| PRECIS support forum http://forum.precisrcm.com/ |
|
| Can not determine new sice master ancillary file. http://forum.precisrcm.com/viewtopic.php?f=1&t=103 |
Page 1 of 1 |
| Author: | kvrmtechsvu [ Fri Jul 08, 2016 9:26 am ] | ||
| Post subject: | Can not determine new sice master ancillary file. | ||
hi, I started a simulation 1989/12/01 with PRECIS for the historical scenario, the model showed the message: ERROR: Failed to update ancillary files please help me to start simulation. with regards, K.Venkatramana
|
|||
| Author: | dmhg [ Fri Jul 08, 2016 11:05 am ] |
| Post subject: | Re: Can not determine new sice master ancillary file. |
Hello, Please try this: # Tidy up cd $ANCIL_MASTER rm *acre* rm *tcr* rm *vtr* # Repair the problem mv HadGEM2-ES_historical_seaice_1950-1955 HadGEM2-ES_historical_seaice.1950-1955 mv HadGEM2-ES_historical_seaice_1960-1965 HadGEM2-ES_historical_seaice.1960-1965 [and so on] also: mv HadGEM2-ES_historical_sst_1950-1955 HadGEM2-ES_historical_sst.1950-1955 [and so on] The issue is caused by the final _ -- it needs to be a . When you rename the files so that the final _ is changed to . then try to restart. Is there a particular reason you're running such a short model run (1989-2000)? Eleven years is the bare minimum (1 year spinup to be discarde, 10 years of model data) but it's always better to run longer integrations in order to capture multi-decadal variability. If you have the boundary data for the future HadGEM2-ES (rcp data 2005-2099) then you'll need to download the ancillary files, because the ones needed are not present in $ANCIL_MASTER To find out what you need, have a look at these files $UMDIR/vn4.5/ctldata/ancil_config/akwsa (Rcp 2.6) $UMDIR/vn4.5/ctldata/ancil_config/akwsl (Rcp 4.5) $UMDIR/vn4.5/ctldata/ancil_config/akwsx (Rcp 8.5) And grab the corresponding files from http://www.precisrcm.com/ANCIL/master (be sure to unzip them after you download them!) David |
|
| Author: | kvrmtechsvu [ Sat Jul 09, 2016 8:24 am ] | ||
| Post subject: | Re: Can not determine new sice master ancillary file. | ||
Hi David, Thank you for reply, I followed u, still I have same issue, please find the leave file as attachment. Actually I am ruing model in two work stations one is from 1970-1990 & other 1990-2005 historical scenario. later I will go for RCP 4.5. please help me to start simulation. with regards, K.Venkatramana
|
|||
| Author: | dmhg [ Sat Jul 09, 2016 3:43 pm ] |
| Post subject: | Re: Can not determine new sice master ancillary file. |
http://www.precisrcm.com/ANCIL/master/HadGEM2-ES_historical_seaice.1985-1989.gz Hello, That file is somehow missing from your $ANCIL_MASTER directory. Download it, unzip it, and copy it to $ANCIL_MASTER and then try to restart. Thank you David |
|
| Author: | sotucker [ Mon Jul 11, 2016 1:36 pm ] |
| Post subject: | Re: Can not determine new sice master ancillary file. |
Dear K.Venkatramana (and other users who may read this post), Please can I give a slight amendment to David's post. Please do NOT perform the step Quote: # Repair the problem mv HadGEM2-ES_historical_seaice_1950-1955 HadGEM2-ES_historical_seaice.1950-1955 mv HadGEM2-ES_historical_seaice_1960-1965 HadGEM2-ES_historical_seaice.1960-1965 [and so on] also: mv HadGEM2-ES_historical_sst_1950-1955 HadGEM2-ES_historical_sst.1950-1955 [and so on] Instead, it is better to remove these files and download new ones. Regards, Simon |
|
| Author: | dmhg [ Mon Jul 11, 2016 2:14 pm ] |
| Post subject: | Re: Can not determine new sice master ancillary file. |
Hi, All the "new ones" are in www.precisrcm.com/ANCIL/master David |
|
| Author: | kvrmtechsvu [ Wed Jul 13, 2016 5:49 am ] |
| Post subject: | Re: Can not determine new sice master ancillary file. |
Hi David, thank you for help. I had other problem, how to start Model from 1975-01-01 ?. its allowing only from 1950/12/1, 1959/12/1,1969/12/1,1979/12/1,1989/12/1,1999/12/1,. with regards, K.Venkatramana |
|
| Author: | dmhg [ Wed Jul 13, 2016 8:24 am ] |
| Post subject: | Re: Can not determine new sice master ancillary file. |
Dear K. Venkatramana, I'm afraid the only start times are 1 Dec ???9 dates (like 1 Dec 1969, 1 Dec 1979). You are best advised to start your experiment on 1 Dec 1969 to cover the 1970s. Remember to remove the first 12 or 13 months of any PRECIS experiment as model spin-up. Don't include this data in any analysis. Thank you David |
|
| Author: | kvrmtechsvu [ Tue Jul 26, 2016 10:23 am ] | ||
| Post subject: | Re: Can not determine new sice master ancillary file. | ||
Hi David, I started historical simulation from 1979/12/01 up to 2005/12/01 with new Ancil/master files. after 5 years model had stopped, so I restarted from command line ( rerun abcaa i4c10 21,3,0,0,0,0 NRUN) for next 21 years. After compilation of one month again it shows NRUN finished at 21:23 25/Jul/2016. Please find .leave file as attachment Please help me to complete simulation. thank you with regards K.Venkatramana, TERI
|
|||
| Author: | dmhg [ Tue Jul 26, 2016 4:56 pm ] |
| Post subject: | Re: Can not determine new sice master ancillary file. |
Hello, Here's the output at the end of the file you kindly attached. Quote: ------------------------------------------------------+------------------------- REAL TIME : 15:32:28 IST Tuesday 26/Jul/2016 | RATE : MODELS : HadRM3P (MOSES2.2), HadGEM2-ES (akwss) | TIME : RUNID : abcaa | DATE : STATUS : NRUN finished at 15:32 26/Jul/2016 | MWLE : 0% ------------------------------+-----------------------+------------------------- start : 1984/12/ 1 0:00z | COMPLETED : 1 Month CURRENT : 1985/ 1/ 1 0:00z | % : 0.392% of 21 Years 3 Months end : 2006/ 3/ 1 0:00z | TIMESTEPS : 25920/6609600 (100 secs/TS) ------------------------------+------------------------------------------------- PRECIS experiments always stop after the first month when an *N*ew run NRUN is issued. That includes rerun-as-NRUN. Try typing Code: restart abcaa And see what happens. It *should* enter the CRUN phase and run *C*ontinuously until the end. Thanks David |
|
| Page 1 of 1 | All times are UTC |
| Powered by phpBB® Forum Software © phpBB Group https://www.phpbb.com/ |
|
