Date of Award

2017

Document Type

Open Access Thesis

Department

Art

Sub-Department

Earth and Environmental Resource and Management

First Advisor

Dwayne Porter

Abstract

The South Carolina Department of Health and Environmental Control (SCDHEC), Surface Water Monitoring Section utilizes a probability survey to assess water quality conditions of streams and rivers in the state. The Generalized Random-Tessellation Stratified survey design prioritizes specific stream order (size) subpopulations and spatial distribution in the generation of potential sample sites, and is intended to survey only perennial stream reaches. Site reconnaissance is conducted to assess site suitability and accessibility. The stream characteristics associated with accepted and rejected sites have been recorded since the late 1990s/early 2000s. Data from 2001-2016 was analyzed to assess apparent trends in both rejected and accepted site characteristics. These trends were evaluated based on frequency of rejection reasons and affected stream orders. Apparent trends in the characteristics of rejected sites were compared against characteristics of sampled sites. A second component to the site rejection assessment determined if there was an association between landcover or land use and site rejection or acceptance. Another focus of the study assessed differences between two digital hydrological networks in South Carolina at the 1:100,000 scale. The EPA Reach File 3.0 network currently used as the survey network was compared with the National Hydrography Dataset Plus (NHDP) network, a combined product between the U.S Environmental Protection Agency and U.S Geological Survey. The two networks were compared for the Pee Dee and Savannah River Basins, and for 4 subwatershed 12-digit Hydrological Code Units (HUCs). The subwatersheds were located in two distinct ecoregions of South Carolina, the Piedmont and Coastal Plain. The advantages and drawbacks of both networks were assessed based on network definition and site reconnaissance.

The leading reasons for sites to be rejected were intermittency and inaccessibility. Intermittency refers to streams that do not flow year-round, and which do not belong in the sample frame. Intermittency was most often observed in the smallest 1st order streams. Inaccessibility affected all orders to some degree, and refers to sites that did not have an acceptable location from which to sample; the sites are assumed to be target population sites as they cannot be verified via site reconnaissance. Landcover associations assessed at the subwatershed regional scale revealed that site accessibility appeared to be a greater issue in networks located in a rural environment than networks located in urban areas. The two hydrological networks were similar in network linear definition, with differences in stream mileage primarily a result of differences in level of stream connectivity rather than spatial disagreement. Variation in stream density in the NHDP 1:100,000 network prevents SCDHEC from utilizing it for the probability survey. However, if the coverage scales were addressed, the NHDP network has useful attributes, such as identifying streams as perennial or intermittent. The network is not presumed to be exact, but the ability to exclude a significant proportion of non-target streams would be advantageous. Further evaluation and statistical analysis are recommended to determine if SCDHEC would benefit from changing the reference digital hydrological network to the NHDP, as these results suggest.

Rights

© 2017, Caitlin Smith

Download

Share

COinS