Assessing the Potential Exposure of Groundwater to Pesticides: A Model Comparison

Abstract

Many pesticide-leaching studies rely on the use of numerical models that account for various physicochemical and biological processes at a range of temporal and spatial scales. In leaching assessments for the registration of pesticides in the European Union and the United States, one-dimensional models are used that describe the vertical transport of soil water and pesticides to groundwater. One-dimensional models are most representative for spray and broadcast applications of pesticides, but they may not be adequate for spatially nonuniform applications such as microdrip irrigation. We tested whether an advanced numerical model capable of one-, two-, and three-dimensional flow and pesticide transport simulation [HYDRUS (2D/3D)] provides results consistent with one-dimensional numerical models commonly used for pesticide registration (PEARL, PELMO). Model comparison was used to assess conceptual uncertainty in soil physical representation of pesticide transport among numerical models across a range of applications. Simulations of soil water flow and pesticide and heat transport were conducted for four different hypothetical pesticides and nine European regulatory standard pedoclimatic scenarios (FOCUS). The results show that the predicted annual amount of pesticide mass leached below the depth of 1 m, for all substances, locations, and simulation years, was on the same order of magnitude among the three models. Additionally, the simulated 80th percentile of annual pesticide concentrations at the 1-m depth was not statistically different among models. HYDRUS (2D/3D) can, therefore, be used as an alternative model for pesticide assessment studies and provides a conceptual framework consistent with PEARL and PELMO but capable of two-and three-dimensional applications as well.

Crop growth and soil water fluxes at erosion‐affected arable sites: Using weighing lysimeter data for model intercomparison

“…However, a recent investigation using ET a in crop model intercomparison studies showed that much of the difference and variability in predicting ET a by crop models was mainly a consequence of using different approaches to estimate potential ET (Kimball et al., 2019). This example demonstrates that in addition to differences in crop model complexity and model functionality, the selection of boundary conditions (Diamantopoulos et al., 2017) can largely affect the outcome of crop models. Thus, not only the selection of the upper boundary (e.g., reference ET) but also the selection of the bottom boundary condition might be relevant in model intercomparison studies if water from deeper soil layers or the water table is available for capillary rise and may serve as an additional water supply for root water uptake and evaporation (Groh, Vanderborght, Pütz, & Vereecken, 2016; Luo & Sophocleous, 2010).…”

Integrating ecosystem services into crop protection and pest management: Case study with the soil fumigant 1,3-dichloropropene and its use in tomato production in Italy

“…Based on information gathered from interviews with farmers, the depth to groundwater at the study locations was generally 3 m or greater. Where concentrations were predicted to exceed the drinking water regulatory limit at 1‐m depth—for the most vulnerable scenarios—further modeling was performed using Hydrus 2D/3D version 2.0 (Sejna et al ; Diamantopoulos et al ). Developed by the US Department of Agriculture, Hydrus is a standard model for detailed simulation of pesticide fate and transport and was recently evaluated by a working group of the Society of Environmental Toxicology and Chemistry (Van den Berg et al ).…”

Pedotransfer Function for the Brunswick Soil Hydraulic Property Model and Comparison to the van Genuchten‐Mualem Model

“…The accurate representation of unsaturated water fluxes in soils is important for reliable descriptions of water and energy fluxes ( van Genuchten & Pachepsky, 2011) at the soil-atmosphere boundary (Vereecken et al, 2019), for the prediction of plant growth and yields (Angulo et al, 2014;Gayler et al, 2009;Hoffmann et al, 2016;Lawless et al, 2008), including irrigation optimization (Elmaloglou et al, 2010), and for the quantification of the environmental fate of agrochemicals (Diamantopoulos et al, 2017;Vereecken et al, 2016). For this, the Richards equation, which is now known as the Richardson equation (Raats & Knight, 2018), is considered as de facto standard (Diamantopoulos & Durner, 2012).…”