Sensitivity and uncertainty analysis on water quality modelling of Aguamilpa reservoir

Abstract

The use of water quality models is determined to a great extent by their ability to accurately reproduce observed data series and by their predictive capability without the need to adjust the calibrated parameters. However, the observed abiotic variables involved in a system are measured with a level of uncertainty. The sensitivity analysis concepts and generalized methodology of uncertainty analysis were used via a computer tool called UNCSIM. As a first step, a parametric optimization model of the Aguamilpa reservoir water quality was accomplished. The aforementioned analysis identified that the wind sheltering coefficient (WSC), Chezy bottom friction solution (FRICC), and coefficient of bottom heat exchange (CBHE) were the parameters of the CE-QUAL-W2 model that significantly influenced the behaviour of temperature and dissolved oxygen concentration in the reservoir. Afterwards, the uncertainty of the water quality model was evaluated through the modification of the hydrological and climatological information, which had a major influence on the simulation of the system. This analysis showed the possible changes in hydrodynamic and water quality characteristics of the reservoir, including an increase in the thermocline due to a possible air temperature increase and a rainfall decrease in the region. The innovative coupling routine of the different modules for the sensitivity and uncertainty analysis developed in this research establishes the basis for the future development of a modelling platform to conduct water quality simulations of the Aguamilpa reservoir in real time through continuous meteorological and water discharge information.