Evaluation of Several Methods of Predicting Historic Water Table Levels

Show full item record

Title: Evaluation of Several Methods of Predicting Historic Water Table Levels
Author: Severson, Erik Donnally
Advisors: Dr. David Lindbo, Committee Chair
Abstract: Redoximorphic features (iron depletions) and their location in the soil indicate seasonal high water tables, which are the critical factors in siting onsite wastewater disposal systems. According to rule ?.1942 Soil Wetness? (15NCAC 18A, 2005), soil wetness conditions are determined by soil morphology is the first occurrence (≥ 2% by volume) of chroma 2 or less iron depletions Soil wetness regulations have created a need for more research in soil hydrology. The general objective of this study is to define relationships between soil morphology and water table in order to confirm or establish three water table monitoring and interpretation procedures; Calibration, Threshold, and Weighted Rainfall Index (WRI) methods. These methods are used or recognized by state regulatory agencies. Specific objectives of this study are; 1. to evaluate relationships between observed water table dynamics and soil morphology; and 2. to compare NCDENR methods of determining soil wetness to the soil morphological indicator of soil wetness Three experimental sites on the Lower Coastal Plain (LCP) were established to encompass different soil types and properties. The Richlands Site contained coarse-loamy textured soils, the Vanceboro Site contained clayey soils, and the Cool Springs Site contained sandy soils. Automated recording water table level wells monitored a total of 20 plots in transects of clayey Aquic Hapludults (Craven) and Aeric Paleaquults (Lenoir); sandy Typic Udipsamments (Tarboro and its wet-end variant) and Aquic Udipsamments (Seabrook); and coarse-loamy Aquic Paleuduults (Foreston and its wet-end variant) and Aeric Paleaquults (Stallings) for over 365 days. These water table and on-site rainfall data (one rain gauge was located at each site) were used as input information for the hydrologic model DRAINMOD (Skaggs, 1978) to independently predict 30-yr historic water table levels for each soil. Using long-term weather data, soil wetness conditions for each method were determined for each of the 20 plots. DRAINMOD was calibrated to achieve a match between measured and observed water tables in each plot by using the methodology outlined by He, et al. (2002) and in accordance to rule .1942 (f) (NCAC 2005). Once the model was calibrated, long-term water table levels were predicted using historic (+30 yr) rainfall data from the nearest weather station in order to determine the probability that a site will be saturated for a 14-d continuous period with a recurrence frequency of 30%. These values were then related to the depth of the first occurrence of three types of redox features (iron concentrations, ≤ 2 and ≤ 3 chroma iron depletions). The Threshold method is a modified form of the Threshold Wetland Simulation (TWS) as reported by Hunt et al., 2001. The requirements for obtaining the soil wetness conditions by the Threshold method were the same as those for the calibration method (14-d continuous period with a recurrence frequency of 30%). This method uses one representative soil in that can be extrapolated to other similar soils. The depths of soil wetness conditions predicted by this method were never more than 12 cm different from those wetness conditions predicted by the Calibration method at all sites. The soil wetness conditions determined by the WRI method is a sliding scale of saturation durations based on monitored rainfall data only. Those values that were obtained were variable within each plot, ranging (at all sites) from ponded conditions (0 cm) to 65 cm below the surface. Therefore, the wetness conditions predicted by this method are split into two groups, WRI-H (water table depth of 0 cm) and WRI-L (water table depth of 65 cm) for each plot. In the case of multiple years of analysis, the shallowest depth predicted by this method is the soil wetness condition (15NCAC 18A .1942, 2005). The soil wetness condition acquired by the WRI method could not be determined for every year of analysis due to lack of required rainfall. The Calibration, Threshold, and WRI methods of determining soil wetness conditions have specific relationships with the soil morphological indicator of wetness (chroma ≤ 2 depletions). For all the soils at the Richlands site and the Lenoir soils of the Vanceboro site, wetness conditions predicted by the different methods (except WRI-H) were higher in the soil profile than the morphology suggests. The level of soil wetness conditions predicted by all methods (except WRI-H) for the Craven soils of the Vanceboro site and the Tarboro (and its variant) were deeper than the levels of chroma ≤ 2 depletion or color occurrences. The wetness conditions predicted by most methods for the Seabrook soils of the Cool Springs site matched the level of the chroma ≤ 2 color occurrence. Soil wetness conditions depths predicted by the Calibration and Threshold methods related best with the depth to different redox features in the following soils. For all soil plots of the Richlands site, the average depth to soil wetness conditions determined by the Calibration method (43 cm) related best with average depth of first occurrence of iron concentrations (43 cm). The average depth of Calibration wetness condition (30 cm) related best with average depth of ≥3 depletions (32 cm) in the Lenoir soils. There was not a consistent relationship in the Aquic Hapludults between the average calibrated soil wetness condition (102 cm) and the average depth of chroma ≤ 2 depletions (52 cm). The average depth to chroma ≥3 depletions (112 cm) of the wet and dry end of the Tarboro soils related best with the average depth of soil wetness determined by Calibration method (110 cm). The average soil wetness depths obtained by the Calibration method for the Seabrook soils (55 cm) related best with depth of chroma ≤2 depletions (52 cm).
Date: 2005-05-02
Degree: MS
Discipline: Soil Science
URI: http://www.lib.ncsu.edu/resolver/1840.16/1812


Files in this item

Files Size Format View
etd.pdf 1.390Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record