following approach 2 at Bahounkpo;
  • GPM-IMERG data historical simulation (2000–2023) following approach 1 at Nalohou.
The combination of GR4J and each rainfall satellite data for high-flow simulation on the studied sites enables an analysis of parameters important for defining sustainable water management schemes, such as discharge occurrence thresholds and trends. The trend analysis was made on annual aggregated data. Lower-Sowé and Nalohou are experiencing a scenario of increased annual rainfall and reference evapotranspiration for the analyzed period. Analysis of the long-term variability of hydrological signatures revealed a distinct increase in runoff in all three sites, probably as a consequence of more water availability, resulting from the balance between rainfall and evapotranspiration (Figure 10, Figure 11 and Figure 12). Seasonal and inter-annual variations in discharge are linked to rainfall patterns. We observe a slight increase in high discharge events, which follow high rainfall events, suggesting a risk of flooding and the need for enhanced, accurate management strategies. Pettit’s tests show a change point in reference evapotranspiration trends at Lower-Sowé, but no break is depicted in rainfall and discharge trends.
Figure 10. Trend analysis of historical climatic (rainfall, reference evapotranspiration, and aridity index) and hydrological parameters (total runoff, high flow, and runoff coefficients) in the Bahounkpo inland valley. Blue markers represent annual estimated values, and red dashed lines indicate linear regression trends.
Figure 11. Trend analysis of historical climatic (rainfall, reference evapotranspiration, aridity index) and hydrological parameters (total runoff, high flow, runoff coefficients) in the Lower-Sowé inland valley. Blue markers represent annual estimated values, and red dashed lines indicate linear regression trends.
Figure 12. Trend analysis of historical climatic (rainfall, reference evapotranspiration, and aridity index) and hydrological parameters (total runoff, high flow, and runoff coefficients) in the Nalohou inland valley. Blue markers represent annual estimated values, and red dashed lines indicate linear regression trends.

3.3. Analysis of Extreme Discharge

Figure 13a, Figure 14a and Figure 15a quantify the probability and uncertainty of extreme discharge events, while Figure 13b, Figure 14b and Figure 15b explores the meteorological drivers behind these events. The fitted probability distribution closely follows the observed values, indicating a reliable representation of extreme flow behavior, while the widening confidence interval at higher return periods reflects increased uncertainty in estimating rare, high-magnitude discharge events. This suggests the need for cautious extrapolation when predicting extreme hydrological conditions for the studied inland valley, as variability increases with less frequent occurrences. A positive correlation between changes in precipitation and corresponding changes in streamflow is observed. The scatter points in Figure 13b, Figure 14b and Figure 15b suggest that although rainfall is a dominant driver of discharge variations, additional factors (such as land cover, soil conditions and antecedent moisture) likely modulate the hydrological response. The presence of outliers at high discharge changes illustrates the influence of nonlinear or threshold-driven processes on the runoff events. The findings emphasize the need for integrated hydrological assessments that account for both statistical and process-based uncertainties in extreme discharge predictions.
Figure 13. Discharge analysis for water management scheme definition: (a) high-flow return period using the GEV model, and (b) cross impacts of rainfall and runoff dynamics in the site of Bahounkpo.
Figure 14. Discharge analysis for water management scheme definition: (a) high-flow return period, and (b) cross impacts of rainfall and runoff dynamics in the site of Lower-Sowé.
Figure 15. Discharge analysis for water management scheme definition: (a) high-flow return period, and (b) cross impacts of rainfall and runoff dynamics in the site of Nalohou.

4. Discussion

4.1. Model Performance

Although the model GR4J has been used for modeling studies in large watersheds across West Africa [13,20,65,66,67,68,69], its applicability to the context of inland valleys, characterized by small catchment areas (<200 km2) with highly complex and varied hydrological dynamics, has been scarcely studied. The results illustrate the effectiveness of the model in simulating daily discharge in inland valleys in Benin. This effectiveness should be compared to the performance obtained in modeling studies of inland valleys with other models [7,10,11,15,16,36]. For instance, for the Nalohou site and compared to the study of [15], our results show that the lumped GR4J model is effective, which underscores the observations of [9]. In such conditions, a model like GR4J can be used as a decision-support tool for water management in the inland valley. It is, however, important to note that the lumped structure of GR4J, with a single production store and fixed flow partitioning, seems to limit its ability to accurately simulate high and low flows, particularly in catchments with intermittent regimes such as inland valleys, where groundwater fluctuations influence runoff generation.
Sensitivity analysis underscores the role of surface–groundwater exchange processes in the hydrological behavior of the inland valleys. Groundwater exchange coefficient appears as a key driver of runoff response in inland valleys. It highlights the importance of collecting groundwater data to improve the modeling of the hydrologic signature of inland valleys. This result indicates areas where further refinement or additional measures might be needed to improve model performance. For example, collecting data on groundwater loss or soil moisture retention could enhance the model’s accuracy and reliability. Soil moisture assimilation seems to be an interesting way to improve the performance of the model. Future work should explore how hybrid or flexible modeling approaches can improve the accuracy of runoff simulations in inland valleys, particularly for extreme (low- and high-) flow conditions.
In the context of the unavailability of accurate in situ rainfall data, satellite-based rainfall products such as CHIRPS can give a good alternative for the estimation of discharge trends and seasonal variations in the three sites. These results are in accordance with those obtained by [41], who found that CHIRPS has the best overall performance over West Africa and northern Central Africa for hydrological modeling at a monthly step. Ref. [70] also demonstrated that the CHIRPS dataset enables accurate streamflow simulation in the Kaboua and Aval Sani catchments, which are located within the same geographical zone as the present study area. GSMAP and GPM-IMERG show mitigated results on the study sites. Given their accuracy against CHIRPS when compared to overall in situ rainfall time series, the low performance of GSMAP and GPM-IMERG in modeling discharge dynamics suggests that overall temporal dynamics do not fully encapsulate all hydrological characteristics. The good performance in modeling the temporal dynamics of discharge when using CHIRPS might be due to its relatively consistent quality across different seasonal conditions. GPM-IMERG demonstrates an interesting ability to simulate high discharge at Nalohou, consistent with its capability to capture extreme rainfall events (>50 mm) at the study sites. GSMAP dataset, while accurate in replicating measured rainfall, seems to be more susceptible to errors such as light precipitation events, which play an important role in hydrological processes. This highlights the importance of considering not only the accuracy of rainfall estimates but also how well each dataset aligns with the hydrological response of the basin.
Since the calibration process of hydrological models can itself introduce biases affecting runoff simulations, it is likely that the model calibration may favor rainfall datasets that better reproduce realistic runoff dynamics, rather than those that solely provide accurate rainfall estimates. In this regard, Ref. [70] emphasizes the importance of calibrating the model individually for each rainfall product, as biases introduced during initial calibration can influence simulation outcomes. The primary objective of model calibration is to adjust parameters to align simulated runoff with observed data, which may inherently advantage datasets that more effectively capture the basin’s hydrological response. Approaches to correcting biases in the presence of satellite-based rainfall data represent an option that can be tested to improve streamflow simulations in inland valleys for operational purposes, mainly for the analysis of extreme flows.
When comparing the two approaches using satellite-based rainfall data explored in this study, one notable advantage of the method involving calibration with in situ rainfall data and validation with satellite rainfall data is that in situ measurements allow optimal parameter adjustment during calibration. However, validating the model using satellite data complicates the assessment of model robustness, as it becomes challenging to discern whether poor performance results from biases inherent in satellite rainfall data or from limited model generalizability. Furthermore, this approach restricts the operational use of the model in contexts relying solely on satellite data. Conversely, calibrating and validating entirely with satellite rainfall data directly tests the applicability of satellite products for hydrological modeling, especially in regions like Benin, where rainfall stations are scarce or nonexistent. The primary difficulty of this approach lies in propagating inherent satellite rainfall biases into the runoff simulations during calibration. Without in situ reference input data, the resulting model exhibits increased versatility and requires recalibration at the same site for different datasets. Additionally, distinguishing whether observed discrepancies between simulations and observations arise from model parameterization issues or biases in the satellite input data becomes challenging. Operationally, the choice between these two approaches will depend on data availability and the specific application context.

4.2. Use of Hydrological Modeling as an Operational Decision Support Tool for Inland Valley Water Management

To control water-related risks, it is important to better understand the hazard and its impact [41]. Lumped models are among the tools for operational water management applications (including training and education) because they propose simplified catchment-scale representations of the transformation of precipitation into river discharge. However, the potential of the GR4J model for operational water management has not been evaluated often enough for inland valleys. The approach used here takes advantage of advances in alternative climate data [71] and efficient user-friendly models to provide an analytic frame for decision-making in inland valley systems. The most important practical advantage of this approach is that it gives a complete flow time series from which various flow characteristics can be extracted. Additionally, various scenarios of water use development can be easily evaluated [72]. In operational terms, for example, the reconstruction of historical flow time series offers a sound basis for defining design parameters such as extreme flow indices or indicators related to favorable periods for agricultural activities. A similar approach has been used to make a hydrological diagnosis of the Bankandi valley in Burkina Faso [3]. One direct application of such an approach in the context of inland valleys water management is related to the trends of high-flow occurrence. Historically, water management schemes in the inland valleys of Benin are based on guidelines derived from data collected in the 70s–90s by CIRAD in cooperation with CIEH and national agricultural research institutions. This research led to the conception of a tool, the “Diagnostic Rapide de Pré-Aménagement (DIARPA)”, or Rapid Appraisal for Inland Valley Development, which defines the type of management scheme for a given inland valley mainly based on five parameters: decennial discharge, longitudinal slope, soil permeability, depth of the impermeable layer and depth of the interflow water table of the study site. Decennial discharge, which is a component of extreme flows in the catchment, is among the most important and difficult-to-estimate parameters used in DIARPA. It can be easily and accurately estimated using historical time series of discharge. Figure 13 presents, as an example, the extreme flow return period analysis based on the simulated historical time series of Bahounkpo, which provides the extreme flow values that must be taken into consideration for water management infrastructure design. Another example of application is related to understanding trends in periods of high-water availability. These periods could be targeted for planting water-intensive crops like rice, while periods of low mean flow might indicate the need for drought-resistant varieties or supplemental irrigation measures. Changes in these periods can also be seen if they exist, and the indices related to favorable periods of agricultural activities in the wet season can be adjusted accordingly. In this context, the consistency of the trend information provided by the model can be reinforced by social surveys related to extreme events.
An additional operational advantage of the approach used here is the possibility to infer groundwater dynamics in the modeled catchments from the GR4J routing-store relative storage (R/X3). Comparing R/X3 with observed piezometric heads at Bahounkpo shows coherent seasonal rises and recessions (Figure 16). Spearman’s rank correlations (rs) are high for most wells. Normalized cross-correlation functions (ρxγ(τ)) peak at positive lags (τ ≈ +2 to +60 days) with coefficients typically varying between 0.82 and 0.93, indicating that, once lagged, R/X3 reproduces most of the temporal variability in groundwater levels with potential predictive implications at Bahounkpo. Consequently, for this site, R/X3 can be used as an operational proxy for groundwater state to support rule-based management, with site-specific lead times informed by the observed lag range.
Figure 16. Time-series comparison and lagged-correlation analysis of groundwater levels and GR4J routing-store relative storage in the Bahounkpo inland valley (2021–2023).

4.3. Limitations of the Research

Despite the interesting results observed in this study, it is important to note that there are some limitations. The limitations are related to the inherent uncertainties in environmental science data and the model, which can affect the results. For certain sites (Bahounkpo and Lower-Sowé), for high flow simulation based on satellite-derived rainfall, the near-zero KGE values show limited improvement compared to the mean-flow benchmark and should therefore be interpreted with caution. Furthermore, extremes estimated using the GEV distribution on simulated data may misrepresent uncertainty in extreme flow analysis, reinforcing the need for careful interpretation.
Another limitation of the approach used here is its direct inapplicability to ungauged sites, as at least one year of calibration/validation is required to ensure the model’s operationality. It would also have been interesting to conduct the present analyses on a larger number of study sites and with tests extended to a wider range of satellite products. Additionally, this study did not explore the potential of employing bias correction techniques to further enhance the performance of satellite rainfall-driven streamflow simulations. The study also did not analyze how rainfall features and inland valley responses contribute to observed differences in the hydrological performance of satellite rainfall data. Investigating such approaches and questions could be beneficial for operational applications, especially for improving the analysis of extreme flow events. These observations, which can be integrated into future studies, do not undermine the results obtained here.

5. Conclusions

This study demonstrates the effective application of the GR4J hydrological model to understand and predict discharge occurrences in the inland valleys of the Sudanian zones of Benin. The model has shown robust performance in simulating discharge events with high seasonal and inter-annual variations. Based on KGE values, CHIRPS stands out as the most reliable and consistent product for water management, while GPM-IMERG and GSMAP show variability and site-specific efficiency, requiring further improvements to be considered dependable. The integration of CHIRPS rainfall data has enabled reasonable runoff simulation, enhancing our ability to understand streamflow trends and seasonal variations based on the reconstruction of historical hydrodynamic conditions. This proactive approach is a precious tool to support informed decision-making by providing a reliable estimation of streamflow thresholds and then building resilience against actual and possible future changes in sustainable water management in the region. The insights gained from this research contribute to a better understanding of the hydrological dynamics in the inland valleys of Sudanian zones of Benin and provide a foundation for developing resilient and sustainable water management practices. The effective implementation and continuous improvement of the model will ensure its long-term utility in sustainable water resource management. Additionally, training and capacity-building efforts are essential to ensure that local authorities and stakeholders effectively adopt and utilize this approach.

Author Contributions

A.M.T.’s contributions include the following: conceptualization, data collection and curation, formal analysis, investigation, methodology, validation, visualization, writing—original draft preparation, writing—review and editing; Q.F.T.’s and P.G.T.’s contributions include the following: conceptualization, formal analysis, methodology, validation, visualization, writing—review and editing, supervision, project administration; P.B.I.A.’s and M.V.’s contributions include the following: conceptualization, formal analysis, methodology, validation, visualization, writing—review and editing, funding acquisition, project administration, resources, supervision. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Wallonie Bruxelles International (WBI) as part of the BAFONBE Project.

Data Availability Statement

The original data presented in this study are openly available in Bafonbe at https://sites.google.com/view/projetbafonbe (accessed on 1 July 2025).

Acknowledgments

This study is the outcome of reflections conducted within the BAFONBE Project, fully funded by Wallonie Bruxelles International (WBI), for which we express our gratitude. We thank the technical teams involved in setting up and collecting data for the BAFONBE project, especially PETIT Sebastien, CHABI N’GOBI Julien, SOMBORO Adam, GOUNOU SAKA Kandine (in memoriam), AFFEDJOU Hyacinthe and AKPONIKPE Thomas. We extend our gratitude to the team of the AMMA CATCH program, particularly the scientific heads of the datasets used in this study. Finally, we thank all the reviewers for their insightful observations and helpful advice, which significantly improved this work.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Dossou-Yovo, E.R.; Baggie, I.; Djagba, J.F.; Zwart, S.J. Diversity of Inland Valleys and Opportunities for Agricultural Development in Sierra Leone. PLoS ONE 2017, 12, e0180059. [Google Scholar] [CrossRef] [PubMed]
  2. Djagba, J.F.; Sintondji, L.O.; Kouyaté, A.M.; Baggie, I.; Agbahungba, G.; Hamadoun, A.; Zwart, S.J. Predictors Determining the Potential of Inland Valleys for Rice Production Development in West Africa. Appl. Geogr. 2018, 96, 86–97. [Google Scholar] [CrossRef]
  3. Yacouba, Y.; Aymar, B.Y.; Amadou, K.; Louis, F.J.; Georges, S.; Thomas, Y.B.; Mouhamed, I.; Bruno, L. Failure of Inland Valleys Development: A Hydrological Diagnosis of the Bankandi Valley in Burkina Faso. Model. Earth Syst. Environ. 2019, 5, 1733–1741. [Google Scholar] [CrossRef]
  4. Djagba, J.F.; Zwart, S.J.; Houssou, C.S.; Tenté, B.H.A.; Kiepe, P. Ecological Sustainability and Environmental Risks of Agricultural Intensification in Inland Valleys in Benin. Environ. Dev. Sustain. 2019, 21, 1869–1890. [Google Scholar] [CrossRef]
  5. Yira, Y.; Diekkrüger, B.; Steup, G.; Yaovi Bossa, A. Impact of Climate Change on Hydrological Conditions in a Tropical West African Catchment Using an Ensemble of Climate Simulations. Hydrol. Earth Syst. Sci. 2017, 21, 2143–2161. [Google Scholar] [CrossRef]
  6. Gabiri, G.; Diekkrüger, B.; Näschen, K.; Leemhuis, C.; van der Linden, R.; Mwanjalolo Majaliwa, J.-G.; Obando, J.A. Impact of Climate and Land Use/Land Cover Change on the Water Resources of a Tropical Inland Valley Catchment in Uganda, East Africa. Climate 2020, 8, 83. [Google Scholar] [CrossRef]
  7. Bossa, A.Y.; Hounkpè, J.; Yira, Y.; Serpantié, G.; Lidon, B.; Fusillier, J.L.; Sintondji, L.O.; Tondoh, J.E.; Diekkrüger, B. Managing New Risks of and Opportunities for the Agricultural Development of West-African Floodplains: Hydroclimatic Conditions and Implications for Rice Production. Climate 2020, 8, 11. [Google Scholar] [CrossRef]
  8. Sezen, C.; Šraj, M. Improving the Simulations of the Hydrological Model in the Karst Catchment by Integrating the Conceptual Model with Machine Learning Models. Sci. Total Environ. 2024, 926, 171684. [Google Scholar] [CrossRef]
  9. Bormann, H.; Giertz, S.; Diekkrüger, B. Hydrological Catchment Models: Process Representation, Data Availability and Applicability for Water Management—Case Study for Benin. In Proceedings of the Symposium S6 Held During the Seventh IAHS Scientific Assembly, Foz do Iguaçu, Brazil, 3–9 April 2005; pp. 86–93. [Google Scholar]
  10. Giertz, S.; Diekkrüger, B.; Steup, G. Physically-Based Modelling of Hydrological Processes in a Tropical Headwater Catchment (West Africa)—Process Representation and Multi-Criteria Validation. Hydrol. Earth Syst. Sci. 2006, 10, 829–847. [Google Scholar] [CrossRef]
  11. Varado, N.; Braud, I.; Galle, S.; Le Lay, M.; Séguis, L.; Kamagate, B.; Depraetere, C. Multi-Criteria Assessment of the Representative Elementary Watershed Approach on the Donga Catchment (Benin) Using a Downward Approach of Model Complexity. Hydrol. Earth Syst. Sci. 2006, 10, 427–442. [Google Scholar] [CrossRef]
  12. Love, D.; Uhlenbrook, S.; Corzo-Perez, G.; Twomlow, S.; van der Zaag, P. Rainfall-Interception-Evaporation-Runoff Relationships in a Semi-Arid Catchment, Northern Limpopo Basin, Zimbabwe. Hydrol. Sci. J. 2010, 55, 687–703. [Google Scholar] [CrossRef]
  13. Cornelissen, T.; Diekkrüger, B.; Giertz, S. A Comparison of Hydrological Models for Assessing the Impact of Land Use and Climate Change on Discharge in a Tropical Catchment. J. Hydrol. 2013, 498, 221–236. [Google Scholar] [CrossRef]
  14. Yira, Y.; Diekkrüger, B.; Steup, G.; Bossa, A.Y. Modeling Land Use Change Impacts on Water Resources in a Tropical West African Catchment (Dano, Burkina Faso). J. Hydrol. 2016, 537, 187–199. [Google Scholar] [CrossRef]
  15. Hector, B.; Cohard, J.-M.; Séguis, L.; Galle, S.; Peugeot, C. Hydrological Functioning of Western African Inland Valleys Explored with a Critical Zone Model. Hydrol. Earth Syst. Sci. 2018, 22, 5867–5888. [Google Scholar] [CrossRef]
  16. Togbévi, Q.F.; Bossa, A.Y.; Yira, Y.; Preko, K.; Sintondji, L.O.; van der Ploeg, M. A Multi-Model Approach for Analysing Water Balance and Water-Related Ecosystem Services in the Ouriyori Catchment (Benin). Hydrol. Sci. J. 2020, 65, 2453–2465. [Google Scholar] [CrossRef]
  17. Idrissou, M.; Diekkrüger, B.; Tischbein, B.; Ibrahim, B.; Yira, Y.; Steup, G.; Poméon, T. Testing the Robustness of a Physically-Based Hydrological Model in Two Data Limited Inland Valley Catchments in Dano, Burkina Faso. Hydrology 2020, 7, 43. [Google Scholar] [CrossRef]
  18. Idrissou, M.; Diekkrüger, B.; Tischbein, B.; de Hipt, F.O.; Näschen, K.; Poméon, T.; Yira, Y.; Ibrahim, B. Modeling the Impact of Climate and Land Use/Land Cover Change on Water Availability in an Inland Valley Catchment in Burkina Faso. Hydrology 2022, 9, 12. [Google Scholar] [CrossRef]
  19. Jaiswal, R.K.; Ali, S.; Bharti, B. Comparative Evaluation of Conceptual and Physical Rainfall–Runoff Models. Appl. Water Sci. 2020, 10, 48. [Google Scholar] [CrossRef]
  20. Le Lay, M. Modélisation Hydrologique Dans un Contexte de Variabilité Hydro-Climatique: Une Approche Comparative Pour L’étude du Cycle Hydrologique à Méso-Échelle au Bénin. Ph.D. Thesis, Grenoble INPG, Grenoble, France, 2006. [Google Scholar]
  21. Jakeman, A.J.; Hornberger, G.M. How Much Complexity Is Warranted in a Rainfall-Runoff Model? Water Resour. Res. 1993, 29, 2637–2649. [Google Scholar] [CrossRef]
  22. Paudel, M.; Nelson, E.J.; Downer, C.W.; Hotchkiss, R. Comparing the Capability of Distributed and Lumped Hydrologic Models for Analyzing the Effects of Land Use Change. J. Hydroinform. 2010, 13, 461–473. [Google Scholar] [CrossRef]
  23. Vansteenkiste, T.; Tavakoli, M.; Steenbergen, N.; Smedt, F.; Batelaan, O.; Pereira, F.; Willems, P. Intercomparison of Five Lumped and Distributed Models for Catchment Runoff and Extreme Flow Simulation. J. Hydrol. 2014, 511, 335–349. [Google Scholar] [CrossRef]
  24. Sinha, S.; Hammond, A.; Smith, H. A Comprehensive Intercomparison Study between a Lumped and a Fully Distributed Hydrological Model across a Set of 50 Catchments in the United Kingdom. Hydrol. Process. 2022, 36, e14544. [Google Scholar] [CrossRef]
  25. Bouadila, A.; Bouizrou, I.; Aqnouy, M.; En-nagre, K.; El Yousfi, Y.; Khafouri, A.; Hilal, I.; Abdelrahman, K.; Benaabidate, L.; Abu-Alam, T.; et al. Streamflow Simulation in Semiarid Data-Scarce Regions: A Comparative Study of Distributed and Lumped Models at Aguenza Watershed (Morocco). Water 2023, 15, 1602. [Google Scholar] [CrossRef]
  26. Banda, V.D.; Dzwairo, R.B.; Singh, S.K.; Kanyerere, T. Hydrological Modelling and Climate Adaptation under Changing Climate: A Review with a Focus in Sub-Saharan Africa. Water 2022, 14, 4031. [Google Scholar] [CrossRef]
  27. Giertz, S.; Diekkrüger, B. Analysis of the Hydrological Processes in a Small Headwater Catchment in Benin (West Africa). Phys. Chem. Earth Parts A/B/C 2003, 28, 1333–1341. [Google Scholar] [CrossRef]
  28. Schmitter, P.; Zwart, S.J.; Danvi, A.; Gbaguidi, F. Contributions of Lateral Flow and Groundwater to the Spatio-Temporal Variation of Irrigated Rice Yields and Water Productivity in a West-African Inland Valley. Agric. Water Manag. 2015, 152, 286–298. [Google Scholar] [CrossRef]
  29. Kamagaté, B.; Séguis, L.; Favreau, G.; Seidel, J.-L.; Descloitres, M.; Affaton, P. Hydrological processes and water balance of a tropical crystalline bedrock catchment in Benin (Donga, upper Ouémé River). C. R.-Geosci. 2007, 339, 418–429. [Google Scholar] [CrossRef]
  30. Guyot, A.; Cohard, J.-M.; Anquetin, S.; Galle, S.; Lloyd, C.R. Combined Analysis of Energy and Water Balances to Estimate Latent Heat Flux of a Sudanian Small Catchment. J. Hydrol. 2009, 375, 227–240. [Google Scholar] [CrossRef]
  31. Séguis, L.; Kamagaté, B.; Favreau, G.; Descloitres, M.; Seidel, J.-L.; Galle, S.; Peugeot, C.; Gosset, M.; Le Barbé, L.; Malinur, F.; et al. Origins of Streamflow in a Crystalline Basement Catchment in a Sub-Humid Sudanian Zone: The Donga Basin (Benin, West Africa). Inter-Annual Variability of Water Budget. J. Hydrol. 2011, 402, 1–13. [Google Scholar] [CrossRef]
  32. Hector, B.; Séguis, L.; Hinderer, J.; Cohard, J.-M.; Wubda, M.; Descloitres, M.; Benarrosh, N.; Boy, J.-P. Water Storage Changes as a Marker for Base Flow Generation Processes in a Tropical Humid Basement Catchment (Benin): Insights from Hybrid Gravimetry. Water Resour. Res. 2015, 51, 8331–8361. [Google Scholar] [CrossRef]
  33. Galle, S.; Grippa, M.; Peugeot, C.; Bouzou Moussa, I.; Cappelaere, B.; Demarty, J.; Mougin, E.; Panthou, G.; Adjomayi, P.; Agbossou, E.K.; et al. AMMA-CATCH, a Critical Zone Observatory in West Africa Monitoring a Region in Transition. Vadose Zone J. 2018, 17, 1–24. [Google Scholar] [CrossRef]
  34. Bodjrènou, R.; Sintondji, L.O.; Comandan, F. Hydrological Modeling with Physics-Based Models in the Oueme Basin: Issues and Perspectives for Simulation Optimization. J. Hydrol. Reg. Stud. 2023, 48, 101448. [Google Scholar] [CrossRef]
  35. Gaba, C.; Alamou, E.; Afouda, A.; Diekkrüger, B. Improvement and Comparative Assessment of a Hydrological Modelling Approach on 20 Catchments of Various Sizes under Different Climate Conditions. Hydrol. Sci. J. 2017, 62, 1499–1516. [Google Scholar] [CrossRef]
  36. Danvi, A.; Giertz, S.; Zwart, S.J.; Diekkrüger, B. Comparing Water Quantity and Quality in Three Inland Valley Watersheds with Different Levels of Agricultural Development in Central Benin. Agric. Water Manag. 2017, 192, 257–270. [Google Scholar] [CrossRef]
  37. Danvi, A.; Giertz, S.; Zwart, S.J.; Diekkrüger, B. Rice Intensification in a Changing Environment: Impact on Water Availability in Inland Valley Landscapes in Benin. Water 2018, 10, 74. [Google Scholar] [CrossRef]
  38. Azuka, C.V.; Igué, A.M.; Diekkrüger, B. Modelling the Hydrological Processes of Koupendri Catchment Northwest, Benin. Int. J. Hydrol. Sci. Technol. 2021, 12, 142–163. [Google Scholar] [CrossRef]
  39. Togbévi, Q.F.; Van Der Ploeg, M.; Tohoun, K.A.; Agodzo, S.K.; Preko, K. Assessing the Effects of Anthropogenic Land Use on Soil Infiltration Rate in a Tropical West African Watershed (Ouriyori, Benin). Appl. Environ. Soil Sci. 2022, 2022, 8565571. [Google Scholar] [CrossRef]
  40. Andrade, J.M.; Ribeiro Neto, A.; Nóbrega, R.L.B.; Rico-Ramirez, M.A.; Montenegro, S.M.G.L. Efficiency of Global Precipitation Datasets in Tropical and Subtropical Catchments Revealed by Large Sampling Hydrological Modelling. J. Hydrol. 2024, 633, 131016. [Google Scholar] [CrossRef]
  41. Kouakou, C.; Paturel, J.-E.; Satgé, F.; Tramblay, Y.; Defrance, D.; Rouché, N. Comparison of Gridded Precipitation Estimates for Regional Hydrological Modeling in West and Central Africa. J. Hydrol. Reg. Stud. 2023, 47, 101409. [Google Scholar] [CrossRef]
  42. Gosset, M.; Viarre, J.; Quantin, G.; Alcoba, M. Evaluation of Several Rainfall Products Used for Hydrological Applications over West Africa Using Two High-Resolution Gauge Networks. Q. J. R. Meteorol. Soc. 2013, 139, 923–940. [Google Scholar] [CrossRef]
  43. Dembélé, M.; Schaefli, B.; van de Giesen, N.; Mariéthoz, G. Suitability of 17 Gridded Rainfall and Temperature Datasets for Large-Scale Hydrological Modelling in West Africa. Hydrol. Earth Syst. Sci. 2020, 24, 5379–5406. [Google Scholar] [CrossRef]
  44. Macharia, D.; Fankhauser, K.; Selker, J.S.; Neff, J.C.; Thomas, E.A. Validation and Intercomparison of Satellite-Based Rainfall Products over Africa with TAHMO In Situ Rainfall Observations. J. Hydrometeorol. 2022, 23, 1131–1154. [Google Scholar] [CrossRef]
  45. Pellarin, T.; Zoppis, A.; Román-Cascón, C.; Kerr, Y.H.; Rodriguez-Fernandez, N.; Panthou, G.; Philippon, N.; Cohard, J.-M. From SMOS Soil Moisture to 3-Hour Precipitation Estimates at 0.1° Resolution in Africa. Remote Sens. 2022, 14, 746. [Google Scholar] [CrossRef]
  46. Perrin, C.; Michel, C.; Andréassian, V. Improvement of a Parsimonious Model for Streamflow Simulation. J. Hydrol. 2003, 279, 275–289. [Google Scholar] [CrossRef]
  47. Bormann, H.; Faß, T.; Giertz, S.; Junge, B.; Diekkrüger, B.; Reichert, B.; Skowronek, A. From Local Hydrological Process Analysis to Regional Hydrological Model Application in Benin: Concept, Results and Perspectives. Phys. Chem. Earth Parts A/B/C 2005, 30, 347–356. [Google Scholar] [CrossRef]
  48. Bormann, H.; Breuer, L.; Giertz, S.; Huisman, J.A.; Viney, N.R. Spatially Explicit versus Lumped Models in Catchment Hydrology—Experiences from Two Case Studies. In Proceedings of the Uncertainties in Environmental Modelling and Consequences for Policy Making; Baveye, P.C., Laba, M., Mysiak, J., Eds.; Springer: Dordrecht, The Netherlands, 2009; pp. 3–26. [Google Scholar]
  49. Peel, M.C.; Finlayson, B.L.; McMahon, T.A. Updated World Map of the Köppen-Geiger Climate Classification. Hydrol. Earth Syst. Sci. 2007, 11, 1633–1644. [Google Scholar] [CrossRef]
  50. Adeoti, K.; Djedatin, G.; Ewedje, E.; Beulé, T.; Santoni, S.; Rival, A.; Jaligot, E. Assessment of Genetic Diversity among Cultivated Pearl Millet (Pennisetum Glaucum, Poaceae) Accessions from Benin, West Africa. Afr. J. Biotechnol. 2017, 16, 782–790. [Google Scholar] [CrossRef]
  51. Ahoyo, C.C.; Houéhanou, T.D.; Yaoïtcha, A.S.; Prinz, K.; Glèlè Kakaï, R.; Sinsin, B.A.; Houinato, M.R.B. Traditional Medicinal Knowledge of Woody Species across Climatic Zones in Benin (West Africa). J. Ethnopharmacol. 2021, 265, 113417. [Google Scholar] [CrossRef]
  52. Coron, L.; Thirel, G.; Delaigue, O.; Perrin, C.; Andréassian, V. The Suite of Lumped GR Hydrological Models in an R Package. Environ. Model. Softw. 2017, 94, 166–171. [Google Scholar] [CrossRef]
  53. Zeng, L.; Xiong, L.; Liu, D.; Chen, J.; Kim, J.-S. Improving Parameter Transferability of GR4J Model under Changing Environments Considering Nonstationarity. Water 2019, 11, 2029. [Google Scholar] [CrossRef]
  54. Adeyeri, O.E.; Laux, P.; Arnault, J.; Lawin, A.E.; Kunstmann, H. Conceptual Hydrological Model Calibration Using Multi-Objective Optimization Techniques over the Transboundary Komadugu-Yobe Basin, Lake Chad Area, West Africa. J. Hydrol. Reg. Stud. 2020, 27, 100655. [Google Scholar] [CrossRef]
  55. Muñoz-Castro, E.; Mendoza, P.; Vásquez, N.; Vargas, X. Exploring Parameter (Dis)Agreement Due to Calibration Metric Selection in Conceptual Rainfall–Runoff Models. Hydrol. Sci. J. 2023, 68, 1754–1768. [Google Scholar] [CrossRef]
  56. Tidjani, A. Hydrological Functioning of Inland Valleys for Water Management in the Sudanian Zones of Benin. Ph.D Thesis, UCL—Université Catholique de Louvain, Ottignies-Louvain-la-Neuve, Belgium, 2025. [Google Scholar]
  57. Bitew, M.M.; Gebremichael, M.; Ghebremichael, L.T.; Bayissa, Y.A. Evaluation of High-Resolution Satellite Rainfall Products through Streamflow Simulation in a Hydrological Modeling of a Small Mountainous Watershed in Ethiopia. J. Hydrometeorol. 2012, 13, 338–350. [Google Scholar] [CrossRef]
  58. Cantoni, E.; Tramblay, Y.; Grimaldi, S.; Salamon, P.; Dakhlaoui, H.; Dezetter, A.; Thiemig, V. Hydrological Performance of the ERA5 Reanalysis for Flood Modeling in Tunisia with the LISFLOOD and GR4J Models. J. Hydrol. Reg. Stud. 2022, 42, 101169. [Google Scholar] [CrossRef]
  59. Knoben, W.J.M.; Freer, J.E.; Woods, R.A. Technical Note: Inherent Benchmark or Not? Comparing Nash–Sutcliffe and Kling–Gupta Efficiency Scores. Hydrol. Earth Syst. Sci. 2019, 23, 4323–4331. [Google Scholar] [CrossRef]
  60. Thiemig, V.; Rojas, R.; Zambrano-Bigiarini, M.; De Roo, A. Hydrological Evaluation of Satellite-Based Rainfall Estimates over the Volta and Baro-Akobo Basin. J. Hydrol. 2013, 499, 324–338. [Google Scholar] [CrossRef]
  61. Sobol′, I.M. Global Sensitivity Indices for Nonlinear Mathematical Models and Their Monte Carlo Estimates. Math. Comput. Simul. 2001, 55, 271–280. [Google Scholar] [CrossRef]
  62. Zhang, C.; Chu, J.; Fu, G. Sobol′’s Sensitivity Analysis for a Distributed Hydrological Model of Yichun River Basin, China. J. Hydrol. 2013, 480, 58–68. [Google Scholar] [CrossRef]
  63. Shin, M.-J.; Jung, Y. Using a Global Sensitivity Analysis to Estimate the Appropriate Length of Calibration Period in the Presence of High Hydrological Model Uncertainty. J. Hydrol. 2022, 607, 127546. [Google Scholar] [CrossRef]
  64. Li, Z.; Huang, G.; Fan, Y.; Xu, J.L. Hydrologic Risk Analysis for Nonstationary Streamflow Records under Uncertainty. J. Environ. Inform. 2016, 26, 41–51. [Google Scholar]
  65. Ruelland, D.; Larrat, V.; Guinot, V. A Comparison of Two Conceptual Models for the Simulation of Hydro-Climatic Variability over 50 Years in a Large Sudano-Sahelian Catchment. IAHS-AISH Publ. 2010, 340, 668–678. [Google Scholar]
  66. Sambou, S.; Boye, M.; Malang, B.A.; Malanda-Nimy, E.N.; Bodian, A.; Mussa, K.; Adama, M.; Fatogoma, B.; Raymond, M.; Alpha, B.; et al. Calage et Validation des Modèles Hydrologiques GR4J et GR2M sur le Bassin du Bafing en Amont de Bafing-Makana: Vers l’étude de L’impact du Climat sur les Ressources en eau de la Retenue de Manantali. 2011, pp. 1–6. Available online: https://www.journees-scientifiques.2ie-edu.org/js2011/sessions/pdf/sossou_s.pdf (accessed on 31 July 2025).
  67. Kodja, D.J.; Mahé, G.; Amoussou, E.; Boko, M.; Paturel, J. Assessment of the Performance of Rainfall-Runoff Model GR4J to Simulate Streamflow in Ouémé Watershed at Bonou’s Outlet (West Africa). Preprints 2018. [Google Scholar] [CrossRef]
  68. Kodja, D.J.; Akognongbé, A.J.S.; Amoussou, E.; Mahé, G.; Vissin, E.W.; Paturel, J.-E.; Houndénou, C. Calibration of the Hydrological Model GR4J from Potential Evapotranspiration Estimates by the Penman-Monteith and Oudin Methods in the Ouémé Watershed (West Africa). Proc. IAHS 2020, 383, 163–169. [Google Scholar] [CrossRef]
  69. Koubodana, H.D.; Atchonouglo, K.; Adounkpe, J.G.; Amoussou, E.; Kodja, D.J.; Koungbanane, D.; Afoudji, K.Y.; Lombo, Y.; Kpemoua, K.E. Surface Runoff Prediction and Comparison Using IHACRES and GR4J Lumped Models in the Mono Catchment, West Africa. Proc. IAHS 2021, 384, 63–68. [Google Scholar] [CrossRef]
  70. Poméon, T.; Jackisch, D.; Diekkrüger, B. Evaluating the Performance of Remotely Sensed and Reanalysed Precipitation Data over West Africa Using HBV Light. J. Hydrol. 2017, 547, 222–235. [Google Scholar] [CrossRef]
  71. Dube, T.; Seaton, D.; Shoko, C.; Mbow, C. Advancements in Earth Observation for Water Resources Monitoring and Management in Africa: A Comprehensive Review. J. Hydrol. 2023, 623, 129738. [Google Scholar] [CrossRef]
  72. Smakhtin, V.U. Low Flow Hydrology: A Review. J. Hydrol. 2001, 240, 147–186. [Google Scholar] [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Journal Statistics
Multiple requests from the same IP address are counted as one view.
Hybrid Agricultural Compensation Mechanism for Protection of Ecological Flow in Rivers in Water-Scarce Areas of Northwest China
Previous
Building Virtual Drainage Systems Based on Open Road Data and Assessing Urban Flooding Risks
Next