TITLE:
A Novel Approach for Optimum Conjunctive Use Management of Groundwater and Surface Water Resources under Uncertainty
AUTHORS:
Gislar Edgar Kifanyi
KEYWORDS:
A Novel Approach, Conjunctive Use Management, Retrospective Optimization Approximation (ROA), Uncertainty
JOURNAL NAME:
Open Journal of Modern Hydrology,
Vol.13 No.1,
January
19,
2023
ABSTRACT: Uncertainty in determining optimum conjunctive water use lies not only on
variability of hydrological cycle and climate but also on lack of adequate data
and perfect knowledge about groundwater-surface water system interactions,
errors in historic data and inherent variability of system parameters both in space and time. Simulation-optimization models are
used for conjunctive water use management under uncertain conditions.
However, direct application of such approach whereby all realizations are
considered at every-iteration of the optimization process leads to a highly
computational time-consuming optimization problem as the number of realizations
increases. Hence, this study proposes a novel approach—a Retrospective
Optimization Approximation (ROA) approach. In this approach, a simulation model
was used to determine aquifer system responses (draw-downs) which were
assembled as response matrices and incorporated in the optimization model
(procedure) as coefficients in the constraints. The sample optimization
sub-problems generated, were solved and analyzed through ROA-Active-Set
procedure implemented under MATLAB code. The
ROA-Active Set procedure solves and evaluates a sequence of sample path
optimization sub-problems in an increasing number of realizations. The
methodology was applied to a real-world conjunctive water use management
problem found in Great Letaba River basin, South Africa. In the River basin,
surface water source contributes 87% of the existing un-optimized total
conjunctive water use withdrawal rate (6512.04 m3/day) and the remaining 13% is contributed by groundwater
source. Through ROA approach, results indicate that the optimum
percentages contribution of the surface and subsurface sources to the total
water demand are 58% and 42% respectively. This implies that the existing
percentage contribution can be increased or reduced by ±29% that is groundwater
source can be increased by 29% while the surface water source contribution can
be reduced by 29%. This reveals that the existing conjunctive water use
practice is unsustainable wherein surface water system is overstressed while
groundwater system is under-utilized. Through k-means sampling technique ROA-Active Set procedure was able to
attain a converged maximum expected total optimum conjunctive water use
withdrawal rate of 4.35 × 104 m3/day within a
relatively few numbers of iterations (6 to 8 iterations) in about 2.30 Hrs. In
conclusion, results demonstrated that ROA approach is capable of managing
real-world regional aquifers sustainable conjunctive water use practice under
hydro-geological uncertainty conditions.