Graduation Year

1998

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Department

Geology

Degree Granting Department

Geology

Major Professor

Mark Stewart, Ph.D.

Committee Member

Mark Ross, Ph.D.

Committee Member

Jeff Ryan, Ph.D.

Committee Member

Carl Steefel, Ph.D.

Committee Member

H. Leonard Vacher, Ph.D

Abstract

Augmentation with reclaimed water is one method for rehydrating wetlands damaged by water-level declines. Augmentation with reclaimed water has been proposed for rehydrating a wetland in the covered-karst terrane of west-central Florida. There is concern because reclaimed water may contain harmful agents that could flow from the wetland 1.4 km to a municipal wellfield that withdraws 30,000 m3/day. Estimates of groundwater flow velocities were calculated from the results of detailed field studies at the wetland. Results indicate that groundwater flows downward in the surficial aquifer at rates of 0.1 to 0.2 m/day and horizontally in the Floridan aquifer at rates of 0.02 to 0.5 m/day. Sinkholes do not appear to be preferential pathways for downward groundwater flow at the site because the confining layer is thin to absent, and low-permeability soil layers may restrict flow within the sinkholes.

A method was developed for incorporating large-scale, vertical fracture zones into two-dimensional, finite- difference models of groundwater flow (MODFLOW) and particle tracking (P A TH3D). Fracture zones are included in the numerical model by increasing transmissivity values in selected cells. The selection of cells is based on existing patterns of fracture zones, which can be inferred from maps of photolineaments. To quantify the uncertainties in knowing the exact location, orientation, and hydraulic properties of the fracture zones, these parameters are treated stochastically through Monte Carlo simulation. Stochastic results from an example simulation of groundwater flow and advective transport between the North Lakes wetland and Section 21 wellfield indicate a 50 and IO percent chance that travel times will be less than 9 and 2 years, respectively.

The probability of rehydrating a wetland with an additional source of water can be estimated by combining a form of the water balance equation with Monte Carlo analysis. Parameters that contain large uncertainties, such as future values of rainfall, ET, and groundwater flow, are treated stochastically, while the additional source of water is treated deterministically. Stochastic simulation results in an ensemble of water-level time series that contain all statistically reasonable predictions. Statistics performed on the ensemble indicate that worst-case predictions are highly improbable, and that a recharge rate of 5700m3/day will maintain the target water level in the example wetland used to illustrate the method.

Included in

Geology Commons

Share

COinS