Validation of GOSSYM: Part II. Mississippi conditions☆

Abstract Validation of the cotton simulation model GOSSYM was carried out with data collected from two locations during three different years in Mississippi to: (1) identify areas of the model which need to be improved to provide a universal cotton simulation model, (2) develop versions of GOSSYM with site-dependent or cultivar-dependent parameters as necessary for immediate use. Equations to calculate canopy temperatures were developed from data collected at Mississippi State Plant Science Farm, and carbohydrate partitioning logic was changed in the model giving high priority for bolls under carbohydrate stress conditions. The GOSSYM simulations were very close to real data collected during two different seasons at Mississippi State. For validating GOSSYM with data from Stoneville, Mississippi, we incorporated lygus damage into the simulated crop to give a realistic comparison with the real crop. We believe the data support a lygus population in the field plots of sufficient numbers to cause damage equal to the fruit loss we put into the simulation. Changes during this validation and the reasons for them are discussed.