Exercise on Agrichemical Transport in the Midwest Rivers

Prepared by Pawel J. Mizgalewicz and David R. Maidment

Center for Research in Water Resources, University of Texas at Austin


Table of Contents

The complete description of the methodology can be found in:
Mizgalewicz, P.J., and D.R. Maidment Modeling Agrichemical Transport in Midwest Rivers using Geographic Information Systems, CRWR Online Report 96-6, December 1996.


Goals of the Exercise


Computer and Data Requirements

This exercise is intended to be performed with a Geographic Information System (GIS) program called ArcView. You may use anycomputer running ArcView 3 or later, under Windows OS. An introduction to ArcView exercise is available on-line and on the GISHydro97 CD rom.

In addition to ArcView you will need some programs developed by our research group. The programs are packaged in something called a project which you can load into ArcView like a file.

The data consists of:

Introduction

Tha primary goal for this exercise is to demonstrate the application of GIS to analyze river pollution. The capabilities of this model will be presented using a herbicide, atrazine, and a nutrient, nitrate plus nitrite as nitrogen. The steps of the exercise are displayed on the light-blue background. Please, write me if you have any comments: pawel@crwr.utexas.edu

  • Determine nitrate plus nitrite as nitrogen concentrations in the Iowa-Cedar Rivers;
  • Draw the bar-charts of the agrichemical concentrations in selected sites along the Iowa river;
  • Discuss the results (magnitude, changes along the river, seasonal variations);
  • Assume nitrogen fertilizer use in 1990;
  • Assume flow conditions of 1990.

The ArcView application is divided into four projects (a project is a collection of associated documents: views, tables, layouts, charts, and scripts). Each project is designed to perform a different task. Four buttons that allow user to switch between projects have been placed in the Button Bar. These buttons have the following functions:

Runs the "model" project. This project is designed to prepare entry data for the agrichemical transport model as well as to calculate concentrations and loads.

Runs the "results" project which is designed to display the results as maps of agrichemical concentration and load.

Executes project "flwprc" which provides visualization of the flow record for a selected modeling unit.

Project "tools" contains the fundamental tools for hydrologic modeling.

The project "model" is the main component of the ArcView application. Projects "results", "flwprc", and "tools" are supporting modules.

Get the necessary data

  • Create a directory (command mkdir )
  • There are numerous ways you can get needed data:
    • download the whole directory agrichem from \ftp\pub\agritot\ (crwr server)
    • or download the file agrichem.zip (about 4 Mb) which is a compressed version of the agrichem directory
    • the file is also available from ftp.crwr.utexas.edu agrichem.zip
      use PKUNZIP or other software to uncompress the zip file. To preserve directory structure use option "d" with pkunzip.exe:
      pkunzip -d agrichem.zip
      You probably have pkunzip.exe somwhere on your computer. If not download it .
    • or copy the agrichem folder from the CD-ROM into the directory you have created. Make sure that the following folders: cruse, crwsd, and support, including all files inside them have the writing permission enabled, i.e. the files can be edited.
      NOTE: ONLY files and directories copied from the CD-ROM are write protected !
      The following error message is displayed when the program can not write to one of the files:

  • Create additional directory in which you will save your results (for example: d: /xresults)
  • Open project Model (file model.prj)

Project Model

When opened, the project Model displays the following maps:

units

Six buttons are designed to select different activities:

Edit Application. This button displays dialog boxes that allow the user to specify the agrichemical application rate. If the view-window Application Rate is active, the application in either all, or selected counties can be specified. If the view-window Modeling units is active, the application rate in each modeling unit can be modified.

Edit Cumulative Flow Rate. This button shows dialog boxes for selecting the flow record (by selecting a year) that will be used for calculations and for adjusting the selected values. For example, all data can be multiplied by a factor that represents extreme conditions. The adjustment may be performed for all months of the specified year or for each individual month.

Select Year. This button allows the user to specify a year for a model that has a trend component. There is no trend in the current model.

Select Model. The dialog box displayed by this button displays available models from which the user has to select an equation. This equation will be used to calculate agrichemical concentrations.

Run. Executes the script that performs the calculations.

First Order Reaction. Calculates concentrations and loads assuming exponential decrease of the agricultural constituent along the flow path. The current model does not contain a map of spatially distributed decay rates that allows one to estimate concentrations by this method.

Core database of the model

The following coverages and data files constitute the core database of the model:

Table 1. Polygon Attribute Table of the cruse coverage

Field (Item) Description
Fips county FIPS code
St state
Cntyname county name
N89kgkm2, N90kgkm2, N91kgkm2 estimated average nitrogen fertilizer application rate for the years 1989, 1990, and 1991, respectively [kg/km2]
A89kgkm2 estimated average atrazine application rate for 1989 [kg/km2]
Temp field reserved for storage of results of partial calculations
Use application rate selected for the estimation of concentrations and loads

Table 2. Polygon Attribute Table of the crwsd coverage

Field (Item) Description
Unit_id ID of the modeling unit
Gswsh ID of the USGS gauging station downstream of the modeling unit
Unit_nx ID of the next unit on the flow path (downstream unit ID)
Order a number that specifies location of the unit on the flow path
Areakm2 area of the unit [km2]
Careakm2 drainage area upstream of the unit outlet [km2]
Alndslp average land slope of unit drainage area
Alndlgkm average length on the flow path from the land to the stream network [km]
Chemuse Agrichemical application rate in the unit [kg/yr/km2]
Cchemuse average agrichemical application rate over the unit drainage area [kg/yr/km2]
Travtime travel time across the unit [d] (no data available)
Losscoef Overall loss coefficient in the unit [1/d] (no data available)
Expofact Export factor, fraction of the applied agricultural chemical in the unit that enters the surface water (no data available)
Tcavg Annual average temperature for upstream drainage area [°C]
Pmmavg Annual precipitation depth, average over upstream drainage area [mm]
Flow01...Flow12 monthly flow rate that flows through the unit outlet [m3/s]
Conc01...Conc12 concentration at the unit outlet (atrazine [mg/m3], nitrate plus nitrite as nitrogen [g/m3])
Load01...Load12 chemical load (flow * concentration, atrazine [mg/s], nitrate plus nitrite as nitrogen [g/s])

Table 3. Fields of the file model2a.dbf -- model specification.

Field (Item) Description
Year a year that is used to estimate the trend coefficient;
Ftrend mathematical description of the trend. Current version of the model assumes no trend, i.e. the trend is described by the following equation: 1.00+(0.0000*Year)
Trendcf trend coefficient which is calculated according to the information stored in items Year and Ftrend
Si01...Si12 seasonal index, 12 values that represent the monthly variations of the agrichemical concentration around the annual average.
Model model name, e.g. Nitrate01, Atrazine01, Atrazine02
Sel contains 1 if the model is selected for calculations, 0 otherwise
Equation equation for estimation of agrichemical concentrations. Four models are available,
  • atrazine model without the flow rate [mg/L]:
    (-1.5575+(0.026*U)+(0.7998*LL)+(0.4559*TA)-(0.0048*PA))*TR*SI
  • atrazine model with the flow rate [mg/L]:
    (-0.8142+(0.0133*U)+(0.3346*LL)+(38.8804*LS)+(0.2732*TA)
    -(0.0029*PA))*(Q^0.2899)*TR*SI
  • nitrate plus nitrite as nitrogen model without the flow rate [mg/L]:
    (-7.424541+(0.001062*U)-(1.033063*TA)+(0.019339*PA))*TR*SI
  • nitrate plus nitrite as nitrogen model with the flow rate [mg/L]:
    (-7.57848+(0.00064886*U)+(0.776683*LL)+(173.6409*LS)-(0.520245*TA)
    +(0.0088545*PA))*(Q^0.3432)*TR*SI

Three multiplicative models are included only for testing purposes:

  • nitrate plus nitrite as nitrogen [g/m3]:
    0.000143208*(U^1.3814)*(A^(-0.5556))*(Q^0.4240)*(LL^0.9522)*TR*SI
  • for atrazine [g/m3]:
    0.0001265546*(U^0.8323)*(SL^0.3591)*(Q^0.0940)*(LL^0.9208)*TR*SI
  • for atrazine, same as (x2) but different output units [mg/m3]
    0.1265546*(U^0.8323)*(SL^0.3591)*(Q^0.0940)*(LL^0.9208)*TR*SI

where: U = application rate, A = drainage area, Q = flow rate, LL = average overland flow length, SL = land slope, TA = average temperature, PA = average precipitation, TR = trend coefficient, and SI = seasonal index.

Comments comments and model description

The ArcView script equat6 executed from the button calculates concentrations and loads. This script first selects the record that contains the value of item [Sel] equal to 1. Then equat6 extracts the value from the item [Year], retrieves the equation that describes the trend from the item [Ftrend], calculates the trend coefficient, and stores it in the item [Trendcf]. In the next step, the script extracts the equation from the item [Equation] and replaces the symbolic names U, A, LS, LL, and Q with the corresponding item names [Cchemuse], [Careakm2], [Alndslp], [Alndlgkm], and one of the item [Qm01]...[Qm12]. The symbolic name TR is replaced by the value stored in the item [Trendcf] and the name SI is replaced by the value from one of the items [Si01]...[Si12]. The concentrations are calculated for each month of the year by the Avenue request calculate and the results are stored in the items [Conc01]..[Conc12]. The loads are estimated by multiplication of the flows and concentrations. The products are stored in the items [Load01]...[Load12].

Calculations of the agrichemical concentration in surface waters, according to the first order process, are performed by the script decay1. This Avenue program was written for future extensions of the agrichemical transport model, for example, to utilize the results of the PRIZM model or results of the CEEPES (Comprehensive Environmental Economic Policy Evaluation System) modeling program developed by Iowa State University's Center for Agricultural and Rural Development, CARD (Bouzaher and Monale, 1993).

Specification of the agrichemical application rate

Two spatial resolutions can be used to specify the agrichemical application rate: 1) by county; and 2) by modeling unit. If the view Application Rate is active the application by county may be specified. If the view Modeling Units is active, then the agrichemical application for each unit can be set.

Seting the agrichemical application rate for counties

The agrichemical application rate in either all, or selected counties can be specified. The following dialog box is opened after the button is clicked and the view Application Rate is active:

The application rate of nitrogen fertilizers estimated for 1989, 1990, and 1991, and application rate of atrazine estimated for 1989 can be selected. A multiplier that increases/decreases application rate by a given percentage also can be used to simulate the different policy scenarios. The third input field allows one to enter the value of the application rate directly.

Advanced: The database of the agrichemical application can be extended by adding an item to the polygon attribute table PAT of the cruse coverage (Table 1).

  • Make sure that the Application Rate view is active (just click the mouse on the title bar)
  • Click button , it is located on the button bar.
  • Enter the name of the databases that you want to use: n90kgkm2

The ArcView model checks the data entered by user. For example, if wrong input is detected in the agrichemical application dialog box, the user will be informed what error and in which entry field the error occurred:

Note: Setting the application rate for the counties change values in modeling units regardless some modeling units are selected of not.

Seting the agrichemical application rate for modeling units

The agrichemical application rate in either all, or selected unit watersheds can be multiplied by a factor or an application value can be specified. The following dialog box is opened after the button is clicked and the view Modeling Units is active:

Note: Since the average upstream application rate is calculates for each unit, the changes made to one modeling unit affect the values of all downstream units.

Specification of the monthly discharge rate

The historical monthly discharge can be extracted from the flow database (polygon attribute table of the unflow coverage) by specifying the year and assigned to either all or selected modeling units.  Also, the flow rate values can be multiplied by a factor to represent extreme flow conditions. The multiplication coefficients can be applied to all months of the year or to each individual month. The following dialog box is opened after the button is clicked and the view Modeling Units is active:

  • Click button ;
  • In the first input field "Assume the flow conditions of:" enter the year 1990. Do not change other fields.

Specification of the year

The button allows one to specify the year for a model that has a trend component. The specification of the trend equation is stored in the field Ftrend, file model2a.dbf. Current version of the agrichemical transport models assume no trend, i.e. the trend is described by the following equation: 1.00+(0.0000*[Year]). The following dialog box is opened after the button is clicked:

Selecting a model

The ArcView model allows to select an equation that describe relation between atrazine concentration or nitrate plus nitrite as nitrogen concentration and the variables such as agrichemical application rate, land slope, flow rate, average temperature and average precipitation. These equations are stored in the field [equation] of the file model2a.dbf (Table 3). The following dialog box is opened after the button is clicked:

  • Click button ;
  • Select the name of model that you want to use: nitrate nq1 mg/L (model for nitrate plus nitrite as nitrogen concentration, discharge is not used to make concentration prediction, output is in mg/L units)

Calculating concentrations and loads

The button starts the script that performs the calculations of chemical concentrations and loads. A YES/NO dialog box which ask for model confirmation is displayed:

Next, a dialog box is displayed in which user can select months of the year for which the calculations will be performed:

If user selects "Cancel" the calculations of the concentrations and loads are stopped. No changes to the attribute tables are made.

  • Click button ;
  • Accept the model nitrate nq1 mg/L ;
  • Perform calculations for all months of the year (type "1" in the entry field "All months" of the Recalculate dialog box;
  • Convert the theme Units into a shapefile: Make sure that the view "Modeling units" and the theme "Units" are active. Select from menu: Theme and Convert to shapefile. Name the shapefile n1990.shp and save it in the folder you created earlier (d:/xresults/). Do not add shapefile to the view;
  • Start project results.apr (click button );
    Do not save changes to model.apr unless you really want to.

Exponential decay model

The exponential decay model estimates loads in rivers assuming the chemical losses in rivers are governed by a first order reaction, i.e. the agrichemical mass exponentially decays as it travels from one modeling unit to the next downstream unit. After the button is clicked a series of dialog boxes are displayed in which user is asked to select input data and model parameters.

First the user is asked to select a field which contains export factors. The export factors represent all chemical losses after the application on the field and just before the chemical enters a stream.

Next, the user is asked to select a field with the cumulative runoff (river discharge):

The following dialog box asks for a name or the field in which the results (concentrations) will be stored:

If the field exists, the values will be overwritten otherwise the program asks user if she/he wants to create a new field. If the answer is YES the user will be prompted to specify a width and a precision of the new field.

The travel time (or travel distance) and the time loss coefficient (or distance loss coefficient) are stored in items [Travtime] and [Losscoeff] respectively.

Project Results

The project Results can be opened by clicking the button . This project reserves the space for preparing the maps of estimated agrichemical concentration and load in surface water. Three scripts, executed from the button bar, buttons: , , and , draw the bar charts of the monthly average chemical concentration, chemical load and flow rate, respectively. These scripts are based on ESRI examples.

The legend's title can be entered in the following dialog box:

Four sizes of the charts are available: Small, Medium, Large, and X-Large.

Similar dialog boxes are displayed before the charts of loads and discharges are drawn:

  • Create a new view;
  • Add the shape file n1990.shp to the view;
  • Add the theme RF1 (maps of rivers) to the view, so you can determine which units ale located along the Iowa River;
  • Edit the legend if necessary (i.e. line thicknes, line color, polygon foreground color, and polygon outline color);
  • Select units along the Iowa River (Make sure that the theme N1990.shp is active. To make a multiple selection the shift key must be pressed);
  • Click the button;
  • Enter the title of the legend (e.g. Conc. N1990 g/m3);
  • Select the size of bar small;
  • Print the results. Example map of concentrations in the Iowa River is shown below.
  • Use different method(s) to present results. Show discharge and loads.
  • Start project flwprc.apr (click button ).

Example of presentation of temporal and spatial distribution of the nitrate and nitrite as nitrogen concentrations in the Iowa River:

Example of the spatial distribution of nitrate concentrations in the Iowa-Cedar Basin rivers in January 1990. Detection of this characteristic spatial pattern of river pollution (Northwest - Southeast) by the traditional methods is very difficult, ...if possible.

Project Flwprc

Button opens the project Flwprc.

The bar charts that represent the monthly flow rate and the average precipitation depth for a selected time period can be drawn at the center of selected modeling units ( draws charts of the flow rate and draws charts of the precipitation depth). The script associated with the button displays all recorded flow rate (monthly values from January 1960 to December 1992) for a specified modeling unit. Since charts of monthly flow rate for multiple year periods can be drawn, the wet or dry years can be easily identified.

After the buttons or are selected the user is prompted to enter the name of the legend, and to specify the time period (months and years):

Than the size of the bars must be selected:

  • Task: Compare discharge in the selected gaging stations located along the Iowa river in years 1989 to 1991.
  • Click on any chart to make it active;
  • Click the button to minimize all charts;
  • Add a new View;
  • Add the theme gsflow (discharge measured in 28 gaging stations in years from 1940 to 1992), make it active and visible
  • Add the map of rivers to the view (coverage rf1);
  • Select stations located on the Iowa River;
  • click the button. Enter the beginning month 1/1990 and the ending month 12/1991. Select small size of bars;
  • Below, an example map is presented. The charts chow that the 1989 year was a dry year compared to years 1990 and 1991.

These tools for drawing charts can be used to draw a discharge profiles along selected rivers.

  • Task: Draw a dischrage profile along the Iowa River
  • Open a new View;
  • Add the theme unflow (discharge from 1033 units for the years 1960 to 1992), make it active and visible
  • Select the Iowa River flow path. Click the button and then click an unit in the upper portion of the Iowa River basin. Wait a moment. Avenue script selects all downstream units to the one you clicked on;
  • click the button. To draw a discharge profile in the Iowa River in January 1990 enter from 1990, January and to 1990, January. Select medium or large size of bars;
  • Add a map of rivers to the view (coverage rf1)
  • Make rivers blue and unflow polygons of coverage white (both, foreground and outline);
  • Note: the decrease in the flow in the middle of the Iowa river is most likely caused by the Coralville Lake losses or management.

The button must be pressed to display all recorded flow rate (monthly values from January 1960 to December 1992) in a form of a dynamic chart. If this button is pressed, user can click on a selected gauging station, theme gsflow, or on a selected modeling unit, theme unflow. Two forms of the charts can be selected: Horizontal--bars move from right to left, or Vertical--bars move from up to down.

After chart form is selected the user is prompted to select the time interval for which the data will be displayed.

Before the charts are drawn, the maximum value of the discharge is found to scale the chart. This value is displayed:

The dynamic chart is displayed after user click on the YES button:

The process of displaying series of charts can be stopped in any moment by pressing the stop button displayed in the lower left corner of the ArcView window (left end of the status bar).

After the data over selected time period is displayed, the user is asked if she/he wants to minimize the chart window:

Project Tools

To use a tool a table has to be opened and active. If the active document is not a table an error message is displayed:

The following tools have been created to support the model of agrichemicals in surface waters:

Determines the unit/stream order in the flow system (Avenue script order6);

Then user has to determine a field in which the order will be stored:

Calculates weighted average for each modeling unit total drainage area (script upavg2);

If the specified field does not exist:

If YES was selected:

Accumulates values going along the flow path (script cumul2);

If the entered field can not be found:

Calculates the difference between the unit inputs and the unit output (script decom2).

If the field does not exist:

The procedure of reverse flow-accumulation is very useful to determine the difference between the outflow from the modeling units and the sum of inflows, and therefore to asses water gains/losses over large areas. Below an example of "un-cumulated" flow in the Iowa-Cedar River basin for June 1990. The losses along the Cedar River are evident. Also, the dark blue region close to the Iowa City indicates release of water from the Coralville Lake.


These materials may be used for study, research, and education, but please credit the authors and the Center for Research in Water Resources, The University of Texas at Austin. All commercial rights reserved. Copyright 1997 Center for Research in Water Resources.