The USJ model paused during scenario runs. It is unclear whether this is a model issue or computing issue, though likely it is the former. In the most recent scenario run (IFRs x prices), pausing occurred for HADGEM2 for 1 hour between 1938 & 1939 and for MIROC5 for ~8.5 hours between 2039 & 2040. In a previous run, the pause also happened after 1938 for HADGEM2 (after which I interrupted the run to start over), which leads me to believe there is some issue with the model, possibly related to not dealing with some extreme condition correctly.
The source of this issue needs to be identified and, if possible, fixed, since it could also affect runs on the cluster as well.
The USJ model paused during scenario runs. It is unclear whether this is a model issue or computing issue, though likely it is the former. In the most recent scenario run (IFRs x prices), pausing occurred for HADGEM2 for 1 hour between 1938 & 1939 and for MIROC5 for ~8.5 hours between 2039 & 2040. In a previous run, the pause also happened after 1938 for HADGEM2 (after which I interrupted the run to start over), which leads me to believe there is some issue with the model, possibly related to not dealing with some extreme condition correctly.
The source of this issue needs to be identified and, if possible, fixed, since it could also affect runs on the cluster as well.