December 20, 2012

U.S. Environmental Protection Agency, Office of Pesticide Programs, Environmental Fate and Effects Division
Health Canada, Pesticide Management Regulatory Agency, Environmental Assessment Directorate

On this Page

• Purpose
• Groundwater Conceptual Model
• Implementation of the Conceptual Model into PRZM
  • Field Capacity
  • Irrigation
  • Dispersion
  • Near Surface Degradation
  • Temperature
• Selection of Scenario Locations
• Groundwater Scenario Development Guidance for PRZM
• References
• Appendix 1. Guidance on Setting up Degradation in the Profile

Purpose

This guidance describes the methods for developing scenarios for the PRZM-GW model. PRZM-GW is an implementation of the Pesticide Root Zone Model (Suarez, 2006) in which a pesticide moves downward into an aquifer according to the NAFTA conceptual model (see below). Scenarios are the subset of input parameters that describe the spatial characteristics of the PRZM-GW parameters. These include weather, soil, and crop characteristics.

Top of Page

Groundwater Conceptual Model

Figure 1 depicts the conceptualization of the groundwater scenario that evolved from meetings between the USEPA and Canada's PMRA and from the 2005 N-Methyl Carbamate SAP meetings (USEPA, 2005a, 2005b). The model represents vulnerable private drinking water wells in the vicinity of agricultural environments. In this conceptualization, the pesticide is applied to the soil surface or plant canopy, and precipitation or irrigation drive pesticide into the soil. Meteorological, crop, biological, chemical, and management processes affect the transport of the pesticide as it moves through the soil and into a saturated zone. Horizontal surface movement of a pesticide (via runoff or erosion) and subsequent removal is assumed negligible for this model.

The saturated zone of the conceptual model is a shallow unconfined aquifer with a water table depth that corresponds to the scenario location. The well-screen extends from the aquifer surface to 1 meter below the surface, but this length is also adjustable according to common practices. The default well-screen length of 1 meter was chosen to sample the higher concentrations expected to be closest to the water table; however, this can be adjusted based on site-specific data. Pesticide concentrations in the well are taken as the average concentration in the screened zone.

The conceptual model includes meteorological events that can significantly affect pesticide transport, including precipitation, evaporation, snow, temperature, and wind. These weather processes vary and are simulated with daily resolution. Daily resolution is required by risk assessment applications within OPP's and PMRA's Health Effects Division.

Soil temperature variations with depth and time have important implications on transport and are thus included in the model. Temperature simulations on and below the field are more important for groundwater simulations than for surface water simulations. In a groundwater simulation, the entire duration of the simulation occurs on or below the field; whereas in a surface water simulation the relevant pesticide is in the field's zone of relevance (top 2 cm) for much shorter periods and will be less likely to be in that surface zone during periods of extreme cold when temperature effects will be significant. It is likely that in many scenarios temperature in the subsurface could be quite different from surface temperatures or from the laboratory experiment temperature of the degradation experiments.

Crop descriptions are needed only in regard to how they impact hydrology (i.e., water balance) or pesticide transport. In most cases, parameterization of pesticide hold up and wash off of foliage will be difficult due to lack of data, thus prudence should be considered in attempting to simulate these processes. A description of canopy coverage and root depth is incorporated to capture the most relevant plant influences on pesticide transport.

Agricultural management practices that could dramatically affect pesticide transport such as irrigation are included in the conceptual model. Because irrigation provides significant amounts of water and is an important transport driver, irrigation is included in those areas where it is used. Simulation of other management practices such as soil manipulation, tile drainage, and pesticide application methods may be desirable but are not considered in the present conceptualization because there is a lack of consensus regarding their importance to transport. Such processes may be considered in subsequent versions if more evidence of their impact becomes available.

The conceptual model includes consideration for pesticide degradation in the soil profile. In general, degradation occurs faster in the top of the profile and decreases at lower depths. As depicted in Figure 1 (by the black line on the left of the figure), the pesticide fate conceptualization uses the pesticide's aerobic soil degradation rate in the top 10 cm, and a linearly declining rate with depth to 1 meter, below which only abiotic processes are assumed to occur. In some cases, there may be evidence that certain pesticides behave differently than this depiction. The conceptual model does not preclude alternative degradation schemes when there is compelling evidence that a pesticide may behave differently than the default conceptualization.

Figure 1
General Groundwater Scenario Concept for Estimating Pesticide Concentrations in Drinking Water
The black line through the soil profile represents the decline in degradation with respect to the depth.

Top of Page

Implementation of the Conceptual Model into PRZM

PRZM is a one-dimensional, finite-difference vertical-transport model that accounts for pesticide fate. PRZM can simulate specific site, pesticide, and management properties, including field properties (organic matter, water holding capacity, bulk density, general hydrology), pesticide application parameters (application rate, frequency, incorporation, and methods), and pesticide environmental fate and transport properties (aerobic soil metabolism half-life, soil-water partitioning coefficients, foliar degradation and dissipation). PRZM is the standard EPA model for simulating transport of pesticides off of a field and into a surface water body.

In order to use PRZM to simulate transport to groundwater rather than to surface water, environmental scenarios need to be developed to represent the conceptual model. These scenarios will differ from surface water scenarios in particular for a few key parameters, including field capacity, irrigation, dispersion, near surface degradation, the non-consideration of erosion and runoff, temperature simulation, and relevant output. The basis for these changes is described below. Detailed guidance for selection of these and other parameters is included in the section - Groundwater Scenario Development Guidance.

Top of Page

• Field Capacity

  Although PRZM was not originally developed to simulate saturated conditions, a review of the model's structure and coding showed that saturated conditions can be effectively simulated by redefining the field capacity parameter. Soil compartments in PRZM are maintained at "field capacity" unless losses occur by evapotranspiration. Below the zone of evapotranspiration (which occurs at the bottom of the defined root zone or at a minimum defined evaporation depth), PRZM maintains the water content at "field capacity." Thus, a saturated zone can be created in PRZM by setting the scenario's field capacity parameter (called THEFC in PRZM) equal to the porosity. By doing this, a constant water table can be simulated (Figure 2).

  PRZM-GW output concentrations are calculated from the spatial average over the depth from the top of the saturated zone to the bottom of the well screen (1 meter default). Depending on precipitation and evapotranspiration characteristics of the particular scenario, this spatial averaging represents the temporal average of six months to greater than one year (i.e., 1 meter of aquifer can hold six months of water from the FL scenarios to greater than a year for eastern Washington scenarios; scenarios described later). Note that during this time period, degradation within the aquifer still occurs, and thus this conceptualization is more realistic with regard to concentrations that may be pulled from a well than the temporal averaging described by the FOCUS group, in which post-breakthrough degradation does not occur.

  In addition, analysis of prospective groundwater (PGW) studies indicates that actual water retention in the soil profile is better characterized by a value between field capacity and saturation (Young and Carleton, 2012). Therefore, the soil profile water holding capacity (field capacity in PRZM) should be set to a value that is the midpoint between field capacity and saturation. In this way, the travel time to the aquifer will be better simulated.

  Figure 2
  PRZM Scenario to Simulate a Fixed Water Table

  In Figure 2, note the unique adjustments to PRZM groundwater scenarios: "field capacity" is set to porosity to simulate a saturated zone; well-screen of 1 meter is the default length.

  Top of Page

  ---

• Irrigation

  Irrigation may be an important contributor to vertical pesticide transport and in many scenarios may be the dominant water source. When a scenario specifies irrigation, it is handled according to OPP guidance on PRZM irrigation (USEPA/OPP, 2007), which modifies the guidance provided in the PRZM manual. PRZM irrigation guidance that appears in the PRZM manual may be inappropriate for deeply rooted plants such as orchard trees. PRZM attempts to satisfy water demand to the bottom of the root zone, whereas in reality water demand would only be satisfied as a result of irrigation to several centimeters depth. Thus too much irrigation would occur unless preventative measures are taken. This should be considered in scenario development if irrigation is to be simulated. Guidance is provided for this case later in the parameter selection section of this document.

  Top of Page

  ---

• Dispersion

  Dispersion (whether numerical or explicitly modeled) has an especially important impact on pesticide concentrations when there is a limited degradation zone that applies only over a short distance, such as in this groundwater conceptual model. In these cases, dispersion will quickly remove the pesticide from the degradation zone into areas of low or no degradation. Thus, higher dispersion leads to higher pesticide concentrations, the opposite of what would be expected for a conservative non-degrading tracer. This characteristic is analogous to the efficiency difference between plug flow and completely mixed reactors, which is well known in chemical and environmental engineering fields. Thus, it is important to standardize the discretization within PRZM.

  In PRZM, the vertical discretization is the dominant means of applying dispersion, which is caused by PRZM's use of backwards spatial differencing for the pesticide transport finite differencing. Although backward differencing is effective for elimination of oscillation at boundaries, it produces high but quantifiable numerical dispersion. Additional numerical dispersion also results from the fully implicit discretization of the temporal term (1 to 2 cm), but users have no control over this characteristic (Young and Carleton, 2012). For a backward finite difference, dispersivity is equal to one half of the spatial discretization.

  EFED evaluation of PGW studies shows dispersivity values much greater than the 2.5 cm used by the European Union, as described in the FOCUS guidance. In an analysis of bromide tracers in PGW studies [see Figure 3], dispersivity increased with depth and ranged from about 10 cm at 1 m, increasing to around 40 cm at 5 m. In order to capture the dispersivity in the PRZM scenarios, the scenarios use discretization of 1 cm for the top 10 cm. For the next 10 cm, discretizations are 5 cm. From 20 cm to 1 m, discretizations are 20 cm. The standard scenarios use discretizations of 50 cm for all depths below 1 m, but the PRZM-GW user interface allows other discretizations if the user has justification for alternate schemes.

  Figure 3
  Dispersivity Determined from Bromide Tracers at 21 PGW Sites
  (Young and Carleton, 2012)

  Top of Page

  ---

• Near Surface Degradation

  Aerobic soil metabolism is modeled as declining with depth and becomes zero at 1 m below the surface. Because typically submitted soil degradation studies do not allow differentiation of water-phase from solid-phase degradation, only total system degradation is typically known. Thus the rates of both solid phase and water phase degradation should be set identically and should decline at the same rate with depth.

  For the groundwater conceptualization, this can cause some difficulties when hydrolysis is significant, especially when the degradation rate in water drops below the hydrolysis rate, creating apparent inconsistencies. In order to overcome this problem, the following recommendations apply:

  1. both solid and aqueous degradation rates (represented initially by aerobic soil metabolism rate) are to decline in a linear fashion with depth;

  2. if the aqueous degradation rate drops below the hydrolysis rate, then the hydrolysis rate will be used for the aqueous degradation rate, while the solid degradation rate remains unaffected by the hydrolysis rate and continues to decline with depth.

  An example of the use of these recommendations is given in Figure 4. Figure 4 shows that the aqueous-phase degradation rate does not decline below the hydrolysis rate, while the sorbed-phase rate continues to decline.

  Figure 4
  A Depiction of Declining Degradation with Depth in the Presence of Hydrolysis

  In this case, the hydrolysis rate is 0.15 times the aerobic soil metabolism rate. Hypothetical partitioning conditions with the compound distributed in water and solids is the ratio of 2.8:1

  Top of Page

  ---

• Temperature

  The temperature routine in PRZM allows for heat transfer between the bottom boundary of the simulation (set at a constant temperature) and the soil surface. The temperature routine requires that solar radiation be included in the meteorological file. Solar radiation has not previously been included in the surface water meteorological files; nevertheless, all current weather files have included solar radiation. Thus, standard weather files will work correctly. There are a few other parameters that are needed to run the temperature routine, which are new to the standard scenario development (e.g., albedo, emissivity), and which are not readily available for specific scenarios. Without further specific information regarding this parameterization, typical values should be chosen as specified in the guidance below.

Top of Page

Selection of Scenario Locations

Appropriate scenario locations should be created in regions that are vulnerable to groundwater contaminations. This includes areas where the depth to groundwater is shallow (<100 feet), where there is high net leaching (or high rainfall with respect to evaporation), and where specific characteristics of the location may make a pesticide more prone to persistence (e.g., high or low pH depending on the pesticide).

Top of Page

Groundwater Scenario Development Guidance for PRZM

Guidance is presented in Table1 on how to develop PRZM scenarios for groundwater estimates. The output of these model runs are contained in the PRZM "zts" file, which will give the daily spatially averaged (vertically over 1 m) aqueous concentration. These values represent high-end drinking water concentrations from a rural well. The scenario file can be constructed in the following manner:

Top of Page

References

Suarez, L.A., 2006. PRZM-3, A Model for Predicting Pesticide and Nitrogen Fate in the Crop Root and Unsaturated Soil Zones: Users Manual for Release 3.12.2. EPA/600/R-05/111 September 2006, revision a.

USEPA, 2005a. N-Methyl Carbamate Pesticide Cumulative Risk Assessment: Pilot Cumulative Analysis, February 15 - 18, 2005, Docket Number: OPP-2004-0405

USEPA, 2005b. Preliminary N-Methyl Carbamate Cumulative Risk Assessment, August 23 - 26, 2005, Docket Number: OPP-2005-0172

USEPA OPP, 2002. Guidance for Selecting Input Parameters in Modeling the Environmental Fate and Transport of Pesticides, available on the OPP Water Models web page.

USEPA/OPP, 2007. WQTT Advisory Note: Irrigation Guidance for Developing PRZM Scenarios, May 9, 2007.

Young, D.F. and Carleton, J.N. (2012) Estimating Dispersivity, Mass Recovery and Water Hold Up in Field-Scale Leaching Studies by Use of a Capacity Model, Abstract H11B-1173 presented at 2012 Fall Meeting, AGU, San Francisco, Calif., 3-7 Dec.

Top of Page

Appendix 1
Guidance on Setting up Degradation in the Profile

The PRZM implementation of the groundwater conceptual model uses eight soil layers as shown in Table A1 and also depicted in Figure A1. The first layer is 10 cm and is discretized at 1 cm increments to allow some precision in defining root depths and evaporations zones and pesticide incorporation into the soil. The next layer is 10 cm thick and is discretized at 5 cm increments. The next four layers are all 20 cm and contain one compartment. These first six layers make up the top 1 meter of soil and represent the aerobic degradation zone. Degradation in the top meter declines proportionally with depth after the first 10 cm. Degradation reduction factors are given in the table below. The degradation rate in each layer is to be multiplied by the respective factor from the table. The next layer represents the depth between the aerobic zone and the water table. This layer is also discretized at 50 cm intervals. The eighth layer represents the water table and also has 50 cm discretization. Aerobic degradation does not occur in the seventh and eighth layers; however, hydrolysis is not precluded in these or any other horizons.

Figure A1
Degradation Reduction Factors and Discretizations
for Layers and Nodes in PRZM Standard Groundwater Scenarios

Top of Page