Thanks a lot for all who responded me regarding to wheat phenology used in DSSAT3 models. To correct some thoughts I found that I have to write a bit more first of all I found the model simulation amazing regarding to all experiment data I have I found a very good accurate agreement between measured and simulated data. To make my problem more clear: As someone also adviced me to translate the Zadoks codes to Ceres scale, actually I have no other way but it is involved with some errors and will affect on my result and maybe anything we publish as our work, so I thought that may be some others faced such a thing as me and found a good way Anyway, to forecast winter wheat yield through Mrkov Chain method I used a multiple-run of the model to find the stochastic transition probabilities, or the rate of change of the system from one stage of life ( phenology ) to the next one, in this way at any phenology stage we can have a stochastic estimation of different probabilities of yield direction towards the next development stage. I finished that but when I got back to see how this forecasting chart behave under the real data, ( validating ) I found that the succeeding phenology stages used in the model is different from what I have as measured values, so to validate this forecasting method I have to find that what are exact equal DSSAT Wheat phenology scales in terms of Zadoks. Anyway, thanks for all directions. Best Regards Mohammad Bannayan