Comparison of 1D and 2D modelling of pesticide transfer in a tiledrained context. Application to la Jaillière site.
Dairon R.1,2*, Carluer N.,1 Dutertre A.,2 Réal B.2 and Leprince F.2
1
Irstea, UR Maly, 5 rue de la doua, Villeurbanne, France.
Arvalis-institut du végétal, 3 rue Joseph et Marie Hackin 75016, Paris, France.
*[email protected]
2
Introduction
One dimensional physically based models are widely use as regulatory tools for pesticide
leaching in the E.U. However, in the particular case of drained soils, water and solute flow
are mainly influenced by the 2D-spatial dynamics of the drained water-table (Paris, 2004).
On the other hand, preferential flow is often invoked (Jarvis, 2007) to explain the observation
of quick pesticide transfer to the drain. The main objective of this study is to evaluate the
role of the drained water table and preferential flow in pesticide export, comparing a one
dimensional model and a two dimensional model.
Material and methods
The la Jaillière experimental station is located in the western part of France (47°27' N, 0°57'
W). The site has been monitored for pesticides since 1994 and provides an important data
set on pollutant transfers. It is currently used as a drainage scenario for European pesticide
registration. The soils are predominantly stagnic luvisol, occurring on a gently sloping
plateau. The fields are tile drained according to French criteria (0.9 m deep and 10-12 m
drain spacing). The study is focused on one plot of the domain, called “T4”, whose
characteristics are provided in Table 1.
Table 1 : Main soil characteristics of the T4 plot.
Horizon
Thickness
Ap
30
E
18
Bt
17
Bt/C
45
a : organic matter content
Clay (%)
Silt (%)
Sand (%)
O.M (%)
20.8
25.9
49.2
42.7
44.6
41.3
35.3
35.8
34.6
32.8
15.5
21.5
2.17
0.77
0.46
0.36
a
Bulk
density
(g.cm3)
pH (water)
Structure
1.55
1.63
1.7
1.7
6.3
7
5.6
4.9
Blocky
Blocky
Prismatic
Blocky
The fate of two substances with widely differing characteristics was investigated. The first,
isoproturon, is moderately sorbed (Koc : 124 L.Kg-1) and weakly persistent in soils (DT50 : 15
Days) whereas the second, diflufenican has a high sorption capacity (Koc : 2000 L.Kg-1) and
is slightly persistent in the environment with DT50 exceeding 140 days. Both substances are
applied during autumn on winter-wheat.
The two models tested in this study were MACRO 5.2 (Larsbo and Jarvis, 2003) and
HYDRUS-2D (Simunek et al., 1999). MACRO is a one dimensional physically based model
with the soil porosity divided into two flow domains (macropores and micropores). Richard’s
equation and the classic convection-dispersion equation (CDE) are used in the matrix, for
water and solute transport, respectively. In the macropore domain, a kinematic wave
approach is used for water flow and convection for solute transport. HYDRUS-2D is a two
dimensional physically based model which allows users to choose from a range of hydraulic
models (e.g single porosity, mobile-immobile water). In order to investigate preferential flow
impact on pesticide transport, we decided to test a single porosity model and a dual
permeability model developed by Gerke and van Genuchten (1993) using Richard’s
equationand the CDE in both domains of porosity.
According to EU procedures (FOCUS, 2001), models were first calibrated against water flow
and bromide for which hourly data sets are available. Bromide has been use in several
studies to investigate and calibrate the mass transfer coefficient (e.g. Stenemo and Jarvis
2007; Kohne et al., 2006) and other parameters (tortuosity, disperivity, etc) which were
evaluated here.
Results
Results show that both models underestimated drainage during the intensive drainage
season (IDS) without appropriate calibration. Another difficulty of the site is to represent the
role of the groundwater table on drainage intensity and depletion. Simulation of bromide
shows discrepancies between models, MACRO can’t’ represent simultaneously the dynamic
of both water and bromide as shown on Fig. 1. Thus, the mass transfer coefficient can not be
estimated with certainty. On the other hand, HYDRUS-2D represents the water and solute
dynamics well for both options. Consequences for pesticide export are assessed using the
mass transfer coefficient calibrated with HYDRUS-2D and MACRO.
Figure 1 : Measured and simulated drain flow and bromide concentration by MACRO, for maximum
tortuosity in all horizons.
References
Adamiade CV (2004). Influence d'un fossé sur les écoulements rapides au sein d'un bassin
versant. Application au transfert des produits phytosanitaires., PhD thesis. Université Paris
VI.
Gerke HH, Van Genuchten MT (1993). A Dual-Porosity Model for Simulating the Preferential
Movement of Water and Solutes in Structured Porous-Media. Water Resources Research
29:305-319.
Jarvis NJ (2007). A review of non-equilibrium water flow and solute transport in soil
macropores: principles, controlling factors and consequences for water quality. European
Journal of Soil Science 58:523-546.
Kohne JM, Kohne S et al. (2006). Multi-process herbicide transport in structured soil
columns: Experiments and model analysis. Journal of Contaminant Hydrology 85:1-32.
Larsbo M, Jarvis N (2003). MACRO 5.0. A model of water flow and solute transport in
macroporous soil, Swedish University of Agricultural Sciences.
Simunek J, Sejna M et al. (1999). The HYDRUS-2D software package for simulating the twodimensional movement of water, heat, and multiple solutes in variably-saturated media,
International Groundwater Modeling Center, Riverside, California: 227 pp.