Article

A refined lack-of-fit statistic to calibrate pesticide fate models for responsive systems

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Calibration by inverse modelling was performed with the MACRO transport and fate model using long-term (>10 years) drainflow and isoproturon (IPU) data from western France. Two lack-of-fit (LOF) indices were used to control the inverse modelling: sum of squares (SS) and an alternative statistic called the vertical-horizontal distance integrator (VHDI), which is designed to account for offsets in observed and predicted arrival times of peak IPU concentration. With these data, SS was artificially inflated because it is limited to comparison of predicted and observed IPU concentrations that are concurrent in time. The LOFs were used along with the index of agreement (d) and the correlation coefficient (r) to ascertain the fit of the calibrated models. Predicted arrival times of peak IPU concentration differed somewhat from observed times. All four indices indicated better model fit for the second of two validation periods when inverse modelling was controlled by VHDI rather than SS (SS = 26.4, d = 0.660, r = 0.606 and VHDI = 1.25). The VHDI statistic was markedly lower compared with the uncalibrated model (38.0) and SS calibration results (24.5). The final maximum predicted IPU concentration (44.5 microg L(-1)) for the calibration period was very similar to the observed value (44 microg L(-1)). VHDI is seen as an effective alternative to SS for calibration and validation of pesticide fate models applied to responsive systems. VHDI provided a more realistic assessment of model performance for the transient flows and short-lived concentrations observed here, and also effectively substituted for the objective function in inverse modelling.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... There seems to be insufficient data to date to bring the MACRO model calibration approaches to some uniformity. In addition, it remains unclear whether the model, once calibrated, will adequately predict targets (percolation and IPU concentration), since in a study by Nolan et al. [22] the calibrated model for one data set gave a better prediction than for another. ...
... µg/L was observed in the period 20-40 days after application. In addition, a rapid breakthrough of isoproturon in the period 6-15 days after application was also observed by Nolan et al. [22]. July 2015 was rainy; in total, from the moment of processing until the end of July, 199 mm of precipitation fell. ...
Article
Full-text available
Assessing the risk of using pesticides for the environment in general, and for groundwater in particular, necessitates prediction of pesticide migration. For this purpose, mathematical models of pesticide behavior are utilized, which must be parameterized and calibrated based on experimental data to make them perform properly. The behavior of the pesticide cyantraniliprole was examined in a long-term lysimetric experiment. The MACRO 5.2 dual porosity model was calibrated based on the percolate and the levels of pesticides in the soil profile and percolate. Despite employing experimentally verified soil parameters and pedotransfer functions (PTF), the model must be calibrated for percolation. This is due to the model’s properties as well as the complexity of the soil as an object of study, and its pore space, which is subject to daily and annual fluctuations. It is the parameters that describe the structure of the pore space that need to be calibrated. Calibrating for pesticide concentrations required a minor revision of the sorption and transformation rates, as well as an increase in the dispersivity and ASCALE values.
... Other examples of methods used to calibrate the MACRO model include Markov chain Monte Carlo methods (Larsbo et al., 2009) and Levenberg-Marquardt fitting (Nolan et al., 2009). Common to these automatic calibration methods is that parameter values in subsequent simulations are dependent on the analysis of the preceding simulations. ...
... Small temporal offsets in drain discharge or pesticide concentration peaks may be of little consequence from a risk assessment point of view but have a large effect on the goodness-of-fit measures (Armstrong et al., 1996). To address this problem, the verticalhorizontal distance integrator (VHDI) was recently intro-duced and evaluated in a MACRO application (Nolan et al., 2009). The VHDI takes both the error in peak values and the error in the timing of peak values into account and seems to be an attractive alternative to the RMSE and NS for macroporous soils where responses to rainfall events are rapid and short-lived. ...
Article
MACRO is a one-dimensional dual-permeability model of variably saturated water flow and reactive solute transport in soil that has been used since the early 1990s as a research tool to investigate the effects of macropore flow on soil hydrology and contaminant transport under transient field conditions. It is also widely used in the form of bespoke versions in pesticide exposure and risk assessments for groundwater and surface waters, e.g., in registration procedures in the European Union (EU). Macropore flow is a highly episodic, fast, non-equilibrium process that can dominate the leaching of reactive solutes in structured soils. This has important consequences for model calibration and validation procedures. Firstly, it means that in addition to water contents and resident solute concentrations, water flows and flux concentrations measured at high time resolution are required. Secondly, it implies that more weight must be placed on flux data obtained during these important but short-lived episodic flow events if parameters controlling macropore flow are to be reliably estimated. Although the choice of approach will vary with the purpose, automatic or hybrid automatic/manual calibration procedures are generally recommended for MACRO, especially global methods that account for uncertainty within a multi-objective framework. Despite the complexity of the processes it attempts to represent, MACRO is a parsimonious model, requiring only five additional parameters to simulate water flow and reactive solute transport, compared to the use of the Richards equation and the advection-dispersion equation. Nevertheless, for practical reasons, the size of the parameter space that can be explored by calibration is often quite limited. This emphasizes the importance of measuring those parameters that can be measured and the role of sensitivity analyses in supporting the choice of parameters to calibrate. This will vary between applications, but at least for structured soils, all five parameters regulating the generation and strength of non-equilibrium flow and transport are usually rather influential. © 2012 American Society of Agricultural and Biological Engineers.
... Indeed, there has been worldwide interest in the application of inverse Accepted Article modeling to various mathematical models for pesticides. 13,[16][17][18][19][20] However, the potential problem of the local optimization algorithm is that the minimized objective function may not be the global minimum but the local minimum, which is significantly affected by the starting value. 16 In recent years, the application of a global optimization algorithm incorporating the Markov-chain Monte-Carlo (MCMC) technique has been focused on to avoid this problem. ...
Article
Full-text available
BACKGROUND In Japan, while experimental data for the dissipation behavior of paddy pesticides under a standardized test system are available, the application of a mathematical model is limited. This paper proposes a new model calibration procedure for inversely deriving the model parameters from the experimental data. This procedure is tested in the open software R by running an improved Pesticide Concentration in Paddy Field‐1 (PCPF‐1) model with R packages to analyze the dissipation of simetryn and molinate in flooded lysimeters and paddy fields. RESULTS The model fitting was performed by a random minimization routine. Furthermore, the uncertainties of the model parameters envisioned by the global sensitivity analysis were successfully reduced using the Markov‐chain Monte‐Carlo technique. The calibrated simulation was validated at each test plot by confirming multiple statistical indices (i.e., Nash‐Sutcliffe efficiency: 0.88–1.00; percent bias: <±5%). The dissipation pathways of two herbicides were quantitatively clarified by the mass balance of calibrated simulations and the effect of the unexpected herbicide runoff was quantified. The case study showed that the adjustment of daily percolation rate in the lysimeter experiment is the key to simulate the actual paddy field condition more accurately, especially in a case where pesticides show higher water solubility and soil mobility. CONCLUSION The developed procedure can analyze the experimental data with acceptable accuracy and extract the unobservable information quantitatively. Our approach is applicable to the optimization of not only the model but also future experimental design. This article is protected by copyright. All rights reserved.
... Cette étape doit être réalisée en comparant les prévisions des modèles à des données obtenues au champ (Carsel et al., 1985 ;Parrish et al., 1992 ;Dubus et al., 2002). Les résultats publiés dans la littérature montrent que PRZM, PEARL, MACRO et PELMO sont capables de reproduire plus ou moins correctement, en fonction du contexte (climat, sol, culture, pesticide), la dissipation et le transfert des pesticides en plein champ (Brown et al., 2004 ;Jackson et al., 2005 ;Scorza Junior et Boesten, 2005 ;Mamy et al., 2008 ;Luo et Zhang, 2009 ;Nolan et al., 2009 ;Rosenbom et al., 2009 ;Undabeytia et al., 2009 ;Leistra et Boesten, 2010 (Saseendran et al., 2004 ;Ma et al., 2005 ;Saseendran et al., 2007). Par ailleurs, la performance de RZWQM pour simuler les transferts de pesticides dans les systèmes de cultures est correcte mais bien souvent le modèle a été calé au préalable (Malone et al., 2004). ...
Thesis
L'objectif de ce projet de thèse consiste à développer une approche de modélisation permettant d'évaluer et de comparer les performances environnementales de prototypes de systèmes de culture conçus pour réduire l'usage des pesticides, dans le cadre du plan Ecophyto, en tenant compte de la variabilité spatio-temporelle des conditions agro-pédo-climatiques, et des principales sources d'incertitudes liées aux molécules, milieux, modèles Les résultats permettront de (1) fournir une méthode d'évaluation quantitative des performances environnementales de systèmes de culture tenant compte des processus conditionnant le devenir des pesticides dans les sols et l'environnement, de la variabilité agro-pédo-climatique et des diverses sources d'incertitudes ; (2) proposer une démarche aidant au choix de systèmes de culture ; et (3) identifier les systèmes de culture permettant d'atteindre les objectifs du plan Ecophyto 2018, donc aider à réduire l'exposition des écosystèmes aux pesticides.
... Preferential solute leaching is "event-driven", which makes the Nash-Sutcliffe model efficiency very sensitive to even slight errors in timing. Nolan et al. (2009) proposed a "refined lack-of-fit statistic" that measures the ability of a model to predict peak 20 concentrations and takes into account possible shifts in time (and scale). It is unfortunately not easily applicable here, as not all our simulations exhibit clearly defined solute leaching peaks. ...
Article
Full-text available
Estimating pesticide leaching risks at the regional scale requires the ability to completely parameterise a pesticide fate model using only survey data, such as soil and land-use maps. Such parameterisation usually rely on a set of lookup tables and (pedo)transfer functions, relating elementary soil and site properties to model parameters. The aim of this paper is to describe and test a complete set of parameter estimation algorithms developed for the pesticide fate model MACRO, which accounts for preferential flow in soil macropores. We used tracer monitoring data from 16 lysimeter studies, carried out in three European countries, to evaluate the ability of MACRO and this "blind parameterisation" scheme to reproduce measured solute leaching at the base of each lysimeter. We focused on the prediction of early tracer breakthrough due to preferential flow, because this is critical for pesticide leaching. We then calibrated a selected number of parameters in order to assess to what extent the prediction of water and solute leaching could be improved. Our results show that water flow was generally reasonably well predicted (median model efficiency, ME, of 0.42). Although the general pattern of solute leaching was reproduced well by the model, the overall model efficiency was low (median ME = −0.26) due to errors in the timing and magnitude of some peaks. Preferential solute leaching at early pore volumes was also systematically underestimated. Nonetheless, the ranking of soils according to solute loads at early pore volumes was reasonably well estimated (concordance correlation coefficient, CCC, between 0.54 and 0.72). Moreover, we also found that ignoring macropore flow leads to a significant deterioration in the ability of the model to reproduce the observed leaching pattern, and especially the early breakthrough in some soils. Finally, the calibration procedure showed that improving the estimation of solute transport parameters is probably more important than the estimation of water flow parameters. Overall, the results are encouraging for the use of this modelling set-up to estimate pesticide leaching risks at the regional-scale, especially where the objective is to identify vulnerable soils and "source" areas of contamination.
... Preferential solute leaching is "event-driven", which makes the Nash-Sutcliffe model efficiency very sensitive to even slight errors in timing. Nolan et al. (2009) proposed a "refined lack-of-fit statistic" that measures the ability of a model to predict peak concentrations and takes into account possible shifts in time (and scale). It is unfortunately not easily applicable here, as not all our simulations exhibit clearly defined solute leaching peaks. ...
Article
Full-text available
Estimating pesticide leaching risks at the regional scale requires the ability to completely parameterise a pesticide fate model using only survey data, such as soil and land-use maps. Such parameterisations usually rely on a set of lookup tables and (pedo)transfer functions, relating elementary soil and site properties to model parameters. The aim of this paper is to describe and test a complete set of parameter estimation algorithms developed for the pesticide fate model MACRO, which accounts for preferential flow in soil macropores. We used tracer monitoring data from 16 lysimeter studies, carried out in three European countries, to evaluate the ability of MACRO and this "blind parameterisation" scheme to reproduce measured solute leaching at the base of each lysimeter. We focused on the prediction of early tracer breakthrough due to preferential flow, because this is critical for pesticide leaching. We then calibrated a selected number of parameters in order to assess to what extent the prediction of water and solute leaching could be improved. Our results show that water flow was generally reasonably well predicted (median model efficiency, ME, of 0.42). Although the general pattern of solute leaching was reproduced well by the model, the overall model efficiency was low (median ME = −0.26) due to errors in the timing and magnitude of some peaks. Preferential solute leaching at early pore volumes was also systematically underestimated. Nonetheless, the ranking of soils according to solute loads at early pore volumes was reasonably well estimated (concordance correlation coefficient, CCC, between 0.54 and 0.72). Moreover, we also found that ignoring macropore flow leads to a significant deterioration in the ability of the model to reproduce the observed leaching pattern, and especially the early breakthrough in some soils. Finally, the calibration procedure showed that improving the estimation of solute transport parameters is probably more important than the estimation of water flow parameters. Overall, the results are encouraging for the use of this modelling set-up to estimate pesticide leaching risks at the regional-scale, especially where the objective is to identify vulnerable soils and "source" areas of contamination.
... 30,2011 metabolite concentrations resulted from trying to match observations near or below the detection limit; some negative correlations were computed where the predicted peaks in concentrations of the primary metabolite were out of phase with the observed maxima. An objective function with more tolerance to phase mismatch may help improve the overall models [33]. ...
Article
In the present study a branched serial first-order decay (BSFOD) model is presented and used to derive transformation rates describing the decay of a common herbicide, atrazine, and its metabolites observed in unsaturated soils adapted to previous atrazine applications and in soils with no history of atrazine applications. Calibration of BSFOD models for soils throughout the country can reduce the uncertainty, relative to that of traditional models, in predicting the fate and transport of pesticides and their metabolites and thus support improved agricultural management schemes for reducing threats to the environment. Results from application of the BSFOD model to better understand the degradation of atrazine supports two previously reported conclusions: atrazine (6-chloro-N-ethyl-N'-(1-methylethyl)-1,3,5-triazine-2,4-diamine) and its primary metabolites are less persistent in adapted soils than in nonadapted soils; and hydroxyatrazine was the dominant primary metabolite in most of the soils tested. In addition, a method to simulate BSFOD in a one-dimensional solute-transport unsaturated zone model is also presented.
Article
Preferential flow from stream banks is an important component of pesticide transport in the mountainous areas of northern Thailand. Models can help evaluate and interpret field data and help identify the most important transport processes. We developed a simple model to simulate the loss of pesticides from a sloped litchi (Litchi chinensis Sonn.) orchard to an adjacent stream. The water regime was modeled with a two-domain reservoir model, which accounts for rapid preferential flow simultaneously with slow flow processes in the soil matrix. Preferential flow is triggered when the topsoil matrix is saturated or the infiltration capacity exceeded. In addition, close to matrix saturation, rainfall events induce water release to the fractures and lead to desorption of pesticides from fracture walls and outflow to the stream. Pesticides undergo first order degradation and equilibrium sorption to soil matrix and fracture walls. The model was able to reproduce the dynamics of the discharge reasonably well (model efficiency [EF] = 0.56). The cumulative pesticide mass (EF = 0.91) and the pesticide concentration in the stream were slightly underestimated, but the deviation from measurement data is acceptable. Shape and timing of the simulated concentration peaks occurred in the same pattern as observed data. While the effect of surface runoff and preferential interflow on pesticide mass transport could not be absolutely clarified, according to our simulations, most concentration peaks in the stream are caused by preferential interflow pointing to the important role of this flow path in the hilly areas of northern Thailand.
Article
Full-text available
Correlation and correlation-based measures (e.g., the coefficient of determination) have been widely used to evaluate the "goodness-of-fit" of hydrologic and hydroclimatic models. These measures are oversensitive to extreme values (outliers) and are insensitive to additive and proportional differences between model predictions and observations. Because of these limitations, correlation-based measures can indicate that a model is a good predictor, even when it is not. In this paper, useful alternative goodness-of-fit or relative error measures (including the coefficient of efficiency and the index of agreement) that overcome many of the limitations of correlation-based measures are discussed. Modifications to these statistics to aid in interpretation are presented. It is concluded that correlation and correlation-based measures should not be used to assess the goodness-of-fit of a hydrologic or hydroclimatic model and that additional evaluation measures (such as summary statistics and absolute error measures) should supplement model evaluation tools.
Chapter
The fit of model simulated values to the observations and prior information tests the ability of the model to realistically represent the simulated system. This chapter presents statistical and graphical methods of evaluating model fit in the order they generally are used in a modeling project.
Article
This study evaluated probabilistic modelling approaches against data on leaching of isoproturon through two contrasting soil types. Leaching through undisturbed lysimeters from a sandy loam (Wick series) and a moderately structured clay loam (Hodnet series) was investigated in seven replicates. The variability of soil properties and of sorption and degradation of isoproturon was estimated by taking 6–14 samples within the areas of lysimeter extraction in the field. Normal distributions were assigned to Koc and DT50 and a large number of values for these two parameters were sampled from each distribution. Parameter values were used to simulate movement of isoproturon through the lysimeters with the preferential flow model MACRO. Uncertainty in output distributions was compared with the variability of measured data. A constrained probabilistic assessment varying only degradation and sorption properties was sufficient to match the observed variability in cumulative leaching from the coarse-textured Wick soil (CV = 79%). Variation of pesticide properties alone could not match observed variability in cumulative leaching from the structured Hodnet soil (CV = 61%) and variability in a number of soil properties was incorporated. For both soils, constrained probabilistic approaches where only the top few most sensitive model inputs were varied were sufficient to match or exceed observed variability.
Article
A field experiment was conducted to investigate preferential transport of herbicides and to explore processes that cause rapid movement. Bromide, chloride, and three herbicides (triasulfuron, atrazine, and terbuthylazine) with different mobility characteristics were applied to six 1.4×1.4 m field plots on a loamy and a sandy soil. At both sites, three of the plots were covered with a plastic roof 1 month before the beginning of the experiment to achieve different initial water contents between the plots. Two days before the beginning of the tracer experiment, crops were removed, and the soil surface was homogenized with a spade to a depth of 15-20 cm. One day after application of the chemicals the plots were irrigated with a sprinkling apparatus. The cumulative amounts of infiltration until the time of sampling were 30, 60, and 90 mm within 1, 2, and 3 days, respectively. A trench was excavated, and soil cores were taken horizontally from a 1×1 m profile in a regular 0.1×0.1 m grid. The loamy and the sandy soil showed completely different transport patterns. In the loamy soil the bulk mass of herbicides remained in the top layer; however, considerable amounts of herbicides were transported below the root zone. A few percent for triasulfuron and atrazine and
Article
Quantitative approaches to the evaluation of model performance were recently examined by Fox (1981). His recommendations are briefly reviewed and a revised set of performance statistics is proposed. It is suggested that the correlation between model-predicted and observed data, commonly described by Pearson's product-moment correlation coefficient, is an insufficient and often misleading measure of accuracy. A complement of differenec and summary univariate indices is presented as the nucleus of a more informative, albeit fundamentally descriptive, approach to model evaluation. Two models that estimate monthly evapotranspiration are comparatively evaluated in order illustrate how the recommended method(s) can be applied.
Article
A simple analytic model is proposed which predicts the unsaturated hydraulic conductivity curves by using the moisture content-capillary head curve and the measured value of the hydraulic conductivity at saturation. It is similar to the Childs and Collis-George (1950) model but uses a modified assumption concerning the hydraulic conductivity of the pore sequence in order to take into account the effect of the larger pore section. A computational method is derived for the determination of the residual water content and for the extrapolation of the water content-capillary head curve as measured in a limited range. The proposed model is compared with the existing practical models of Averjanov (1950), Wyllie and Gardner (1958), and Millington and Quirk (1961) on the basis of the measured data of 45 soils. It seems that the new model is in better agreement with observations.
Book
This is the original first edition published as a physical book by Elsevier. It is woefully out of date. An updated electronic version was published in 2002 by the U.S. Geological Survey, and a completely revised 2020 version with updated methods and supporting materials is listed in my publication list, and is available for download at https://doi.org/10.3133/tm4a3 .
Article
This review discusses the causes and consequences of ‘non-equilibrium’ water flow and solute transport in large structural pores or macropores (root and earthworm channels, fissures and interaggregate voids). The experimental evidence suggests that pores larger than c. 0.3 mm in equivalent cylindrical diameter allow rapid non-equilibrium flow. Apart from their large size and continuity, this is also due to the presence of impermeable linings and coatings that restrict lateral mass exchange. Macropores also represent microsites in soil that are more biologically active, and often more chemically reactive than the bulk soil. However, sorption retardation during transport through such pores is weaker than in the bulk soil, due to their small surface areas and significant kinetic effects, especially in larger macropores. The potential for non-equilibrium water flow and solute transport at any site depends on the nature of the macropore network, which is determined by the factors of structure formation and degradation, including the abundance and activity of soil biota such as earthworms, soil properties (e.g. clay content), site factors (e.g. slope position, drying intensity, vegetation) and management (e.g. cropping, tillage, traffic). A conceptual model is proposed that summarizes these effects of site factors on the inherent potential for non-equilibrium water flow and solute transport in macropores. Initial and boundary conditions determine the extent to which this potential is realized. High rain intensities clearly increase the strength of non-equilibrium flow in macropores, but the effects of initial water content seem complex, due to the confounding effects of soil shrinkage and water repellency. The impacts of macropore flow on water quality are most significant for relatively immobile solutes that are foreign to the soil and whose effects on ecosystem and human health are pronounced even at small leached fractions (e.g. pesticides). The review concludes with a discussion of topics where process understanding is still lacking, and also suggests some potential applications of the considerable knowledge that has accumulated in recent decades.
Article
Inverse models using, for example, nonlinear least-squares regression, provide capabilities that help modelers take full advantage of the insight available from ground-water models. However, lack of information about the requirements and benefits of inverse models is an obstacle to their widespread use. This paper presents a simple ground-water flow problem to illustrate the requirements and benefits of the nonlinear least-squares regression method of inverse modeling and discusses how these attributes apply to field problems. The benefits of inverse modeling include: (1) expedited determination of best fit parameter values; (2) quantification of the (a) quality of calibration, (b) data shortcomings and needs, and (c) confidence limits on parameter estimates and predictions; and (3) identification of issues that are easily overlooked during nonautomated calibration.
Article
A one-year field-scale leaching experiment was conducted on a structured clay soil for the purpose of evaluating the dual-porosity/dual-permeability model MACRO. The model was first calibrated against measurements of water contents, drainflow, and bromide contents in the soil profile and concentrations in drain discharge. Bentazone transport was then simulated without any further model calibration. The model gave acceptable predictions of the water balance, providing the significant water inflow into the plot from the surrounding areas was accounted for. Simulated bromide contents in the soil were, for the most part, within one standard deviation of the measured values. Bromide was measured in tile drainage 26 days after spraying at concentrations >3 mg l−1 (after 43 mm of precipitation), while groundwater concentrations at 2 m depth were as large as 10 mg l−1 only 42 days after spraying. This is a strong indication of macropore flow. The agreement between model predictions and bentazone measurements was on the whole good, especially for the depth profiles and the initial breakthrough in the drainflow, whereas short-term fluctuations in drainage water concentrations were poorly captured. This was probably caused by the model description of first-order mass-exchange between micro- and macropores, which neglects lateral concentration gradients. Judging from statistical criteria, the model accurately predicted bentazone amounts in the soil profile (model efficiency 0.87), while 66% and 89% of the simulated bentazone concentrations in tile drainage were within a factor of 2 and 5 of measured values respectively. Simulations run without macropore flow overestimated bentazone leaching by ca. 20%. In other words, macropore flow reduced leaching in this clay soil, because much of the bentazone was ‘protected’ against bypass flow in macropores, being stored in micropore water moving at a ‘reduced’ convective transport velocity.
Article
A decision-support tool (MACRO—DB) for predicting pesticide fate and mobility in soils is described. MACRO—DB consists of soil, pesticide, climate and crop databases linked to parameter estimation routines and a simulation model (MACRO). The system currently allows access to three soils databases: SEISMIC (c. 400 UK benchmark soils), MARKDATA (26 Swedish soils), and a database that the user can develop independently. Automatic estimation procedures (pedo-transfer functions) translate the soil information into model parameter values. Two pesticide databases are available. The PETE database contains information for over 600 common compounds. The user can also develop a parallel database for new compounds. Sorption and degradation constants are calculated automatically, combining soil and compound properties. A weather database contains long-term daily meteorological data and a weather generator enables synthetic daily weather data to be derived from long-term average climatic data. A separate database contains information on typical planting and harvest dates, root depths, etc. for some common agricultural crops. To account for parameter uncertainty, the user can simulate ‘worst-case’ or ‘average-case’ soil scenarios, utilizing available information on the mean and typical ranges of soil organic carbon content and pH. The user can view and analyse simulation results with flexible, in-built, graphical procedures.
Article
Many environmental studies on the protection of European soil and water resources make use of soil water simulation models. A major obstacle to the wider application of these models is the lack of easily accessible and representative soil hydraulic properties. In order to overcome this apparent lack of data, a project was initiated to bring together the available hydraulic data which resided within different institutions in Europe into one central database. This information was then used to derive a set of pedotransfer functions applicable to studies at a European scale. These pedotransfer functions predict the hydraulic properties from parameters collected during soil surveys and can be a good alternative for costly and time-consuming direct measurement of these properties. A total of 20 institutions from 12 European countries collaborated in establishing the database of draulic operties of uropean oils (HYPRES). This database has a flexible relational structure capable of holding a wide diversity of both soil pedological and hydraulic data. As these data were contributed by 20 different institutions it was necessary to standardise both the particle-size and the hydraulic data. A novel similarity interpolation procedure was successfully used to achieve standardization of particle-sizes according to the FAO clay, silt and sand particle-size ranges. Standardization of hydraulic data was achieved by fitting the Mualem-van Genuchten model parameters to the individual θ(h) and K(h) hydraulic properties stored in HYPRES. The HYPRES database contains information on a total of 5521 soil horizons (including replicates). Of these, 4030 horizons had sufficient data to be used in the derivation of pedotransfer functions. Information on both water retention and hydraulic conductivity was available for 1136 horizons whereas 2894 horizons had only information on water retention. Each soil horizon was allocated to one of 11 possible soil textural/pedological classes derived from the six FAO texture classes (five mineral and one organic) and the two pedological classes (topsoil and subsoil) recognised within the 1:1 000 000 scale Soil Geographical Data Base of Europe. Next, both class and continuous pedotransfer functions were developed. By using the class pedotransfer functions in combination with the 1:1 000 000 scale Soil Map of Europe, the spatial distribution of soil water availability within Europe was derived.
Article
The leaching of the pesticide isoproturon from the macroporous clay soil at Brimstone Farm was modelled using four alternative models (MACRO, CRACK-NP, SIMULAT and PLM). Model results are presented for two test periods, the whole of one winter for which daily observations are available, and a short subset for which hourly data were presented. The best results are those given by MACRO with an expert user, although satisfactory results were also obtained from CRACK-NP and for the longer test period by PLM. SIMULAT was less successful in modelling the site because it did not include an adequate representation of the site hydrology, it was unable to predict the leaching of pesticide. MACRO was also used by a second modelling group who were less familiar with both the code and the site. Although the initial uncalibrated runs from this group were poor, the final calibrated results were almost as good as those derived by the ‘expert’ user. The simulations showed the difficulty of deriving adequate representations, even where relatively complete soil physical data are available. A shortcoming of the dataset provided was the lack of detailed soil moisture observations, particularly to define the initial conditions. From a well-monitored site, many observations of site hydrology (water table position, drainflow and surface flow) were available, but significantly, fewer pesticide concentrations in either the soil or the discharges were available. Models could thus be evaluated only in terms of their ability to predict the magnitude and timing of major pesticide leaching events.
Article
Field and laboratory tests were conducted to investigate the disappearance and decrease in availability of the herbicide isoproturon in soil with time. Monitoring the fate of 14C-labeled isoproturon gave the persistence of the herbicide under different conditions. The calculated half-life (DT50) of the extractable parent product was close to 19 days for the two experiments, whereas half-lives of exhaustive extractable 14C-residues averaged 22 days under controlled conditions versus 48 days in the field. Moreover, if nonextractable 14C-residues were taken into account in estimating isoproturon fate, the DT50 of radioactive residues in soil (extractable parent and degradation products + nonextractable residues) reached 1 to 1.4 years under field and controlled conditions, respectively. The variability of disappearance rates for the different kinds of residues is discussed within a wider definition of persistence.
Article
The uncalibrated predictive ability of four preferential flow models (CRACK-NP, MACRO/MACRO_DB, PLM, SWAT) has been evaluated against point rates of drainflow and associated concentrations of isoproturon from a highly structured and heterogeneous clay soil in the south of England. Data were available for four plots for a number of storm events in each of three successive growing seasons. The mechanistic models CRACK-NP and MACRO generally gave reasonable estimates of drainflow over the three seasons, but under-estimated concentrations of isoproturon over a prolonged period in the first season and over-estimated them in the two remaining seasons. CRACK-NP simulated maximum concentrations of isoproturon over the first two events of each of the three seasons of 156, 527 and 24.4 micrograms litre-1, respectively, and matched the observed data (465, 65.1 and 0.65 micrograms litre-1) slightly better than MACRO (69.1, 566 and 58.5 micrograms litre-1). Automatic selection of parameters from soils information within MACRO_DB reduced the emphasis on preferential flow relative to the stand-alone version of MACRO. This gave a poor simulation of isoproturon breakthrough and simulated maximum concentrations were 0, 50.1 and 35.1 micrograms litre-1, respectively. The capacity model PLM gave the best overall simulation of total drainflow for the first two events in each season, but over-estimated concentrations of isoproturon (967, 808 and 51.3 micrograms litre-1). The simple model SWAT represented total drainflow reasonably well and gave the best simulation of maximum isoproturon concentrations (140, 80.2 and 8.2 micrograms litre-1). There was no clear advantage here in using the mechanistic models rather than the simpler models. None of the models tested was able to simulate consistently the data set, and uncalibrated modelling cannot be recommended for such artificially drained heavy clay soils.
Article
Sensitivity analyses for the preferential flow model MACRO were carried out using one-at-a-time and Monte Carlo sampling approaches. Four different scenarios were generated by simulating leaching to depth of two hypothetical pesticides in a sandy loam and a more structured clay loam soil. Sensitivity of the model was assessed using the predictions for accumulated water percolated at a 1-m depth and accumulated pesticide losses in percolation. Results for simulated percolation were similar for the two soils. Predictions of water volumes percolated were found to be only marginally affected by changes in input parameters and the most influential parameter was the water content defining the boundary between micropores and macropores in this dual-porosity model. In contrast, predictions of pesticide losses were found to be dependent on the scenarios considered and to be significantly affected by variations in input parameters. In most scenarios, predictions for pesticide losses by MACRO were most influenced by parameters related to sorption and degradation. Under specific circumstances, pesticide losses can be largely affected by changes in hydrological properties of the soil. Since parameters were varied within ranges that approximated their uncertainty, a first-step assessment of uncertainty for the predictions of pesticide losses was possible. Large uncertainties in the predictions were reported, although these are likely to have been overestimated by considering a large number of input parameters in the exercise. It appears desirable that a probabilistic framework accounting for uncertainty is integrated into the estimation of pesticide exposure for regulatory purposes.
Article
Laboratory studies were conducted to determine the sorption behaviour of six commonly used pesticides (acetochlor, atrazine, carbendazim, diazinon, imidacloprid and isoproturon) on Hungarian brown forest soil with clay alluviation (Luvisol) using the batch equilibrium technique. The sorption isotherms could be described by the Freundlich equation in non-linear form (n < 1) for all compounds, however in case of diazinon using the extended Freundlich equation proved to be a better approach. The adsorption constant related soil organic carbon content (Koc) calculated from Freundlich equation were 314 for acetochlor, 133 for atrazine, 2805 for carbendazim, 1589 for diazinon, 210 for imidacloprid and 174 for isoproturon. The octanol-water partition coefficients (Pow), which can be a useful parameter to predict of adsorption behaviour of a chemical on soil, and dissociation coefficients of these pesticides were calculated based on the chemical structure of them using a computerized expert system. The octanol-water partition coefficients were determined experimentally from high performance liquid chromatographic parameters as well. Good agreement was observed between experimental and the computer expert system estimated data. Computer estimated log Pow values ranged 0.5 and 3.86 for the examined pesticides, with imidacloprid and diazinon being the least and most hydrophobic respectively. Experimentally determined logPow ranged between 0.92 and 3.81 with the same tendency. It can be concluded that the Freundlich adsorption constants (Kf) are slightly related to the octanol-water partition coefficients of investigated chemicals, nevertheless no close correlation could be established because of the influence of further characteristics of solutes and soil.
Article
Calibration of pesticide leaching models may be undertaken to evaluate the ability of models to simulate experimental data, to assist in their parameterisation where values for input parameters are difficult to determine experimentally, to determine values for specific model inputs (e.g. sorption and degradation parameters) and to allow extrapolations to be carried out. Although calibration of leaching models is a critical phase in the assessment of pesticide exposure, lack of guidance means that calibration procedures default to the modeller. This may result in different calibration and extrapolation results for different individuals depending on the procedures used, and thus may influence decisions regarding the placement of crop-protection products on the market. A number of issues are discussed in this paper including data requirements and assessment of data quality, the selection of a model and parameters for performing calibration, the use of automated calibration techniques as opposed to more traditional trial-and-error approaches, difficulties in the comparison of simulated and measured data, differences in calibration procedures, and the assessment of parameter values derived by calibration. Guidelines for the reporting of calibration activities within the scope of pesticide registration are proposed.
Article
Although macropore flow is recognized as an important process for the transport of pesticides through a wide range of soils, none of the existing spatially distributed methods for assessing the risk of pesticide leaching to groundwater account for this phenomenon. The present paper presents a spatially distributed modelling system for predicting pesticide losses to groundwater through micro- and macropore flow paths. The system combines a meta version of the mechanistic, dual porosity, preferential flow pesticide leaching model MACRO (the MACRO emulator), which describes pesticide transport and attenuation in the soil zone, to an attenuation factor leaching model for the unsaturated zone. The development of the emulator was based on the results of over 4000 MACRO model simulations. Model runs describe pesticide leaching for the range of soil types, climate regimes, pesticide properties and application patterns in England and Wales. Linking the MACRO emulator to existing spatial databases of soil, climate and compound-specific loads allowed the prediction of the concentration of pesticide leaching from the base of the soil profile (at 1 m depth) for a wide range of pesticides. Attenuation and retardation of the pesticide during transit through the unsaturated zone to the watertable was simulated using the substrate attenuation factor model AQUAT. The MACRO emulator simulated pesticide loss in 10 of 12 lysimeter soil-pesticide combinations, for which pesticide leaching was shown to occur and also successfully predicted no loss from 3 soil-pesticide combinations. Although the qualitative aspect of leaching was satisfactorily predicted, actual pesticide concentrations in leachate were relatively poorly predicted. At the national scale, the linked MACRO emulator/AQUAT system was found to predict the relative order of, and realistic regional patterns of, pesticide leaching for atrazine, isoproturon, chlorotoluron and lindane. The methodology provides a first-step assessment of the potential for pesticide leaching to groundwater in England and Wales. Further research is required to improve the modelling concept proposed. The system can be used to refine regional groundwater monitoring system designs and sampling strategies and improve the cost-effectiveness of the measures needed to achieve 'good status' of groundwater quality as required by the Water Framework Directive.
Article
Field monitoring and scenario-based modelling were used to assess exposure of small ditches in the UK to the herbicide sulfosulfuron following transport via field drains. A site in central England on a high pH, clay soil was treated with sulfosulfuron, and concentrations were monitored in the single drain outfall and in the receiving ditch 1 km downstream. Drainflow in the nine months following application totalled 283 mm. Pesticide lost in the first 12.5 mm of flow was 99% of the total loading to drains (0.5% of applied). Significant dilution was observed in the receiving ditch and quantifiable residues were only detected in one sample (0.06 microg litre(-1)). The MACRO model was evaluated against the field data with minimal calibration. The parameterisation over-estimated the importance of macropore flow at the site. As a consequence, the maximum concentration in drainflow (2.3 microg litre(-1)) and the total loading to drains (0.76 g) were over-estimated by factors of 2.4 and 5, respectively. MACRO was then used to simulate long-term fate of the herbicide for each of 20 environmental scenarios. Resulting estimates for concentrations of sulfosulfuron in a receiving ditch were weighted according to the prevalence of each scenario to produce a probability distribution of daily exposure.
Article
The leaching model PESTRAS was used to estimate sorption and degradation values for bentazone from three lysimeter datasets using the inverse modelling package PEST. Investigations were undertaken to assess the influence on calibration results of (1) values attributed to uncertain parameters not included in the calibration, and (2) starting values supplied to the inverse modelling package. Automatic calibrations with different realistic values for the Freundlich exponent n(f) yielded different combinations of K(om) and DT50. Similarly, the supply of different starting values for K(om) and DT50 revealed that different combinations of these two parameters equally calibrated PESTRAS for two of the three lysimeters. Examination of the error surface, ie the forward running of the model for different combinations of K(om) and DT50 values, and the calculation of the goodness-of-fit to the experimental data, was found useful for identifying those instances where non-uniqueness in the calibration is likely to occur. Although the derivation of sorption and degradation values through inverse modelling is expected to offer significant benefits over laboratory determinations, care should be exercised when examining values derived through this approach. Research is needed to identify data requirements for robust estimation of sorption and degradation parameters through calibration of pesticide fate models against leaching data.
Article
Testing of pesticide leaching models against comprehensive field-scale measurements is necessary to increase confidence in their predictive ability when used as regulatory tools. Version 5.1 of the MACRO model was tested against measurements of water flow and the behaviour of bromide, bentazone [3-isopropyl-1H-2,1,3-benzothiadiazin-4(3H)-one-2,2-dioxide] and imidacloprid [1-(6-chloro-3-pyridylmethyl)-N-nitroimidazolidin-2-ylideneamine] in a cracked clay soil. In keeping with EU (FOCUS) procedures, the model was first calibrated against the measured moisture profiles and bromide concentrations in soil and in drain water. Uncalibrated pesticide simulations based on laboratory measurements of sorption and degradation were then compared with field data on the leaching of bentazone and imidacloprid. Calibrated parameter values indicated that a high degree of physical non-equilibrium (i.e. strong macropore flow) was necessary to describe solute transport in this soil. Comparison of measured and simulated bentazone concentration profiles revealed that the bulk of the bentazone movement in this soil was underestimated by MACRO. Nevertheless, the model simulated the dynamics of the bentazone breakthrough in drain water rather well and, in particular, accurately simulated the timing and the concentration level of the early bentazone breakthrough in drain water. The imidacloprid concentration profiles and its persistence in soil were simulated well. Moreover, the timing of the early imidacloprid breakthrough in the drain water was simulated well, although the simulated concentrations were about 2-3 times larger than measured. Deep groundwater concentrations for all substances were underestimated by MACRO, although it simulated concentrations in the shallow groundwater reasonably well. It is concluded that, in the context of ecotoxicological risk assessments for surface water, MACRO can give reasonably good simulations of pesticide concentrations in water draining from cracking clay soils, but that prior calibration against hydrologic and tracer data is desirable to reduce uncertainty and improve accuracy.
The MACRO Model (Version 4.3) – Technical Description Available: http://www-mv.slu.se/bgf/Macrohtm
  • Nj Jarvis
26 Jarvis NJ, The MACRO Model (Version 4.3) – Technical Description. [Online]. Available: http://www-mv.slu.se/bgf/Macrohtm/ macro43b/ TechMacro43.pdf [November 2006].
29 FOCUS, Surface Water Scenarios in the EU Evaluation Process under 91 Development and use of a database of hydraulic properties of European soils
  • Y Mualem
  • Jhm Lilly
  • A Nemes
  • A Bas
Mualem Y, A new model for predicting the hydraulic conductivity of unsaturated porous media. Water Resour Res 12:513–522 (1976). 29 FOCUS, Surface Water Scenarios in the EU Evaluation Process under 91/414/EEC. EC Document Reference SANCO/4802/2001-rev. 2 (2001). 30 W ¨ osten JHM, Lilly A, Nemes A and Bas CL, Development and use of a database of hydraulic properties of European soils. Geoderma 90:169–185 (1999).
ContributionàContribution`Contributionà l'amélioration de l'´ evaluation du risque a priori pour les eaux superficielles par les phytosanitaires: soutienàsoutien`soutienà la construction de scénariossc´scénarios nationaux pour l'´ evaluation des préparations
  • N Carluer
  • Ig Dubus
  • V Gouy
  • Jj Gril
  • N Surdyk
Carluer N, Dubus IG, Gouy V, Gril JJ and Surdyk N, ContributionàContribution`Contributionà l'amélioration de l'´ evaluation du risque a priori pour les eaux superficielles par les phytosanitaires: soutienàsoutien`soutienà la construction de scénariossc´scénarios nationaux pour l'´ evaluation des préparations.
MACRO 5.0. A model of water flow and solute transport in macroporous soil. Technical description. Swedish University of Agricultural Sciences Emergo
  • M Larsbo
  • Jarvis Nj
Larsbo M and Jarvis NJ, MACRO 5.0. A model of water flow and solute transport in macroporous soil. Technical description. Swedish University of Agricultural Sciences Emergo 20-03 : 6 Report (2003).
and six other computer codes for universal sensitivity analysis, calibration, and uncertainty evaluation
  • Ep Poeter
  • Mc Hill
  • Er Banta
  • S Mehl
  • S Christensen
  • Ucode
Poeter EP, Hill MC, Banta ER, Mehl S and Christensen S, UCODE 2005 and six other computer codes for universal sensitivity analysis, calibration, and uncertainty evaluation. US Geological Survey Techniques and Methods 6-A11 (2005).
Report on Soil Profile Characteristics and Topsoil Structure at the Arvalis Experimental Site
  • G Beard
23 Beard G, Report on Soil Profile Characteristics and Topsoil Structure at the Arvalis Experimental Site, LaJaillì ere, France. [Online]. Available: http://www.landlook.fsnet.co.uk/default.htm [September 2006]. 24 FOOTPRINT Pesticide Properties Database. [Online]. Available: http://www.herts.ac.uk/aeru/footprint/en/ [July 2007].
Approach of degradation ways of two selected substituted ureas, isoproturon and chlortoluron in the water/sediments system of a swamp's drainage ditches
  • L Patria
  • M Griseau
  • Le Brun
  • O Merlet
Patria L, Griseau M, Le Brun O, Merlet N and Dore M, Approach of degradation ways of two selected substituted ureas, isoproturon and chlortoluron in the water/sediments system of a swamp's drainage ditches. J Eur d'Hydrol 26:101–120 (1994).
A screening model to estimate pesticide concentrations in drinking water
  • First Usepa
Contribution à l'amélioration de l'évaluation du risque a priori pour les eaux superficielles par les phytosanitaires: soutien à la construction de scénarios nationaux pour l'évaluation des préparations
  • V Carluer N Dubus Ig Gouy
  • J J Gril
  • Surdyk
Approach of degradation ways of two selected substituted ureas, isoproturon and chlortoluron in the water/sediments system of a swamp's drainage ditches
  • Patria
UCODE_2005 and six other computer codes for universal sensitivity analysis, calibration, and uncertainty evaluation
  • E P Poeter
  • E R Hill Mc Banta
  • S Mehl
  • Christensen
  • Hill