Graduation Year

2022

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Civil and Environmental Engineering

Major Professor

Mahmood H. Nachabe, Ph.D.

Committee Member

Mauricio E. Arias, Ph.D.

Committee Member

Hadi Charkhgard, Ph.D.

Committee Member

Lu Lu, Ph.D.

Committee Member

Louis H. Motz, Ph.D.

Keywords

North Florida, GCM, Hindcasting, Predevelopment, pumping

Abstract

Long-term groundwater management relies on forecasts of decadal or longer groundwater levels driven by nested scales of variability in climate. Discerning the impacts of anthropogenic stressors on groundwater is also critical for establishing predevelopment benchmark groundwater conditions and developing climate change adaptation strategies to reduce risks and increase resiliency. This research demonstrates the development and successful applications of a new methodology to predict and assess multidecadal groundwater dynamics for understanding long-term effects of climate change and groundwater withdrawals.

A Physically Constrained Wavelet-Aided Statistical Model (PCWASM) is first introduced to analyze and predict monthly aquifer levels on multidecadal time scales. The approach retains the simplicity of regression modeling but is constrained by temporal scales of processes responsible for groundwater level variation, including aquifer recharge and pumping. The methodology integrates statistical correlations enhanced with wavelet analysis into established principles of groundwater hydraulics including convolution, superposition, and the Cooper-Jacob solution. The systematic approach includes (1) identifying hydrologic trends and correlations using cross-correlation and multi-time scale wavelet analyses, (2) integrating temperature-based evapotranspiration and groundwater pumping stresses, and (3) assessing model performances using fixed-block k-fold cross-validation and split calibration-validation methods. The approach is applied three hydrogeologicaly distinct sites in North Florida using over 40 years of monthly groundwater levels. The models performed well for predicting monthly groundwater levels from 7 to 22 years with less than 2.1 feet errors. Results reveal low-frequency climatic signals and long-term variation in cumulative ET are critical for multidecadal aquifer level predictions.

The PCWASM is later utilized to hindcast the predevelopment groundwater levels back to early 1900s using limited records and to examine the temporal evolution of climatic and pumping impacts nested within periodic natural cycles and trends. The model and hindcasts are tested using traditional split calibration-verification methods for the period of record and with the documented historical drought and wet years. The pumping impact is quantified over time and compared with regional groundwater models, revealing that withdrawals are responsible for 30 to 70% of the declines in levels since 1960s. Hindcasting yielding 110 years of monthly levels is used to assess the effect of climate change and pumping on the frequency of critical low levels. At all three sites, the frequencies of critical low levels increase significantly in the 1960-2015 period when compared to the 1904-1959 period. For example, at site 1, the return period of the critical low level is shortened by 3.9 years due to climate change and 2.2 years due to pumping.

Finally, long-term groundwater level trends are forecasted and examined using a large ensemble of global climate model (GCM) projections under low and medium emission scenarios. The forecasts from 2020 to 2099 indicate groundwater levels may continue to decline, however, at an accelerated pace after 2040s reaching critical levels by the end of this century. Pumping impact constitutes 10 to 45% of future declines but is amplified by enhanced drought. As groundwater responds to temperature, rainfall and pumping differently, influence of each driver on future trends is also examined separately. Results show highly divergent groundwater response to projected hydroclimatic changes in that future long-term rainfall trend may lead to rising groundwater levels, which, however, may be overshadowed by heightened ET loss driven by global warming and increased groundwater abstraction, hence, causing steep declines. This study also reveals poor performance of predictions driven by GCM projections in replicating the timing of high and low extremes, attributed to failure of GCM projections and downscaling methods to capture the timing of climatic cycles, controlling hydrologic memory. Additionally, a multidecadal harmonic trend analysis exposes presence of potential centennial cyclic trends in groundwater levels, critical for future predictions. Hence, GCM-based forecasts are recommended to be cautiously utilized for groundwater resource planning when significantly departing from historical long-term cyclic patterns.

The PCWASM is a computationally efficient parsimonious approach to predict multidecadal groundwater dynamics with the ability to discern impacts of pumping and climate change on aquifer levels and can be implemented using publicly available datasets. Thus, it should provide a versatile tool for managers and researchers for predicting and analyzing multidecadal monthly aquifer levels under changing climate and pumping stresses.

Download

Share

COinS