Estimation of Runoff and nonpoint source pollution using GIS

Abedalrazq F. Khalil

CEE 6440:  GIS in Water Resources

Department of Civil and Environmental Engineering

Utah State University

Fall 2002

ABSTRACT

 "Comprehensive watershed management requires the determination of both point and non-point sources of pollution within a watershed" (Said, 2001). The concerns of non-point sources pollutants (NPS) are mainly regarding nitrogen and phosphorous which most often cited as causing water quality impairment. Pollutants might resulted from the following sources: a) soil erosion, b) surface water runoff entering streams or groundwater, c) agricultural activities and wastewater treatment plants which are the principal sources of nitrogen in surface waters (said et. al.2001) d.) Human or animal wastes, fertilizers, and plants, which chiefly produce Phosphorus. In this class term paper the focus is to calculate the runoff from a basin and then to conduct some calculation using runoff water volume and the expected mean concentration (EMC) to provide the both annual loading of both Phosphorus and Nitrogen, and a visual display of the interaction between these loadings and the land use.

INTRODUCTION

In this study, the GIS technology used, basically to calculate the average annual runoff for a land use configuration based on precipitation layers. The runoff values are calculated using the Curve Number (CN) method. With climate record of the area, with the annual runoff depth, it is possible to calculate runoff volume. A simple calculation using volume and the expected mean concentration (EMC) will provide the annual loading of chemicals. The expected mean concentration (EMC) of nitrogen and phosphorus depend on the each land use. High values are expected in agricultural systems, followed by residential, commercial, grass/pasture, and forest. (Nerilde Favaretto, 2000).

Figure 1: Methodology used in the project.

Table 1: Shows Expected mean concentration (EMC) of Total N, Total P in each land use class*.

*Source: http://pasture.ecn.purdue.edu/~engelb/agen526/projects00/favarett/LTHIAproject.htm

AREA of STUDY

This project deploys the GIS layers used by Said A. et. al. 2001 for Snake River Basin in Idaho, in there study of the relationship of land-use to total Nitrogen/Phosphorus in streams, while the focus in this paper is just related to landuse.

OBJECTIVES

The objectives of this study are to estimate the total phosphorus and the total nitrogen concentrations and the to display a visual relationships between these parameters and the land-use.

            In addition, this study provide visual estimate the relationships between total phosphorus and total nitrogen concentrations and the relationships between these parameters and the land-use.

METHODOLOGY

Most of the papers estimate nutrient loading on a large basin, using the Geographic Information System (GIS) version of the Long Term Hydrologic Impact Assessment and Non Point Source Pollutant (L-THIA NPS) model which  can be downloaded from(http://danpatch.ecn.purdue.edu/~sprawl/LTHIA). What used in this study is something close the concept of LTHIA but different in some sense.

Curve Number

                Direct average runoff volume Q is predicted for average daily rainfall by using the SCS curve number equation 1 (U.S. Department of Agriculture, Soil Conservation Service 1972):

                                       (1)

where Q is the daily runoff, P is the daily rainfall that can be estimated from the long term precipitation or GRID GIS theme, and S is potential infiltration. S changed among watersheds because soils, land use, and slope and with time because of changes in soil water content. The parameter S is related to a soil type curve number (CN) by the SCS equation 2 (U.S. Department of Agriculture, Soil Conservation Service 1972):

                                            (2)

CN is a function of land use, surface condition, and soil group.

Expected Mean Concentration

The expected mean concentration (EMC) of nitrogen and phosphorus depends on the land use. High values are expected in agricultural areas, followed by residential, commercial, grass/pasture, and forest. Table 1 shows the EMC for TN and TP used by the model for the calculations in this study (Nerilde Favaretto, 2000).

Steps:

• The first step in this project was to locate the necessary data sets. Some data downloaded from “Idaho department of Water Resources”  and other data set made available to this term paper by Said A. from his research on this area. The projection is  North American Datum of 1927Clarke 1866 Projection.

Figure 2: Map shows the chosen HUCs of Idaho State HUCs.

Figure 3: Shows the counties of the chosen area.

Figure 4: Shows the river of the chosen area.

• Once the data layers were added to the GIS, some tabular manipulation was necessary in order to prepare it for use with the ArcGIS, e.g. the landuse classification tables, hydrologic soil group data.

Figure 5: Map Shows Landuse distribution.

Figure 6: curve number distribution.

• Next, a table containing the hydrologic soil group data was joined with the table of the soils theme.

• Curve number grids were generated using the land use and soil grids. The same soil grid was used for each curve number grid.

• With the precipitation data obtained earlier from the web site. Runoff depth grids were created with the curve number grids.

 Figure 7: Map shows runoff depth distribution in ‘cm’.

Table 2: Rainfall statistics in (cm)

• Union the shape number and the hydrologic soil group maps with the map of antecedent runoff condition (ARC), include the soil data theme where soils are classified into four HSG’s (A, B, C, and D) according to their minimum infiltration rate, which is obtained for bare soil after prolonged wetting.

• Runoff volume grids were calculated using the runoff depth grids, and land use/soil index, and the curve number using this equation:

Figure 8: Map shows Runoff Volume in m³.

The estimated average annual runoff depth and volumes are greatest in the cities. This is because of the impervious nature of an urban landscape due to concrete, pavement, and density of buildings. The areas with lowest runoff depth are located near the rivers and streams. These are the lowest elevation areas and are also the outlets of the watershed.  The runoff depth in the agricultural areas is medium to low. 

• The expected mean concentration (EMC) of nitrogen and phosphorus which depend on the each land use are entered to tables.

 Figure 9: Expected mean concentration for both N and P.  P to N ratio normally doesn’t exceed 0.0625.

• Finally, NPS pollutant loading grids were generated via the runoff volume grids. Union the Landuse them and the Runoff Volume theme; multiply Runoff volume times the expected values of P and N for each land use.

Figure 10: Shows the Nitrogen loads in ( Kg/ha/year)

Table 3 : Statistics of ‘N’ flow in the river in mg in Total Runoff volume:

Figure 11: Shows the phosphorus loads in ( Kg/ha/year)

Table 4: Statistics of ‘P’ flow in the river in mg in Total Runoff volume:

It is shown clearly on these maps that the areas of greatest risk are the commercial, industrial, and the high-density residential areas. Phosphorus is often more problematic in an urban setting.

• The final step was preparation of the data for presentation.

• Map layouts were prepared and exported as bitmaps for presentation. 

CONCLUSIONS

The previous calculated layouts show:

• The higher levels of nitrogen and phosphorus tend to occur in the agricultural areas where they are used heavily as fertilizer inputs. 
• The runoff depths values are largest in the city due to the large amounts of concrete, pavement, and buildings. 
• Runoff depth was generally lower in the agricultural areas and virtually non-existent near the creeks and streams because they have the lowest elevation and is the outlets for the watershed.

I feel that this was a nice exercise in using GIS. These types of results can be very useful to raise awareness of the potential long-term effects of land use changes within a watershed. This can facilitate better planning for the future in order to limit the extent of possible harmful effects due to pollution and runoff. City planners and engineers could use this data as a tool to aid in the design of future urban development.

Acknowledgment

I would like to thank Ahmad Said for the extremely valuable inputs he has made to the work reported in this term paper and for his initial idea.  I want to thank also Ashraf Shagadan for his help and Dr. David Tarboton, course instructor, for his time in helping me to learn.

References

 1.)Ahmed Said, Stevens, D. K. and Sehlke G. The Relationship of Land-Use to Total Nitrogen/Phosphorus in Streams, 2001 AWRA Annual Water Resources Conference; Albuquerque, New Mexico; Nov 12-15, 2001

2.) Nerilde Favaretto, 2000. Nitrogen And Phosphorus Losses In Surface Water From The Indian Pine Watershed, ABE 526- Watershed System Design, Final Project, May 2000, http://danpatch.ecn.purdue.edu/~sprawl/LTHIA7/lthia/lthia_index.htm.

3.) http://pasture.ecn.purdue.edu/~engelb/agen526/projects00/favarett/LTHIAproject.htm