Abstract
The objective of this thesis is twofold: to develop time series analysis methods for the estimation of aquifer parameters and recharge to be used in groundwater models and to develop time series analysis methods for the identification and quantification of a regime change.
In Chapter 2, a pumping test is replaced by time series analysis of heads measured in the vicinity of a well field with a strongly varying pumping regime. The step response function obtained with time series analysis provides an estimate of the steady response to pumping that would be achieved if the pumping rate was constant. The resulting virtual steady state cone of depression of the well field allows for a straightforward calibration of a regular groundwater model to estimate aquifer parameters. In addition, time series analysis can be used to determine the type of reaction, phreatic or semi-confined, in the different monitoring wells.
In Chapter 3, stream-aquifer interaction is analyzed with a time series model using a response function that is a solution to the groundwater flow equation. Head fluctuations in the vicinity of a river are analyzed, which result directly in estimates of aquifer parameters, including the resistance to flow at the interface between the stream and the aquifer. For the study site, the resistance to flow between the stream and the aquifer can be explained by stream line contraction rather than by the presence of a semi-pervious layer at the bottom of the river.
In Chapter 4, time-averaged groundwater recharge is estimated from time series models of groundwater heads that are fitted under an additional constraint that aims at better identifying the influence of evaporation. The constraint is that the seasonal harmonic of the observed head is reproduced as the response of the seasonal harmonics of precipitation, evaporation, and pumping. Better identification of the influence of evaporation results in more reliable recharge estimates to be used in regular groundwater flow models.
In Chapter 5, time series analysis is applied to identify and analyze a transition in the groundwater regime of an aquifer. The groundwater regime is defined as the range of head variations of a time series throughout the seasons. A new time series modeling approach is proposed to simulate the transition from an initial regime to an altered regime. In the case study, the estimated timing and magnitude of the transition provides strong evidence that the transition is the result of dredging works in the main river draining the aquifer. The existence of the transition of the groundwater regime had gone unnoticed, despite intensive groundwater monitoring.
This thesis showed how time series analysis can be applied to estimate the magnitude of groundwater model parameters or recharge and be applied as a tool to gain insight in the functioning of groundwater systems.
A crucial issue when estimating aquifer parameters or recharge from time series models is the uncertainty of the estimates. A modified Gauss Newton approach was used in this thesis. This approach converges quickly and provides an estimate
of confidence intervals of the estimated parameters. The systematic comparison of different estimation procedures, including Markov Chain Monte Carlo, is recommended for future study.
Groundwater modeling is based on a conceptual model of a groundwater system to simulate groundwater flow, while time series analysis can be used to estimate groundwater model parameters and identify possible changes in regimes for use in groundwater models. Both modeling approaches are complementary and it is recommended that they be applied together in a systematic fashion.
In Chapter 2, a pumping test is replaced by time series analysis of heads measured in the vicinity of a well field with a strongly varying pumping regime. The step response function obtained with time series analysis provides an estimate of the steady response to pumping that would be achieved if the pumping rate was constant. The resulting virtual steady state cone of depression of the well field allows for a straightforward calibration of a regular groundwater model to estimate aquifer parameters. In addition, time series analysis can be used to determine the type of reaction, phreatic or semi-confined, in the different monitoring wells.
In Chapter 3, stream-aquifer interaction is analyzed with a time series model using a response function that is a solution to the groundwater flow equation. Head fluctuations in the vicinity of a river are analyzed, which result directly in estimates of aquifer parameters, including the resistance to flow at the interface between the stream and the aquifer. For the study site, the resistance to flow between the stream and the aquifer can be explained by stream line contraction rather than by the presence of a semi-pervious layer at the bottom of the river.
In Chapter 4, time-averaged groundwater recharge is estimated from time series models of groundwater heads that are fitted under an additional constraint that aims at better identifying the influence of evaporation. The constraint is that the seasonal harmonic of the observed head is reproduced as the response of the seasonal harmonics of precipitation, evaporation, and pumping. Better identification of the influence of evaporation results in more reliable recharge estimates to be used in regular groundwater flow models.
In Chapter 5, time series analysis is applied to identify and analyze a transition in the groundwater regime of an aquifer. The groundwater regime is defined as the range of head variations of a time series throughout the seasons. A new time series modeling approach is proposed to simulate the transition from an initial regime to an altered regime. In the case study, the estimated timing and magnitude of the transition provides strong evidence that the transition is the result of dredging works in the main river draining the aquifer. The existence of the transition of the groundwater regime had gone unnoticed, despite intensive groundwater monitoring.
This thesis showed how time series analysis can be applied to estimate the magnitude of groundwater model parameters or recharge and be applied as a tool to gain insight in the functioning of groundwater systems.
A crucial issue when estimating aquifer parameters or recharge from time series models is the uncertainty of the estimates. A modified Gauss Newton approach was used in this thesis. This approach converges quickly and provides an estimate
of confidence intervals of the estimated parameters. The systematic comparison of different estimation procedures, including Markov Chain Monte Carlo, is recommended for future study.
Groundwater modeling is based on a conceptual model of a groundwater system to simulate groundwater flow, while time series analysis can be used to estimate groundwater model parameters and identify possible changes in regimes for use in groundwater models. Both modeling approaches are complementary and it is recommended that they be applied together in a systematic fashion.
Original language | English |
---|---|
Awarding Institution |
|
Supervisors/Advisors |
|
Award date | 7 Jan 2020 |
Print ISBNs | 978-94-028-1859-8 |
DOIs | |
Publication status | Published - 12 Dec 2019 |
Keywords
- time-series analysis
- groundwater modeling
- parameter estimation
- response function
- seasonal harmonic
- groundwater regime