In an agricultural ecosystem, soil temperature can affect the growth o
f plants and organisms, the fate and transport of chemicals, and many
other natural phenomena. Simulation of soil temperature is essential t
o support many agricultural models. Modeling the fluctuations of soil
temperature at different depths is complicated considering the great n
umber of variables. In this study, a simple model, based on an artific
ial neural network (ANN), was developed to simulate daily soil tempera
tures at 100, 500 and 1500 mm depths, in a soil from Ottawa, Ontario,
Canada, in an attempt to develop a simple, fast, and more accurate ANN
model than the conceptual models currently used to simulate soil temp
erature. The inputs for the ANN model included: daily rainfall, potent
ial evapotranspiration, maximum and minimum air temperature, and the d
ay of the year These input factors are all easy to obtain and are meas
ured at most weather stations world-wide. The parity between the measu
red and the simulated data, resulting from ANNs, shows the ability of
simple ANN models to simulate soil temperature. The results obtained f
rom ANN models varied within a root-mean-square difference range from
0.63 to 1.39 degrees C, standard deviations from 0.61 to 1.39 degrees
C and coefficients of determination (r(2)) from 0.937 to 0.985. The ac
curacy of the simulations shows the simplicity with which ANNs can be
used to model complicated phenomena in agricultural systems. The short
time of execution (a few seconds for a one-year simulation) is anothe
r benefit of ANN models. Many simulation models, such as for pesticide
fate and transport, nutrient movement in soils, and soil bioremediati
on, require timely fluctuations of soil temperatures. For such uses, t
he fast execution of ANNs is very helpful. Therefore, this technology
could prove very useful for decision support systems which require rea
l-time control in agricultural applications.