Date of Award

8-2014

Document Type

Thesis

Degree Name

Master of Science (MS)

Legacy Department

Forest Resources

Advisor

Jayakaran, Anand D

Committee Member

Hitchcock , Daniel R

Committee Member

Chow , Alex T

Committee Member

Post , Christopher J

Abstract

Flooding and stormwater control is a critical issue in coastal South Carolina because of shallow water table elevation, topography and rapid urbanization in the region. A best management practice (BMP) using low impact design (LID) principles known as a bioretention cell (BRC) is gaining popularity for stormwater management. Five BRCs in four landscape positions (well-drained uplands, tidal-proximal, poorly-drained-uplands, and floodplain) were instrumented for microclimate, soil moisture, and water table elevation for hydraulic efficiency and for water quality measurements. Three BRCs did not have an overflow outlets, one BRC (floodplain) employed an underdrain system, and one BRC (tidal proximal) had an overflow outlet. Temporal analysis of water table data showed that water table elevation exhibited seasonal fluctuations at all the sites. The well drained uplands and poorly drained uplands BRCs had a shallow water table during the growing season and a deep water table during the dormant season. Groundwater at the tidal proximal BRC reflected semidiurnal fluctuations in level but on a seasonal basis was relatively static compared to the seasonal variation of groundwater at other sites. In situ conductivity measurements of groundwater at the tidal proximal BRC, showed a spike in conductivity between October and December 2012 after the passage of Hurricane Sandy. The floodplain BRC water table was localized by a confining clay layer and showed little seasonal variation, much like the poorly drained uplands BRC. However, within storm events, groundwater variation at these two BRCs were large compared to the well-drained uplands and tidal proximal BRC. Small diurnal fluctuations in water table elevation occurred during dry days caused by potential evapotranspiration (PET). A linear regression analysis showed a significant relationship between each individual BRC's daily variation in soil moisture content (dry days only) and the daily PET. Soil moisture content was monitored in three zones (surface, shallow, and deep). Soil moisture content in the surface and shallow zones increased during the growing season and decreased during the dormant season within the well-drained upland BRCs. However, this trend was reversed within the poorly drained upland and floodplain BRCs. Seasonal trends in soil moisture within the tidal proximal BRC were not measured due to sensor malfunction. Infiltration rates measurements at every BRC exceeded published infiltration rates for the surrounding parent material. The highest infiltration rate was at the upland BRCs (93.7 cm/hr) where the media and parent material was predominantly sand. The lowest infiltration rate was found at the poorly drained upland BRC (19.9 cm/hr). There was a significant relationship by linear regression between the peak infiltration rate and peak rainfall intensity, while no significant relationships were found between the peak infiltration rate and soil moisture content or peak infiltration rate and water table elevation. The BRCs were sampled for water quality during storm events along a primary stormwater path (inflow through the soil profile to the groundwater) and secondary path (inflow to the overflow/outlet where available) for water quality and instrumented for water table elevation and soil moisture. Samples were also collected during non-storm days to estimate ambient nutrient concentrations in the groundwater. Linear regressions were used to evaluate the removal (slope) and determine a calculated ambient concentration based on the y-intercept of the regression line. Nutrients and bacteria tested were nitrate, ammonia, total dissolved nitrogen (TDN), non-purgeable organic carbon (NPOC), Phosphate (PO43-), total coliform bacteria, and E. coli bacteria. The primary pathway within all the BRCs showed removal of 87 to 98% for NO3-, with the exception of the poorly drained uplands BRC (56-66%). The removal of NO3- along the secondary pathway was 26-32%. The removal of ammonia along the primary pathway within floodplain and tidally proximal BRCs was 74-96%, while at the remaining two upland BRCs and along all the secondary stormwater pathways, ammonia was being exported suggesting a short-circuiting along the secondary pathways. All the BRCs showed removal of TDN (78-99%) along the primary stormwater path with the exception of the floodplain BRC that exported TDN. The total coliform bacteria percent removal for all the BRCs was high (89-99%) with the exception of the tidally proximal and poorly drained upland BRCs along the primary stormwater path. All the sites had high removals of E. coli bacteria (93-97%) along the primary stormwater pathway. There was variability in nutrient and bacteria removal rates that appeared to be linked to differences in landscape position, water table, and soil moisture. Multivariate linear regressions were used to incorporate the variable hydrology of the coastal landscape to understand how hydrology and hydraulics affect nutrient and bacteria removal at varying depths.

Share

COinS