Dear CMAQ users,
I met a problem when I running CMAQ. The information of error as listed below.
H-eddy DT & integration steps: 7.2000000E+02 1
after HDIFF G 1.3190703E+03 A 2.4345098E+12 N 3.9007351E+00
after DECOUPLE_ G 1.2988125E-01 A 1.8667142E+08 N 3.9435661E-04
after CLDPROC G 1.2989776E-01 A 1.8669941E+08 N 3.9408138E-04
after CHEM G 1.2988505E-01 A 1.8669941E+08 N 3.9408138E-04
*** ERROR ABORT in subroutine Soa_Defn on PE 007
*** Problem in NEWT at Layer = 5
PM3EXIT: date&time specified as 0
Date&time specified as 0
I will appreciate a lot for all suggestions.
What version of CMAQ are you using? The Newton-Raphson method in the NEWT subroutine was replaced with a bisection method in 2011. Did you modify the code to set newtpartition to .TRUE.?, and if so, why?
The version of CMAQ is 5.0.1.
I am not sure. Where can I check the newpartition?
The comments in the current version of SOA_DEFN.F indicate the new partitioning scheme was coded in July 2011, but this version was apparently not released until CMAQv5.1 in November 2015. The algorithm is documented at the bottom of this page, just before the References.
CMAQv5.0.1 came out in July 2012. There have been six releases of CMAQ since then. You are strongly encouraged to use an updated version of the model. If you cannot do so, then you might try commenting out the call to M3EXIT on line 947 of SOA_DEFN.F. It is possible the model will be able to recover, and this error can be ignored. But if the error persists, you will have to debug the issue. The algorithm and its references are given in the code.
1 Like
Thank you, your advice is really helpful. I will try this method later.