Hello Dr. Shurpali:

Thank you for posting your DSSAT data question to the listserver.   I think
that you will find that not only DSSAT, but other modeling systems have
similar data requirements, some of them more stringent than DSSAT.  Note
that the IBSNAT and DSSAT development group has been at the forefront of
defining a minimum data set for running crop models.  In general we also
have the concept that if your input data are of poor quality, the results
of the model will be poor (garbage in ==> garbage out).

I realize that these answers do not help you very much. Model users in both
developing and developed countries are faced with similar problems of how
to define the inputs for the model.  Methods have been developed to
estimate soil parameters from general soil data bases.  Similarly weather
generators are being used to generate weather data based on climate
information if detailed weather data are not available.  Andre mentioned
that he is developing methods to estimate genetic coefficients based on
variety trial data.  Others are involved in similar projects.

You might be surprised how much local data is available for model
calibration and evaluation.  It might require that you have to talk to your
local agronomist or other scientist involved in field work.  There is not
always a need to start a new trial if data have already been collected in
the past.  Modeling is a systems approach, requiring collaboration of
scientists of different disciplines.  I have been pleasantly surprised
about the amount of detailed data that are available in India.  ICRISAT has
been involved in establishing modeling networks in India for various crops,
including peanut and sorghum.  I highly recommend that you locate the
publications by Dr. Piara Singh, Dr. Virmani, Dr. Alagarswamy and others
who have been at the forefront of evaluating the DSSAT models for India and
developing applications for India.  Related to that, please note that many
resources  are already available, including the papers published about DSAT
and the CROPGRO/CERES and other models as well as the book "Understanding
Options for Agricultural Production."  Others have faced similar problems
you have identified in your e-mail and some solutions have been found.

Again, thank you for posting your question.  High quality input data
continuous to be one of the limiting factors for practical application of
the crop models.

Regards,

Gerrit

=====================================================================
Gerrit Hoogenboom
Professor
Department of Biological and Agricultural Engineering
The University of Georgia
Griffin, Georgia 30223-1797, USA

Phone: 770-229-3438
Fax:     770-228-7218
[log in to unmask]
www.georgiaweather.net
www.gerrithoogenboom.com
=====================================================================