Hi Luis,
This post descibes the process of downloading the correct ancillaries for the particular GCM that is driving HadRM3P (the RCM).
viewtopic.php?f=1&t=66In your case, I strongly recommend that you consider using HadCM3Q0 (aenwh) if you have it, or ECHAM5. The ECHAM4 GCM is now deprecated. We still "allow" echja (the A2) for 1960-2100 but I think B2 echjd is now limited to 2070-2100. So I would not proceed with the experiment bolac. Create "bolad" with the same parameters except using a different driving GCM (aenwh, echvc).
If you need new boundary data, just let us know. We can also set you up with a newer version of PRECIS.
Thanks
David