Estimation of groundwater recharge in Bugesera region (Burundi) using soil moisture budget approach
Laboratory for Applied Geology and Hydrogeology, Ghent University, Krijgslaan 281-S8, B-9000 Gent, Belgium; University of Burundi, Department of Earth Sciences, Bujumbura, Burundi
Laboratory for Applied Geology and Hydrogeology, Ghent University, Krijgslaan 281-S8, B-9000 Gent, Belgium
Laboratory for Applied Geology and Hydrogeology, Ghent University, Krijgslaan 281-S8, B-9000 Gent, Belgium
Abstract
Groundwater recharge for Bugesera region, a potable water-scarce area in northeastern Burundi, is computed using the soil moisture budget technique. Five evapotranspiration methods including Hamon, Hargreaves, Thornthwaite and two modifications of the original Thornthwaite method are evaluated in comparison to the reference evapotranspiration method, i.e. the FAO Penman-Monteith equation for years where complete climatic data is available. The evapotranspiration calculated by the aforementioned methods along with rainfall data are used to compute the soil moisture budget. The latter is calculated using the methodology devised by Thornthwaite & Mather (1955). Recharge calculation is performed using both the Thornthwaite Monthly Water-Balance Model (henceforth TMWB model) and excel sheets wherein each term of the soil moisture budget is computed separately. The results of evapotranspiration calculations show that, while the other evaporation methods slightly to moderately underestimate or overestimate the potential evapotranspiration in comparison to the FAO Penman-Monteith method, Hargreaves equation aberrantly overestimates this parameter. Likewise, groundwater recharge estimated using Hargreaves’ evapotranspiration is dramatically reduced in comparison to the other evapotranspiration methods. Moreover, this study clearly shows that the time discretisation used in recharge calculations has important consequences, the use of smaller time steps leading to enhanced recharge. This better corresponds to reality. Compared to the recharge values obtained on a daily basis with Penman-Monteith evapotranspiration, the TMWB model which is on a monthly basis, using Hamon’s evapotranspiration, gives the best approximations of reality, with the advantage of needing much less data. The distribution pattern of monthly recharge features a bimodal pattern somewhat similar to that of the monthly rainfall with an important peak in April.