Modelling Soil Organic Matter Dynamics – Global Challenges - Sustainable Management of Soil Organic Matter - page 69

F temp =( T / T opt ) 2 , T ≤ T opt (3b)

F moist =1 − ( Σ E os )/ E os, max , ( Σ E os )< E os, max (3c)

F contact = 1.54 − 0.36 R , 1.5 < R ≤ 3 (3d)

where j is the coefficient determining the slope of Equation 3a, T is the
average daily air temperature, E os is the potential daily soil evaporation
and the subscripts 'opt' and 'max' refer to the optimum and maximum
parameter values, respectively. Outside the given limits, the factor values
are set equal to 1 or 0, except for F contact which has a minimum ( F contact, min )
of 0.46 in the default values (Table 2.4.1).
For parameterization, APSIM-Residue was programmed into a
spreadsheet and the inputs for calculating the factors from Equation 3, i.e.
residue C : N, daily air temperature and soil evaporation, were taken from
measured data. Soil evaporation data were derived from daily water balance
calculations performed on the soil moisture data from the experiments.
Using the input data, daily values of k were calculated from Equation 2, and
daily residue mass from Equation 1. Optimal values of r max , T opt , E os, max ,
F contact, min were determined by the generalized reduced gradient optimiza-
tion method of Fylstra et al . (1998) to minimize an objective function, D
(Table 2.4.1). F C:N was not considered explicitly (i.e. was not included in
Equation 2) in the optimization since it is constant once calculated from
initial residue C : N, which did not vary greatly (75-95) between the three
sites. Instead, the optimal value of r max included the C : N limitations to
decomposition, which were assumed to be uniform across the experiments.
To derive more general parameter values from the optimization, values
of r max , T opt , E os, max and F contact, min were optimized for both the Mackay
Early and Mackay Late experiments, i.e. to minimize the sum of the
objective function from the two experiments. To provide some test of their
generality, the optimal parameter values were used to predict residue
decomposition at the third site, Harwood Early.

Table 2.4.1. Parameter values and resultant values of the objective function (D) in each
experiment for two parameterization schemes, the default parameters (Probert et al., 1998b;
which included those for F C:N , not shown here) and the result of the optimization.

Parameter value Values of D a

Scheme r max E os, max T opt F contact, min Mackay Early Mackay Late Harwood Early

Default 0.100 20 20 0.46 100.3 384.9 129.0

Optimal 0.007 9 23 1 .46 10.2 21.4 18.8

a D= Σ (|R p − R o | + 1) 2 , where the subscripts p and o refer to predicted and observed values,
respectively.