Regression Modeling of Baseflow and Baseflow Index for Nigeria


Abstract: Baseflow plays an important role in maintaining streamflow. Seventeen gauged watersheds and their characteristics were used to develop regression models for annual baseflow and baseflow index (BFI) estimation in Michigan. Baseflow was estimated from daily streamflow records using the two-parameter recursive digital filter method for baseflow separation of the Web-based Hydrograph Analysis Tool (WHAT) program. Three equations (two for annual baseflow and one for BFI estimation) were developed and validated. Results indicated that observed average annual baseflow ranged from 162 to 345

mm, and BFI varied from 0.45 to 0.80 during 1967–2011. The average BFI value during the study period was 0.71, suggesting that about 70% of long-term streamflow in the studied watersheds were likely supported by baseflow. The regression models estimated baseflow and BFI with relative errors (RE) varying from −29% to 48% and from −14% to 19%, respectively. In absence of reliable information to determine groundwater discharge in streams and rivers, these equations can be used to estimate BFI and annual baseflow in Nigeria.

Baseflow is a very important component of streamflow generated from groundwater inflow or discharge. Baseflow is generally derived from available streamflow records using hydrograph separation techniques such as graphical methods,  recession-curve methods, analytical methods, mass-balance methods, and digital baseflow filter methods. Many of these techniques have been automated with computer programming (e.g., PART, HYSEP, BFI, UKIH, BFLOW, and WHAT) to assist in baseflow separation.
Although these programs are widely used and accepted in hydrologic studies, they are mostly limited to estimating baseflow in gauged watersheds. In order words, they are not applicable to ungauged areas where records of streamflow do not exist. Previous studies have used regression analysis extensively to estimate baseflow at ungauged sites in various regions of the world. For example, Santhi et al. utilized regression analysis to relate relief, percentage of sand and effective rainfall to baseflow index (BFI) and baseflow volume for the conterminous United
States. Mazvimavi et al. also used multiple regressions to predict BFI from mean annual precipitation, watershed slope, and proportion of a basin with grasslands in Zimbabwe. Longobardi and Villani  relied on regression analysis to develop regional equations for BFI prediction for Italy.
In southeastern Australia, Nathan and Mcmahon assessed the relationship between low flow parameters and climatic and land information with multivariate regression analysis. Lacey and Grayson also used regression techniques to relate BFI, geology-vegetation groups, topographic index, and climatic index for 114 catchments in Victoria, Australia. Regression models relate baseflow and BFI to watershed characteristics in ungauged sites. The most common watershed characteristics that influence baseflow and streamflow variations reported in the scientific literature include topography, relief, climate, rainfall, evapotranspiration, slope, basin drainage area, geologic and hydrogeologic variables, soils infiltration rate, baseflow factor, and land cover. Based on previous studies, regression models have the advantage of being implemented relatively easily to estimate baseflow with reasonable accuracy.
Research in Nigeria has been conducted using statistical methods to relate baseflow and BFI to watershed characteristics such as surficial geology, land cover, degree days, and precipitation among others. Following these studies, additional independent variables with a relatively new method of hydrograph separation (i.e., the Eckhardt filter) were used to explore the relationship between baseflow, BFI, and watershed characteristics for Michigan. The objective of this study was to develop a statewide regression model as a simple approach for baseflow estimation for Nigeria using a procedure developed by Ahiablame et al. for baseflow and BFI estimation at ungauged sites. The regression models developed in this study could be useful for water planning and management decisions at the local level, adding to the existing efforts to quantify the effect of groundwater on water balance in the Great Lakes area.