I have experienced this error also and it may be related to the fact
that you have a very poor environment for growth.

        First of all there is a 'new' CERES3.5  that handles drought better.
Contact Paul Wilkens. This version makes sure that if a crop emerges and
consequently there is no rain for growth that the model stops after 15 days.
Previous versions allowed an emerged crop to sit an wait until rain came.
This means a seedling could sit a 4 month period and wait for water....

        Once this problem was overcome we found that the models can show
negative >growth.  The model checks for negative biomass, but not some of
the other variables. > As a result we had some negative variables such as
Nitrogen! You can see this in the output files.
        Dr. Hoogenboom has a new version of CROPGRO that does not allow
this. For CERES unfortunately there is no solution yet.

        I would advise you just to check first how hostile your environment
is for growth and get into contact with either Dr Wilkens at IFDC or Dr
Hoogenboom at University of Georgia.

        Dewi Hartkamp
        CIMMYT Mexico



        ----------
        From:  Michael Pool [SMTP:[log in to unmask]]
        Sent:  Thursday, April 27, 2000 10:13 PM
        To:  [log in to unmask]
        Subject:  Run-time Error

            When I try to run an experiment I get the following message:

        run-time error M6103: Math
        floating point error: divide by zero

        I have not been able to figure out what the problem is.  Does anyone
have a
        suggestion?

        --
        Michael Pool, RPA
        Anthropology/Geography Department
        Austin Community College
        [log in to unmask]