-
|
I've been getting errors in some, but not all, ED2 runs that are part of an ensemble run (using Sometimes the reason is "Failed finding the second guess for regula falsi". Are there any "usual suspects"—e.g. values in config.xml—that I should check to figure out why this is happening? |
Beta Was this translation helpful? Give feedback.
Replies: 3 comments 3 replies
-
|
This can happen if under certain environmental conditions. Is the wind very strong? I'd check the met drivers during the time they are crashing. Also check that your reference height (zref) is larger than your vegetation? This is required if you're using icanturb = 2. You can also try a different icanturb.
…________________________________
From: Eric R. Scott ***@***.***>
Sent: Wednesday, February 15, 2023 2:55 PM
To: EDmodel/ED2 ***@***.***>
Cc: Subscribed ***@***.***>
Subject: [EDmodel/ED2] Convergence errors with `zoobukhov8` in `canopy_air_coms.f90` (Discussion #380)
I've been getting errors in some, but not all, ED2 runs that are part of an ensemble run (using PEcAn). Many of them are similar to the one below:
--------------------------------------------------------------
!!! FATAL ERROR !!!
--------------------------------------------------------------
---> File: canopy_air_coms.f90
---> Subroutine: zoobukhov8
---> Reason: Zeta didn't converge, giving up!!!
--------------------------------------------------------------
Sometimes the reason is "Failed finding the second guess for regula falsi".
Are there any "usual suspects"—e.g. values in config.xml—that I should check to figure out why this is happening?
—
Reply to this email directly, view it on GitHub<#380>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AH77Q3ZXLJWHJJV5WFBX6PTWXVGEFANCNFSM6AAAAAAU5MJGG4>.
You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
-
|
is zref something set in ED2IN? |
Beta Was this translation helpful? Give feedback.
-
|
Found this issue: #58 (comment), which looks like what you're talking about. My |
Beta Was this translation helpful? Give feedback.
-
|
I had this issue in the past due to some variable not being correctly initialised in the cohort or patch dynamics step. If the crash is happening on the first day of the month, then it is likely not a problem with zoobukhov8, it is just that the function received a NaN that caused the solution to fail. |
Beta Was this translation helpful? Give feedback.
-
|
This seems like a more likely culprit. In a more recent attempt at an ensemble run, I'm getting this along with budget check errors, "attempt to set a site with 0 patches" errors, and some SIGABRT errors. |
Beta Was this translation helpful? Give feedback.
-
|
I made my grasses and forbs have a higher DBH (1 cm instead of 0.6) in my .css file and it seems to have solved all the bugs (except the SIGABRT on the last day of simulation). Can't remember where I saw it, but there was an issue in the repo that made me think that small and/or sparse plants were causing ED2 to have difficulty calculating air turbulence or something. I have no clue what a DBH of 1cm means for a grass, but hopefully it is a reasonable fix?? |
Beta Was this translation helpful? Give feedback.
I made my grasses and forbs have a higher DBH (1 cm instead of 0.6) in my .css file and it seems to have solved all the bugs (except the SIGABRT on the last day of simulation). Can't remember where I saw it, but there was an issue in the repo that made me think that small and/or sparse plants were causing ED2 to have difficulty calculating air turbulence or something. I have no clue what a DBH of 1cm means for a grass, but hopefully it is a reasonable fix??