Impact of Streamflow Gap Filling on QSWAT+ and GR4J Rainfall–Runoff Modeling in the Mouhoun River Basin, Burkina Faso

Rainfall-discharge modeling is essential for water resources management in West African river basins, which are often characterized by high hydroclimatic variability and incomplete hydrometric data. This study evaluates the effect of flow series reconstruction on the performance of the QSWAT+ and GR4J models in the Mouhoun River basin at Banzon, in order to identify the approach best suited to a context of limited data availability. The methodology consists of two distinct components: The first component involves quality control and gap analysis, gap filling using missForest and Mice-PMM. In this study, missForest was selected as the preferred gap-filling method because it consistently outperformed MICE-PMM across the KGE, NSE, R2, RMSE, MAE, and PBIAS values. The second component includes the configuration and model run, sensitivity analysis, calibration, and validation of the models. Sensitivity, calibration, and validation performance is evaluated primarily using the Kling–Gupta Efficiency (KGE). Using the missForest-reconstructed streamflow dataset, GR4J outperformed QSWAT+ during both calibration (1981–1995) and validation (1996–2014). GR4J achieved KGE and NSE values of 0.73 and 0.48 during calibration, and 0.62 and 0.44 values during validation, compared with values of 0.66 and 0.30, and 0.54 and 0.26, respectively, for QSWAT+. This superior performance was also reflected in lower RMSE values for GR4J (7.16 and 8.50 during calibration and validation, respectively) than for QSWAT+ (8.25 and 9.76, respectively). Despite its spatialized representation of processes, QSWAT+ appears to be more sensitive to uncertainties related to data and parameterization. These results show that combining reconstruction of discharge data with a parsimonious model constitutes a relevant approach for hydrological modeling in the context of limited data.

Authors

Institutions

Publication Details

Journal
Limnological Review
Published
2026-11-19
DOI
https://doi.org/10.3390/limnolrev26040066
Primary Topic
Hydrology and Watershed Management Studies
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Impact of Streamflow Gap Filling on QSWAT+ and GR4J Rainfall–Runoff Modeling in the Mouhoun River Basin, Burkina Faso

Bambara Telado Luc, Moumouni Djibo, Ali Doumounia, Wenceslas Somda et al.
Limnological Review
Hydrology and Watershed Management Studies
article

Impact of Streamflow Gap Filling on QSWAT+ and GR4J Rainfall–Runoff Modeling in the Mouhoun River Basin, Burkina Faso

Bambara Telado Luc, Moumouni Djibo, Ali Doumounia, Wenceslas Somda, Dioviel Dominique Somda, François Zougmore
article en

Abstract

Rainfall-discharge modeling is essential for water resources management in West African river basins, which are often characterized by high hydroclimatic variability and incomplete hydrometric data. This study evaluates the effect of flow series reconstruction on the performance of the QSWAT+ and GR4J models in the Mouhoun River basin at Banzon, in order to identify the approach best suited to a context of limited data availability. The methodology consists of two distinct components: The first component involves quality control and gap analysis, gap filling using missForest and Mice-PMM. In this study, missForest was selected as the preferred gap-filling method because it consistently outperformed MICE-PMM across the KGE, NSE, R2, RMSE, MAE, and PBIAS values. The second component includes the configuration and model run, sensitivity analysis, calibration, and validation of the models. Sensitivity, calibration, and validation performance is evaluated primarily using the Kling–Gupta Efficiency (KGE). Using the missForest-reconstructed streamflow dataset, GR4J outperformed QSWAT+ during both calibration (1981–1995) and validation (1996–2014). GR4J achieved KGE and NSE values of 0.73 and 0.48 during calibration, and 0.62 and 0.44 values during validation, compared with values of 0.66 and 0.30, and 0.54 and 0.26, respectively, for QSWAT+. This superior performance was also reflected in lower RMSE values for GR4J (7.16 and 8.50 during calibration and validation, respectively) than for QSWAT+ (8.25 and 9.76, respectively). Despite its spatialized representation of processes, QSWAT+ appears to be more sensitive to uncertainties related to data and parameterization. These results show that combining reconstruction of discharge data with a parsimonious model constitutes a relevant approach for hydrological modeling in the context of limited data.

Limnological ReviewVol. 26(4)
Université Joseph Ki-Zerbo (BF), Université Daniel Ouezzin Coulibaly (BF)
Hydrology and Watershed Management Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon