Ghamari,

Here are some ideas.  You calibrated HWAM based on P5 (set to MDAP, after setting P1 and P2 to get to ADAP) plus G2 and G3.  So, P5, G2, and G3 are dominantly what affects HWAM, and CWAM is additionally affected by RUE (which you should not touch) and SLPF (soil fertility value) in the soil's file.  You could reduce the SLPF in soil file from 1.00 (best possible soil) to a lower value, to reduce the RMSE and improve fit to CWAM.  This "might" move LAIX??? and should also affect HWAM as well (but that latter can always be "re-calibrated" by setting G2 and G3 again).

As for LAIX, it is not so easy.  CERES-Maize functions for LAI are not very flexible, and are dominantly driven by what you simulate for ADAP and leaf number, although leaf number "could" be modified by changing PHINT.  So, SLPF and PHINT are your only available coefficients left to change to improve fit to CWAM and LAIX, but change to those, would require re-calibration of G2 and G3 again, to get good fit of HWAM.

Why are you turning off water and N?  That is not particularly a safe thing, if you are trying to mimic a "real" field yield.  This last bit of advice may be more critical then all of the so-called recalibrations to genetic coefficients.

Good luck,
Ken Boote

-----Original Message-----
From: DSSAT - Crop Models and Applications [mailto:[log in to unmask]] On Behalf Of mohsen
Sent: Thursday, January 29, 2009 3:09 AM
To: [log in to unmask]
Subject: CERES-Maize calibration

dear all
I have a calibrated CERES-Maize model under potential condition (water & nitrogen are swich off), that simulated adjectives such as ADAP, MDAP and HWAM gives the least values of RMSE, but simulated LAIX and CWAM are less than observad values and gives excessive value of RMSE.
 How can i fix this discord?

Best regards
ghamari