Hi there;
Please do not get me wrong and try to Re-read my other mail.
The basic rule as everybody knows, one should split the data in
two groups: one for calibration and one group for evaluation.
Calibration the DSSAT models, Derek needs at least to run the
GENCALC but according to what he mentioned about his data,
looks he doesnot posses enough data to run the GENCALC, that
is why he tried to use the closest cultivar he thinks shows similar
behaviour to his wheat cultivar and no wonder he got wrong results.
In this case I suggested him to run the model and 'play' with coeff.
till he get the best similarity between his simulations and
observations (the first group of data). After this step, he would be
able to evaluate the model with the rest of his data no matter, if his
data are from another experiment at some other year/s or at any
other site. So nobody suggeted him to calibrate the model for each
year (or site) of experiment. Once you calibrate the model, it
should be able to be used for other years and/or sites as long as
you use the same cultivar.
I cannot see any danger in this approach.
Best
Mohammad


> Dear Derek,
>
> I think, the proposal from Mohammad is dangerous. Because, when you change or
> 'play' with the genetic coeffients (the genetic coeffiients are constant and
> independent from the special environment), your are adapting the model about
> these coeffients to your special environment conditions. That means in
> reality, the model doesn't work with the input parameters of your environment
> (weather, soil, production techniques etc.).
> Securely, you can have success with this model, in case you determine the
> genenetic coeffients for each of yours production fields. But this - from
> scientific view - does not satisfie and is a much of work - from practice
> side - .
> What is your target? To predict yields or to optimise fertilising ? May be, I
> can help with the long term model N-Prog (s. www.n-prog.de).
>
> Please tell me and all the best
>
> Peter
> ---------------------------------------------------------
> Schoop, Peter, Dr. sc. agr., Dipl. Ing.
> Institut of Agri- and Environment Research
> Sonnenweg 1, 21647 Moisburg
> Germany
> Phone: +49-4165-224 700
> Fax:     +49-4165- 224 701
> E-Mail: [log in to unmask]
> Internet: www.n-prog.de
> ---------------------------------------------------------