Numerical Groundwater Flow Modeling of the Akaki River Catchment

Ebasa, Oljira (2006) Numerical Groundwater Flow Modeling of the Akaki River Catchment. Masters thesis, Addis Ababa University.

PDF (Numerical Groundwater Flow Modeling of the Akaki River Catchment) Ebasa, Oljira.pdf - Accepted Version Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

Akaki River Catchment is a sub catchment of Awash drainage basin with an approximate surface area of around 1462km2, boundary length of 216km and it lies at the eastern edge of the Western Ethiopian plateau that descends to the Main Ethiopian Rift. The capital city, Addis Ababa and other smaller towns are found in this catchment. The catchment is totally covered with volcanic rocks of various ages that correspond to different stratigraphic units. The rocks were subjected to rift tectonics that is manifested by a number of fault systems having a general trend of the rift system (NE – SW). As numerical groundwater flow models represent the simplification of complex natural systems, different parameters were assembled into conceptual model to represent the complex natural system in a simplified form. The conceptual model was input into the numeric model to examine system response. Numerical groundwater simulation was carried out using MODFLOW, 1996 (McDonald and Harabaugh, 1988). Two dimensional profile model was developed considering the system to be under steady state condition and assuming flow system view point. Three scenarios of increased withdrawals and one scenario of decreased recharge were simulated to study system response. Model calibration was carried out by trial and error calibration method using groundwater contours constructed from heads collected in 122 observation points. The calibration showed that about 81% of simulated heads were within the calibration target and the overall root mean square error for simulated hydraulic heads is about 10.42m. The poor fit at some points was due to numerous limitations associated with the model. Model sensitivity analysis was conducted by taking recharge and hydraulic conductivity as the model is most sensitive to them. A change in recharge by 20%, 40%, 60%, -20%, - 40%, and -55% resulted in RMS head changes from the calibrated value by 17%, 58%,106%, 26%, 86% and 147%, respectively. Equal changes in hydraulic conductivity (in the order mentioned for recharge) resulted in RMS head changes from calibrated value by 12%, 33%, 55%, 19%, 90% and 193%, respectively. In addition, the effect of varying these two parameters on stream leakage was tested. Accordingly, changes in steady state estimated recharge by 15%, 30%, 45%, -15%, -30%, and -45% resulted in change in stream leakage from calibrated value by 12%, 24%, 36%, -20%, -24% and -28%, respectively. The same changes in hydraulic conductivity (in the order mentioned for recharge) resulted in stream leakage changes from calibrated value by 1.9%, 3.7%, 5.5%, -2.1%, -4.2% and -6.9%. The results of the numerical simulations showed that increased well withdrawals by 15%, 25% and 45% resulted in RMS (Root Mean Square) head changes of 0.7m, 1.3m, and 2m, respectively. The same change in well withdrawal resulted in 1.4%, 2.4% and 3.7% respective changes in river leakages compared to the steady state simulated value. Similarly, these increased withdrawals resulted in reduction of the calibrated subsurface outflow by 3.4%, 6.4%, 9.1%, respectively. On one hand 20% decrease in steady state simulated recharge resulted in reductions of groundwater level by 5.6m, stream leakages by 15.8% and subsurface outflows by 5%.

Item Type:	Thesis (Masters)
Subjects:	Q Science > Q Science (General) Q Science > QE Geology
Depositing User:	Selom Ghislain
Date Deposited:	27 Jun 2018 11:54
Last Modified:	27 Jun 2018 11:54
URI:	http://thesisbank.jhia.ac.ke/id/eprint/6042

Actions (login required)

View Item