Rural Clean Water Program Publications
National Rural Clean Water Program Symposium Proceedings
EPA. 1992. Seminar Publication: Proceedings of the National Rural Clean Water Program Symposium - 10 Years of Controlling Nonpoint Source Pollution: The RCWP Experience, U. S. Environmental Protection Agency, Cincinnati, Ohio, EPA/625/R-92/006, 400p.The proceedings of the National Rural Clean Water Program Symposium held in Orlando, Florida, in September, 1992, are now available. General topics include water quality and land treatment monitoring; relating water quality to land treatment; land treatment and operation and maintenance of BOPS; project coordination and farmer participation; institutional arrangements, program administration and project spin-offs; information and education; socioeconomics, technology transfer, and lessons learned; and research needs and future vision.
Copies of the proceedings are available (free) from USEPA, CERI, Document Distribution
(G-72), 26 Martin Luther King Dr., Cincinnati, OH 45268.
Rural Clean Water Program Evaluation Summary
Gale, J. A., D. E. Line, D. L. Osmond, S. W. Coffey, J. Spooner, and J. A. Arnold. 1992. Summary Report: Evaluation of the Experimental Rural Clean Water Program. National Water Quality Evaluation Project, NCSU Water Quality Group, Biological and Agricultural Engineering Department, North Carolina State University, Raleigh, NC. 38p.The Summary Report: Evaluation of the Experimental Rural Clean Water Program is based on-site evaluations of the 21 Rural Clean Water Program (RCWP) projects, a short answer questionnaire completed by project personnel, project ten-year and annual reports, and technical assistance provided by NWQEP to the RCWP projects. The report includes a summary of lessons learned from the RCWP divided into two sections focusing on 1) program or national level topics and 2) project or local level topics. The lessons are meant to provide guidance for decision-makers at the national, state, and project levels in developing, implementing, and evaluating agricultural NPS pollution control programs and projects. Learning from past experiences and building on successes is vital to the development of effective and efficient NPS pollution control programs. A brief synopsis of each of the RCWP projects is presented.
Copies of the report (WQ-75) may be ordered from Mrs. Janet Young, NCSU Water Quality Group, 615 Oberlin Road, Suite 100, Raleigh NC 27605-1126. Copies are currently being distributed free; when the present supply is exhausted, the report will be reprinted and available at cost. A more comprehensive report on the RCWP evaluation will be published in 1993.
TECHNICAL NOTES
Over the past decade, North Carolina's Albemarle and Pamlico Sounds have experienced increasing water quality problems ranging from fish kills to declining populations of aquatic vegetation. In response to these and other problems, the Albemarle-Pamlico Estuarine Study (A/P Study) was initiated in 1987 to characterize the watersheds draining into the sounds and to identify potential management strategies to address and improve water quality. The A/P Study is a five-year cooperative effort by the North Carolina Department of Environment, Health and Natural Resources and the U.S. Environmental Protection Agency to prepare a comprehensive conservation management plan for the Albemarle-Pamlico region. The plan is intended to recommend priority corrective actions and compliance schedules addressing point and nonpoint sources of pollution to restore the chemical, physical, and biological integrity of the estuary, the wildlife of the estuary, and the production levels of recreational and commercial fisheries of the estuary. A first draft Comprehensive Conservation Management Plan was released for public review in July, 1992.
The following article describes the first of a series of projects performed by researchers at the Research Triangle Institute. The work is focused on assisting resource managers with the state Department of Environment, Health and Natural Resources in understanding the sources of nutrients entering the Albemarle-Pamlico estuarine system and their magnitude and in developing an integrated watershed management approach to reduce nutrient inputs to the estuaries. Except where otherwise noted, the information used in this article is drawn from the 1992 report entitled Watershed Planning in the Albemarle-Pamlico Estuarine System: Report I - Annual Average Nutrient Budgets (1).
North Carolina's Albemarle-Pamlico (A/P) system is the second largest estuarine complex in
North America. With a total watershed area of 30,880 square miles, the A/P watershed (Figure
1) is home to nearly 2 million permanent residents. In addition to the systems many recreational
values, it is a significant nursery area for Atlantic coast fisheries (2). However, as a result of
agricultural, industrial, and municipal loadings, the A/P system is experiencing water quality
problems ranging from nuisance algae blooms to outbreaks of fish diseases.
In 1987, the A/P system was designated as an estuary of national significance and was selected
for study as part of USEPA's National Estuary Program. The resulting Albemarle-Pamlico
Estuarine Study was initiated as a cooperative program between USEPA and the North Carolina
Department of Environment, Health and Natural Resources. The purpose of the A/P Study is to
evaluate the nature of the basins' environmental problems and to determine how the estuaries can
best be preserved and managed. (3)
Nutrients have been identified as key factors in the health of the Albemarle-Pamlico estuarine
system. Nitrogen and phosphorus are known to be primary causes of algal blooms and other
problems, such as decreased dissolved oxygen levels. Government agencies, water users, and
concerned citizens all need to know where nutrients found in the estuaries originate. This
information provides valuable information for use in setting priorities for management strategies
and, in the longer-term, is crucial to evaluating the effectiveness of management effort.
As a first step toward controlling nutrient loadings, a screening approach designed to narrow the
focus of future nonpoint source (NPS) control efforts has been developed (phase 1). The
approach consists of developing nutrient budgets within the 68 subbasins comprising the A/P
study area - These budgets can then serve as a tool for screening out "critical" areas of high
nutrient loading. (3)
The information gained from this process would allow the state to target BMP cost share monies
to the areas causing the greatest nutrient loadings. In addition, these areas could be identified for
the development of Total Maximum Daily Loads (TMDLs). (The TMDL process was
established by section 303 of the Clean Water Act. A TMDL is a tool for implementing state
water quality standards and is based on the relationship between pollution sources and in-stream
water quality conditions. The TMDL establishes the allowable loadings for a water body, thereby
providing the basis for states to establish water quality-based controls.)
Further efforts of the ongoing study include:
For nonpoint sources, an export coefficient approach was used. Because key data sources
included land use classified from LANDSAT images in 1987-88, the temporal resolution was
similar to that of the point source estimates. In recognition of the considerable uncertainty in
choosing coefficients appropriate for particular land cover categories, coefficients representing
"high" "most likely," and "low" estimates were chosen after an extensive literature review.
In 1991, the Research Triangle Institute (RTI), the North Carolina Center for Geographic
Information and Analysis (NC CGIA), and NCDEM completed a watershed delineation project
in the A/P Study Area, with watersheds determined by the topography and hydrology shown on
1: 24, 000 maps. Watershed outlet points were developed using three criteria. First, all
boundaries were consistent with USGS cataloging units (CUs), as shown on a 1: 500, 000 scale
base map. Second, subbasins were delineated based on a revised version of 25 subbasins defined
by NCDEM in the 1970s. This reflects an effort to ensure consistency between NCDEM and
USGS CU boundaries. Finally, watersheds were identified with outlets at the locations of gaging
stations with continuous flow recorders and good water quality records. This,scheme resulted in
identification of 68 polygons in the North Carolina portion of the A/P Study Area.
In Virginia, the Soil Conservation Service, in conjunction with several state agencies, completed
an extensive project to delineate and digitize a statewide watershed network in 1990. These
watersheds cover between 30,000 and 60,000 acres (4). USGS CUs are also available for the
Virginia portion of the A/P Study Area. Watersheds corresponding to gaging stations have not
been identified in Virginia.
Land Use Cover
Land use/land cover data from a 1987-88 LANDSAT classification study were obtained from NC
CGIA. This study resulted in six general (Level 1) and 20 detailed (Level 2) classifications. The
20 classifications were divided into four categories for purposes of calculating runoff and atmos-
pheric inputs: agricultural, forest/wetland, developed, and marsh/water.
Information on the extent and location of urban lands from the satellite classification was used,
with the knowledge that this classification was the least accurate. Because the study area is
largely rural, tills issue is probably not important for area-wide or basin-wide projections,
although land use data for small watersheds that are relatively highly developed may be less
representative of actual conditions.
Export Coefficient Approach
Agricultural Land, Forest/Wetlands, and Developed Land..... Nutrient export coefficients,
reflecting average nutrient loading (kg/ha*yr), were used to evaluate nutrient contributions.
Export coefficients were chosen for four categories of land cover. For any type of land use, the
nutrient export coefficient was multiplied by the total area of that use in the watershed to obtain
a loading estimate. These products were summed to obtain an estimate of total runoff nutrient
loading into surface waters in the study area.
Export coefficients were chosen based on a literature review (summarized in Table 1). No
attempt was made to choose different export coefficients for a given category for different
hydrologic units based on hydrologic or other distinctions. The interquartile range of export
coefficients reported was used as a measure of the variability in areal loading rates. This
approach was chosen because its inherent simplicity is appropriate for predictions in the large
A/P Study Area and the data requirements of alternative nonpoint source modeling approaches
was prohibitive.
Direct Atmospheric Loading .....Atmospheric phosphorus inputs were estimated using
literature values of export coefficients as well. In addition, data generated by EPA's Regional
Atmospheric Deposition Model (RADM) were used to estimate nitrogen loading. RADM was
calibrated using monitoring data from 1985. Latitude and longitude coordinates for point
estimates were obtained and a map was created using ARC/INFO (a geographic information sys-
tem). This point map was contoured using the TIN module of ARC/INFO to produce contour
maps for wet and dry deposition. These maps were overlain on the subbasin map, and areally
weighted averages for each deposition type were calculated. Wet nitrogen deposition estimates
were available from RADM for nitrate only. Dry nitrogen deposition estimates were available
for nitrate, nitric acid, and nitrite. Therefore, the RADM estimates do not include dry deposition
of ammonium, wet or dry organic deposition, or total wet nitrogen oxides.
Upstream Reservoir Discharge .....For the Falls Reservoir (NC), previous estimates of
nutrient loading at the gage below the lake were used. Therefore, estimates presented for the
upper Neuse River Basin do not explicitly include point sources, runoff, or atmospheric inputs.
For Lake Gaston (NC/VA line), USGS records of the long-term average nitrogen and phosphorus
concentration and flow at Roanoke Rapids were used to estimate flux.
SAS, EXCEL, and ARC/INFO were used to manage, analyze, and present data. Results are
available as EXCEL spreadsheets. '
As with any such study, the effects of errors in estimates or results must be considered.
Unfortunately, an analysis to determine quantitative estimates of error was beyond the scope of
the study. However, the following comments are offered for the reader who is interested in a
quantitative assessment of error.
Dillaha (9) suggests that, regardless of whether a lumped (as in this study) or distributed model
is used, nonpoint source models are accurate only within a factor of two or three. A wide range
of export coefficients have been reported in the literature, often without detailed information
about important covariates, such as soil type or runoff, making it difficult to choose an
appropriate export coefficient for any given land cover or land parcel.
There is an acknowledged source of bias in that point source inputs are underestimated. This bias
is a function of the decision not to estimate loading for unmonitored point sources (not thought
to be a large source of bias for North Carolina; possibly larger in Virginia).
Another, perhaps more important, factor that may result in the overestimation of runoff inputs,
in both relative and absolute terms, is the inability of export coefficients to account for loss of
nutrients (to biomass, sediments, and the atmosphere) in the process of being delivered to the
estuaries. Because export coefficients reported in the literature typically represent field-scale or
small-watershed-scale delivery, it is conceivable that additional nutrient loss in the large stream
and river corridors occurs that is not represented by the coefficients.
One ramification of these biases is the possibility not only of attributing a sufficient portion of
the budgets to these source categories, but of overestimating the relative contributions of other
categories.
The usefulness of the models for management purposes is conditioned by tills uncertainty and
variability. Lack of knowledge of model uncertainty makes it difficult for policy makers to
anticipate the effect of a policy intervention such as revising a county land use plan or
reallocating cost share funds. The approach for dealing with uncertainty for the largest source
category (runoff) chosen for this study, professional judgment regarding " high " "most likely,"
and "low" coefficients, is a first cut at addressing this issue. Although the method is riot
statistically rigorous, it does attempt to present useful information concerning plausible model
error.
Two other important issues must be considered that are not reflected in these budgets:
temporal variability and bioavailability. Numerous studies have demonstrated that there is
considerable year-to-year, seasonal, and storm-event-related variability in nutrient loading,
which has important implications for the ecological health of the estuarine system. The
ultimate effect of annualizing inputs is to develop a long-term context in which to consider
them.
There is also evidence that nutrients from the different source categories considered (point
sources, atmospheric deposition, runoff) may be used by primary producers to varying
degrees depending on the chemical form in which the nutrient is delivered. An estimate of the
relative "weight" of a unit mass nutrient loading to the primary producers from each source
category was not provided in this study.
Management issues associated with sources for which widely available data do not exist (such
as confined animal operations, local atmospheric sources, and septic tanks) may be quite
important on local as well as regional scales.
An important product of this study is the development of a watershed-oriented data base, which
includes estimates of land cover and nutrient loadings for 68 and 44 hydrologic units in North
Carolina and Virginia, respectively. Enhancement of the data base will require continuing
interagency coordination to incorporate attributes such as additional watershed boundaries, more
detailed land use/land cover data, and best management practices data. This effort will be critical
to monitoring the effectiveness of nutrient control programs.
The identification of land use and nutrient budgets within these hydrologic units may provide
useful input into geographic targeting efforts for nonpoint source management programs.
Although these are certainly not the only factors to consider in setting geographic priorities for
management efforts, they are nevertheless a key indicator of nonpoint source pollution.
2. Steel, J. (ed). 1991. Albemarle-Pamlico Estuarine System, Technical Analysis of Status and
Trends. Albemarle-Pamlico Estuarine Study Report No. 90-01. NC Dept of Environment, Health
and Natural Res., Raleigh, NC.
3. Tippett J. P. July 1992 Draft. TMDL Case Study. Basin-wide Nutrient Screening for the
Albemarle-Pamlico Estuarine Study. Office of Wetlands, Oceans, and Watersheds, U.S.
Environmental Protection Agency.
4. Soil Conservation Service. 199 1. Virginia Hydrologic Unit Atlas. U.S. Department of
Agriculture, Richmond, VA.
5. North Carolina Division of Environmental Management 1990. Chowan River Water Quality
Management Plan - 1990 Update. Division of Environmental Management, North Carolina Dept
of Environment, Health, and Natural Res., Raleigh, NC.
6. North Carolina Division of Environmental Management 1988. A Report of the Proceedings
to Reclassify the Neuse River as Nutrient Sensitive Waters. Division of Environmental
Management, North Carolina Dept of Environment, Health, and Natural Resources, Raleigh,
NC.
7. North Carolina Division of Environmental Management 1997. Surface Water Quality
Concerns in the Tar Pamlico River Basin. Division of Environmental Management, North
Carolina Dept of Environment, Health, and Natural Resources, Raleigh, NC.
8. North Carolina Division of Environmental Management. 1991. An Evaluation of the Effects
of the North Carolina Phosphate Detergent Ban. Report No. 91-04. Division of Environmental
Management North Carolina Dept. of Environment, Health and Natural Resources, Raleigh, NC.
9. Dillaha, T. A. 1990. Role of best management practices in restoring the health of the
Chesapeake Bay system, In: Haire, M. and E. C. Krome (eds). Perspectives of the Chesapeake
Bay, 1990, U. U. Environmental Protection Agency.
Spooner, J., R.L. Huffman, D.E. Line, J.A. Gale, G.D. Jennings, S.W. Coffey, J.A. Arnold,
L. Wyatt, D.L. Osmond, S.K. Haseeb. 1992. Nonpoint sources,Water Environment Research
64(4):503-514.
The 1991 annual literature review of nonpoint source prepared by the NCSU Water Quality
Group and others includes 243 references. The review was published in the June, 1992, issue
of Water Environment Research (formerly the Journal of the Water Pollution Control
Federation). The review covers nonpoint source policy, economics, and planning; water quality
and water resources; best management practices for nonpoint source control; and. nonpoint
source pollution monitoring and modeling.
Free copies may be ordered from Publications Coordinator, NCSU Water Quality Group, 615 Oberlin
Rd., Suite 100, Raleigh NC 27605-1126. (Please refer to WQ-74.)
Randall C. Dodd, Research Triangle Institute
Gerard McMahon, Center for Urban and Regional Studies, University of North Carolina at
Chapel Hill
John P. Tippett, Research Triangle Institute
Steven Stichter, Research Triangle Institute
Judith A. Gale, NCSU Water Quality Group
Scope of the Study
Goals of the nutrient budget phase of the study were to:
These ongoing efforts will expand the study in order to:
Approach
The initial phase consisted of two components. For point sources, data in digital format from
1989-90 compliance monitoring records tracked by the NC Division of Environmental
Management (NCDEM) and the Virginia Water Control Board (VWCB) were collected and
analyzed.Methods
Data used to calculate current (1989-90) inputs were
obtained from NCDEM and VWCB and included estimates of monthly median flow and
concentration where available.
Watershed Delineation
Table 1: Export Coefficients Literature Review (1)
(kg/ha*yr)
Land Use/Land Cover Categories
Agric Forest/ Devel Atmos
Wetland
Total Phosphorus
Low (25%) 0.55 0.09 0.45 0.25
Median 0.99 0.13 1.06 0.65
High (75%) 2.03 0.21 1.5 0.69
Total Nitrogen
Low (25%) 5 0.69 5 8.7
Median 9.8 2.33 7.5 12.4
High (75%) 14.3 3.8 9.72 24
No. of Studies 77 36 78 6
Results
Point source estimates for North Carolina (data are very limited for Virginia dischargers) are
summarized in Figure 2. Although no area-wide point source data collection efforts were
conducted prior to the RTI study, it is clear in comparing these estimates to previous studies by
NCDEM (5,6,7,8), that point source phosphorus loading declined substantially in the 1980s,
while point source nitrogen loading changed very little.
Estimated nonpoint source nitrogen and phosphorus contributions are summarized in Figure 3.
Agricultural inputs dominate runoff contributions; developed lands contribute a very small
proportion of the total runoff inputs. Exceptions to this generalization certainly occur locally as
a result of local land use patterns and deviations from the export, or yield, represented by the
coefficients selected (due to both natural features and management factors).
Discussion
The "screening" nature of the study and the goal to develop area-wide budgets required use of
data coverages available for the entire area. Relying on state point source, LANDSAT, and
RADM data and literature estimates of export coefficients allowed for assessments of the relative
magnitude of sources and source areas on large temporal (annual) and spatial (regional) scales.
It is apparent based on this analysis that long-term
management of nutrient inputs to the estuaries and sounds of the Albemarle-Pamlico system will
be challenging. As a function of the landscape, most of the area-wide loading is generated from
agricultural land, with atmospheric sources of nitrogen and point source phosphorus inputs being
the next largest source categories. Principal challenges in managing agricultural loading include:
targeting key source areas and sources; tracking management efforts; and quantifying the
effectiveness of management efforts, on the field, watershed, and basin scale.References
1. Dodd, R.C., G. McMahon and S. Stichter. 1992. Watershed planning in the
Albemarle-Pamlico Estuarine System, Report I - Annual Average Nutrient Budgets. Report No.
92-10. Albemarle-Pamlico Estuarine .Study, NC Dept of Environment, Health and Natural
Resources, Raleigh, NC.For Further Information
Copies of the report on which this article is based (1) are available from the Albemarle-Pamlico
Estuarine Study, NC Department of Environment, Health, and Natural Resources, P. 0. Box
27687, Raleigh, NC 27611.
INFORMATION