Regression Modeling of Base Flow Index (BFI) Using Watershed Characteristics for Bua River, Malawi

Abstract:

Many watersheds in Africa including Malawi are characterized by shortages of data such as river discharge data which is a global problem. In the absence of reliable data to determine groundwater discharge in rivers, this study presented a first comprehensive modeling technique to estimate base flow index (BFI) in Bua River watershed where discharge data records occurred at irregular intervals. Stochastic flow duration curve (FDC) analysis based on the principle of order statistics was presented in this study to evaluate various flow indices such as Q70₁₀ (which is the discharge equaled or not exceeded 70% of the time with a 10-year return period). This study used hydrological and meteorological data (34 years); and spatial data from the Malawi Government. The ‘smoothed minima’ base flow separation technique was used to quantify the BFI. Prior to model development, the Spearman’s correlation test was used to determine the predictors amongst 20 variables for model development. Multiple linear regression was used to develop equations for estimating annual BFI. The results showed that the Bua River has a moderately high base flow component of approximately 68% to 81%. The findings showed that the significant explanatory variables to estimate BFI (Model 1) included annual precipitation (AP), cropland cover (CLC) and Q70₁₀, while AP, BFI and grassland cover (GLC) were used as explanatory variables to estimate Q70₁₀ (Model 2). The adjusted R² ranged from 0.61 to 0.95 for the regression models, showing that they were suitable for estimating the flow indices. These methods and findings could be used to develop a database of base flow indices for watersheds in Malawi and other overseas countries. In addition, the findings from this research would contribute to the achievement of the Sustainable Development Goals; 2, 6 and 15. In conclusion, these methods and findings would also have important implications for future studies in hydrology and water resources management in watersheds.