Page 81 - Modelisation du devenir des pesticides...

3.2 Materials and methods

pesticide fate models and their Use (FOCUS). 40,41 The objectives of this paper are
i) to present the development of the pesticide module in the crop model STICS and
ii) to evaluate the capability of the model to simulate the fate of three herbicides
using different field datasets and model comparisons. The evaluation is based on a
non-calibration approach in order to illustrate the ability of the model to predict
the pesticide behavior with the measurements or pesticide properties available in
a database. 42 After a brief description of the model basis, the implementation of
the pesticide scheme is described. Results are discussed on the basis of observation
data and compared to the performance of previous pesticide models previously run
for the three study sites.

3.2 Materials and methods

3.2.1 The STICS agronomic model
STICS is a daily time step crop model that uses climate, soil, and agricultural
practices data to simulate yield and environmental variables such as carbon or ni-
trate flux. 5,43,44 Soil evaporation is calculated in two steps : potential evaporation
related to the energy available at the soil level and then actual evaporation rela-
ted to water availability. It is then distributed over the soil profile. The potential
evaporation is related to plant cover above the ground (using the Leaf Area Index
LAI) and can be either calculated according to a Beer Law equivalent approach,
applied to the potential evaporation/reference evapotranspiration ratio (Penman)
with a constant extinction coefficient, or computed by a simplified daily energy
balance approach. The partition of actual evapotranspiration between transpira-
tion and bare soil evaporation is then computed based on LAI. The plant uptake
is distributed in the soil profile based on the root depth and density. 45,46 The soil
profile is simulated using 1cm elementary layers. The soil water transfer is based
on a reservoir analogy. The net infiltration fills the reservoir, and a downward flux
occurs as soon as the field capacity is reached. The drainage flow is assumed to be
fast enough for the soil moisture to be usually limited to the field capacity, except
when some options are activated. 46 Soil properties (wilting point, field capacity
and soil density) may not be vertically uniform, with at most 5 different values at
given soil horizons. Crop and soil temperatures are key variables as they are invol-
ved both in the plant growth and yield, and in the mineralization of organic matter.
The crop temperature is determined either by a simple energy balance method or
by an empirical approach, both approaches being based on an estimation of the
sensible heat flux from the net radiation and evaporation. 47 The soil temperature
is then computed using the mean and amplitude of the daily crop temperature.
The distribution of soil temperature with depth is based on the formalization of

69