Search

Many sampling projects have been initiated or modified in recent years to quantify the effects of water quality protection and enhancement programs. Although comprehensive references on the theory and procedures related to discharge data collection have been published, similar guides to water quality sampling are not available. Several sources provide general guidance on sampling project design and on manual sampling procedures, but only recently has detailed information on automated storm water quality sampling been developed. As a result, a compilation of available information on the design of water quality sampling projects is needed to support sound decision-making regarding data collection resources and procedural alternatives. Thus, the objective of this article is to compile and present practical guidance for collection of discharge and water quality constituent data at the field and small watershed scale. The guidelines included are meant to increase the likelihood of project success, specifically accurate characterization of water quality within project resource constraints. Although many considerations are involved in establishing a successful sampling project, the following recommendations are generally applicable to field and small watershed studies: (1) consider wet-weather access, travel time, equipment costs, and sample collection method in the selection of sampling site numbers and locations; (2) commit adequate resources for equipment maintenance and repair; (3) assemble a well-trained, on-call field staff able to make frequent site visits; (4) establish reliable stage-discharge relationships for accurate discharge measurement; (5) use periodic manual grab sample collection with adequate frequency to characterize baseflow water quality; (6) use flow-interval or time-interval storm sampling with adequate frequency to characterize storm water quality; and (7) use composite sampling to manage sample numbers without substantial increases in uncertainty.

The Better Assessment Science Integrating point and Nonpoint Sources (BASINS) system was developed by the U.S. Environmental Protection Agency to facilitate developing total maximum daily loads (TMDLs). The Soil Water Assessment Tool (SWAT) is one of the watershed-scale simulation models within BASINS. Because of the critical nature of the TMDL process, it is imperative that BASINS and SWAT be adequately validated for regions on which they are being applied. BASINS and SWAT were tested using six years of hydrologic data from a 22 km2 subwatershed of the Little River in Georgia. Comparisons were made between water balance results obtained using high and low spatial resolution data as well as those obtained using default initial parameters versus those modified for existing groundwater conditions. In general, all scenarios simulated general trends in the observed flow data. However, for the years with lower precipitation, the total water yields simulated with the low spatial resolution data and the default initial conditions were overpredicted by up to 27% of the annual precipitation input. Total water yields simulated using the high spatial resolution input data were within 20% of the observed yields for each year of the assessment. Nash-Sutcliffe model efficiencies (E) for monthly total water yields were 0.80 using the high spatial resolution data with the modified initial conditions and 0.64 using the low spatial resolution data with the default initial conditions. While the model simulated general streamflow trends, discrepancies were observed between observed and simulated hydrograph peaks, time to peak, and hydrograph durations. A one-day time lag between the simulated and observed time to peak was the primary cause of large errors in daily flow simulations. Model modification and more extensive calibration may be necessary to increase the accuracy of the daily flow estimates for TMDL development.

The Riparian Ecosystem Management Model (REMM) was used to simulate shallow groundwater movement, water table depths, surface runoff, and annual hydrologic budgets for a Coastal Plain riparian buffer system near Tifton, Georgia, USA. The riparian buffer consisted of zone 3 (grass downslope from a row-crop field); zone 2 (mature pine forest downslope from zone 3); and zone 1 (mature hardwood forest downslope from zone 2, adjacent to stream). Measured surface runoff and shallow groundwater movement from the adjacent agricultural field were used as the hydrologic input to REMM. Uncalibrated simulation results for a five-year period were compared to measured values for the same time period. The overall error in zone 2 and zone 1 mean water table depths was about 0.07 m, although absolute errors were higher. The water table dynamics simulated by REMM were similar to observed although lags were observed in the response of the simulated water table to large rainfall events. Mixed results were obtained from observed versus simulated surface runoff comparisons, primarily due to large variability in observed runoff depths along the riparian transect. Simulated surface runoff depths for zone 3 were within one standard deviation for four out of the five years. For zone 2, surface runoff depths could only be simulated within one standard deviation for two out of the five years. Simulated seasonal total depths of surface runoff did not always agree with observed values but usually followed both similar temporal and spatial patterns. Annual hydrologic budgets produced total streamflow comparable to those estimated for the riparian buffer site. These results provide an adequate basis for subsequent testing of other REMM model components including water quality and nutrient cycling.

The Georgia Coastal Plain of the United States holds significant agricultural and hydrologic importance. The profitability and sustainability of agriculture in this and other regions are dependent upon climatic patterns and, in particular, rainfall. The temporal and spatial variability of this rainfall play key roles in agricultural management. Developing an understanding of seasonal patterns and individual storm characteristics is critical. Thirty years of rainfall data collected from a dense rain gage network on the 334 km2 Little River Watershed near Tifton, Georgia, were analyzed for this purpose. Storm patterns were characterized by season to establish means and trends. Individual storm characteristics and spatial correlation patterns within storms were quantified. Rainfall patterns, although highly variable from year to year, show rainfall to be greatest in the midsummer months with high intensity, convective thunderstorms. While these summer storms yield relatively low rainfall depths, they occur more frequently than during other seasons. Frontal storms with moderate rainfall amounts are typical of the winter and spring months. The fall months generally have low rainfall totals and storms during this time occur less frequently than during the other months. For larger storm events, defined here as those where at least one rain gage in the network measured 25.4 mm or greater during the event, the mean storm depth weighted over the mean storm coverage of 295 km2 was 20.6 mm, while the mean storm duration was 7.2 h. For these larger events, the summer storms are separated by the least time between events (mean of 116 h) while the fall events are separated by the greatest time (mean of 291 h). For summer events, rain gage depths for individual storm events collected by gages separated by 1.9 km or less are likely to be highly correlated (r iÝ 0.9). This distance increases to 9.2 km for the winter. High correlations (r iÝ 0.9) are expected up to distances of 5 km throughout most of the year, except summer. This analysis establishes the basis for detailed modeling across this and other watersheds in the region.

The effectiveness of mature riparian forests in reducing the impact of agriculture on the quality of the nation’s water resources has been documented, but the impact of forest management practices implemented within riparian forest buffers on their water quality function has not been evaluated. This article examines the effect of forest management within a Coastal Plain riparian forest buffer system (RFBS) on runoff and sediment transport over a four year period. The RFBS, which conformed to USDA-FS and USDA-NRCS best management recommendations, included a narrow strip of undisturbed forest located adjacent to the stream drainage system (Zone 1), a wide managed pine forest downslope from the grass filter (Zone 2), and a narrow grass filter strip immediately downslope from an agricultural field (Zone 3). Forest management treatments evaluated within Zone 2 were mature forest, clear-cut, and selectively-thinned. Significant reductions in runoff and sediment transport were measured under all three forest management treatments. The primary zone of runoff and sediment reduction was within the grass filter portion of the RFBS. These results indicate that riparian forests within a RFBS may be managed for economic return to the land owner without adversely affecting the runoff and sediment reduction function performed by these buffer systems.

The objective of this article was to develop hydrologic information needed for nonpoint source water quality screening and assessment on agricultural watersheds in the Gulf-Atlantic Coastal Plain of the southeastern U.S. The USDA-Agricultural Research Service (ARS) has instrumented study areas in the Coastal Plain that provide rainfall and streamflow data from seven agricultural watersheds ranging from 15.7 to 334 km2 (6.0-129 mi2). Hydrologic budgets were developed for Little River experimental watersheds using long-term (up to 25 yr) water balance analyses and results of previous USDA-ARS research. The primary water balance loss component was evapotranspiration, which accounted for 69% of annual rainfall. Streamflow, the other major loss component, accounted for 30% of annual rainfall. Water balance analyses provide insight into the relative magnitudes of hydrologic budget components and their interactions that is useful for understanding landscape and watershed scale hydrologic and water quality processes on Coastal Plain watersheds where hydrologic response is influenced by riparian storage. Two forms of simple rainfall-streamflow relationships were evaluated for estimating annual water yields: constant percentage (runoff coefficient) relations and linear rainfall-streamflow regressions. Annual runoff coefficients, which averaged 0.30 for the study areas, were related to relative watershed riparian area. Watershed riparian area was more effective than upland land use for predicting annual runoff coefficients. Simple linear rainfall-streamflow regressions were found to more adequately represent year-toyear variability in streamflow production. Use of linear regression models eliminates the bias in water yield estimates inherent to constant percentage relations. Averages of 716 to 772 mm (28.2-30.4 in.) of rainfall were required annually to meet evapotranspiration and other losses on study areas before significant streamflow was produced. About 75% of rainfall in excess of the threshold was measured as streamflow. The effect of rainfall distribution on annual streamflow volume was evident, with two distinct seasons for production of streamflow noted. Rainfall-streamflow relations developed provide regional hydrologic information needed for nonpoint source water quality screening and assessment, and for water resources planning and management on low-gradient, humid region watersheds of the Coastal Plain of the southeastern U.S.

A low-impact surface flow sampler was developed for riparian studies conducted in the Coastal Plain region of the southeastern United States. The device consists of two primary components, a splitter and a collector, which were used for unattended sampling of surface flow in riparian buffer study areas. This low-cost device provides a composite event sample at selected locations within experimental areas. The quantity of sample is adequate for laboratory analyses of dissolved and suspended constituents for both large and small flow events, and permits estimation of the volume of surface flow at the sampling location. Installation and operation of the device requires little disturbance to the riparian buffer ground surface and vegetation, or to surface flow within experimental areas.

Water table gradients were measured and saturated flow rates estimated for a hillslope consisting of a tilled upland field and a downslope riparian forest buffer system located in the Gulf-Atlantic Coastal Plain Tifton-Vidalia Uplands. Three years of water table measurements and estimates of saturated hydraulic conductivity were used to evaluate and quantify saturated water flow gradients, directions, and rates. Forest treatments consisting of clear cutting, thinning, and no cutting were examined. The gradient of the water table from the top of the landscape to the bottom varied from 0.9 to 0.2%, less than the 1.5% land slope. The direction of groundwater flow generally followed the land slope. However, during summer months the hydraulic gradient within the forested buffer reversed direction. Water table data indicate the riparian area was saturated from January through March. During this time, flow direction in the shallow aquifer is from the top of the field to the stream bottom. During summer months, high rates of forest evapotranspiration created large water sinks in the shallow subsurface and large local hydraulic gradients. Examination of water table elevations indicates the seasonal water demand of the forest shifts the direction of shallow subsurface aquifer flow. During these periods flow direction within the riparian buffer was from the lowest landscape position to the riparian forest, reversed from winter months. Total subsurface flow within the hillslope was calculated as 35 mm yr–1, 3% of average annual precipitation. Average groundwater linear velocity was calculated as 1.4 mm h–1. Evapotranspiration loss was estimated as 67% of average annual precipitation.

A data management system (DMS) was developed for use with hydrologic databases collected on Little River Experimental Watersheds (LRW) for the 25-year period, 1968 to 1992. The system consists of data management protocols and computer programs that edit field records, detect and correct erroneous data, and provide initial levels of summarization. The system provides logic for filling of missing records, as well as provisions for storage, retrieval, and backup of data. This DMS has produced quality hydrologic databases for use in water resource and environmental quality research, and has resulted in timely availability and summarization of data with low percentages of missing or estimated records.

Available procedures for estimating storm hydrograph time parameters performed poorly in applications on low-gradient drainage systems of the Coastal Plain and Flatwoods regions of the southeastern United States. Existing empirical relationships generally underpredicted observed hydrograph time parameters on nine Coastal Plain and Flatwoods watersheds, with the standard error of estimate ranging from 63 to 132% of observed means. Hydrograph time parameters from flatland study areas were related to watershed physical characteristic and geomorphic data. The simple parameter, length of main channel, proved superior to all other simple or complex watershed characteristics for explaining observed variations in watershed time-of-concentration and hydrograph time-to-peak. Simple empirical relationships developed for estimating hydrograph time parameters for flatland areas provide needed information for watershed scale hydrologic design and environmental resource modeling applications on low-gradient drainage basins similar to those of the coastal regions of the southeastern United States.

The Rainfall Erosion Index (EI) is one of the primary factors in the Universal Soil Loss Equation. Calculation of EI is based on breakpoint rainfall records where breakpoints separate periods of constant rainfall intensity. Breakpoint data are determined from continuously recording, chart-type raingage records. This article examines the use of rainfall records for EI determination where accumulated depth is recorded at fixed time increments. Various time (1, 2, 3, 5, 6, 10, 15, 30, and 60 min) and depth (0.25, 0.5, 0.75, 1.0, 2.0, 3.0, 4.0, and 5.0 mm) resolutions were superimposed on breakpoint rainfall records from Tifton, GA to obtain equivalent non-breakpoint records. Results of statistical analyses relating non-breakpoint EI values to breakpoint EI values are presented, along with correction factors for adjusting non-breakpoint EI to breakpoint EI and a relationship defining optimum time and depth resolution combinations for accurate EI calculation. KEYWORDS. Rainfall, Erosion, Runoff.

Sediment and nutrient loadings in the Little River Research Watershed in south central Georgia were modeled using the continuous simulation Annualized Agricultural Nonpoint–Source Pollution (AnnAGNPS) model, part of the AGNPS suite of modeling components. Specifically, nitrogen, phosphorus, sediment, and runoff were predicted over a seven–year period. Land under cultivation makes up approximately 25% of the 333 km2 watershed. Livestock facilities include swine, poultry, dairy cows, and beef cattle. Results from the simulation were compared to seven years of monitoring data at the outlet of five nested subwatersheds and at the outlet of the Little River Research Watershed (LRRW). The average annual predicted runoff in the upper part of the watershed was one–third to half of observed runoff. In contrast, predicted runoff in the lower part of the watershed was close to observed, and was 100% of observed at the outlet of the watershed. Runoff underprediction was attributed to the method of landcover discretization. The extent of forest land in the upper watershed (55% to 63%) and the fragmented landscape that has relatively small fields surrounded by riparian forests and tracts of forest resulted in overestimation of forested area in the watershed. In addition to runoff, sediment and nutrient loads were also underpredicted in the upper part of the LRRW. Two factors are most likely responsible for underprediction. Runoff is underpredicted at these sites, which reduces the carrying capacity of sediment loads. In addition, the overestimation of forested areas at these sites coincides with underestimation of sediment–producing areas, such as cropland. In contrast to the upper part of the watershed, sediment and nutrient loads were overpredicted in the lower part of the watershed. This may have resulted from inadequately simulating nonpoint–source pollution attenuation by the extensive riparian forests and forested in–stream wetland areas found in these watersheds. Prediction results can be improved through better input into the model, as well as modification of the processes within the model to account for forest and riparian conditions.

An empirical equation based on storm data from the midwestern U.S. has been used in a number of water resource and water quality models for estimating peak flows for storm events on ungaged watersheds. Limited testing of the equation has shown a tendency to overpredict storm peaks in regions of low topographic relief. The objective of this study was to test the Smith and Williams equation and develop improved methods for estimating peak flows from ungaged watersheds in the southeastern U.S. Coastal Plain, a region characterized by low–gradient drainage networks and extensive riparian storage. The equation was evaluated on 58 storm events from seven experimental watersheds. Results indicate that the Smith and Williams equation significantly overpredicts peak flows in the Coastal Plain region, with an average error in peak estimation of over 250%. Regional peak flow equations were developed using stepwise linear regression analyses of log–transformed storm peak flows and watershed and storm event characteristics. These equations, which relate storm peaks to drainage area, runoff amount, and a drainage network parameter, explained up to 94% of the variability in measured peaks from Coastal Plain watersheds. The Coastal Plain peak flow equations provide improved algorithms for estimating storm peaks from ungaged watersheds in the region.

CHEMICAL load of nitrate-plus nitrite-nitrogen and orthophosphate-phosphorus in runoff from a complex cover agricultural watershed in the Georgia Coastal Plain was found to be less than the rainfall input of these chemicals. Chloride load in runoff exceeded chloride load in rainfall. Measured stream flow nutrient losses showed a range in nitrate-plus nitrite-nitrogen load from 0.14 to 0.24 kg/ha/yr and a range in the orthophosphate-phosphorus load from 0.141 to 0.144 kg/ha/yr.

SEDIMENT delivery ratios (SDR) were computed for a 1736 ha, mixed land use, agricultural Coastal Plain watershed, based on observed sediment yield and com-puted gross erosion by the USLE method and an experi-mentally determined soil erodibility factor. Seasonal variation in SDR values was observed. This variation was not adequately explained by seasonal variation in rainfall and/or runoff. Relationships for predicting SDR based on season and climatic variables were developed for the study area. Season was found to be the predominate in-dividual predictor of observed variations in SDR values.

Adequate information on the seasonal distribution of rainfall erosivity required to predict rainfall erosion losses was not available previously for peninsular Florida. In this area, the oceanic and continental climatic influence causes significant seasonal and year-to-year differences in precipitation amount and distribution. Thunderstorms may occur at any time of the year, resulting in high rainfall energy and potentially producing considerable soil detachment by raindrop impact in the absence of adequate plant cover. Daily precipitation data from selected locations in Florida were analyzed to determine seasonal variation in the rainfall erosion index (EI). Long-term seasonal rainfall EI distributions were used to delineate EI-distribution regions, thereby extending applicability of the Universal Soil Loss Equation for erosion modeling in peninsular Florida. Design values for annual total and maximum 1-day EI, and a revised map of average annual rainfall erosivity are presented, also.