Error- init routine for LSM + DA instance Noah-MP.4.0.1+ESA CCI soil moisture is not defined #1084
-
|
Hi, I'm trying to assimilate ESA CCI version 6.1 soil moisture into Noahmp 4.0.1 LSM and got an error Program received signal SIGSEGV: Segmentation fault - invalid memory reference. Backtrace for this error: config file is attached. Thank you |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 34 replies
-
|
Hi @Gayathridevi123, The error you encountered indicates that NoahMP401 version hasn't been set up with the ESA CCI SM DA obs. When I look in our make/default.cfg file, it looks like the plugins are actively there in the latest LIS code, but the connections need to be here, if using the master branch: https://github.com/NASA-LIS/LISF/blob/master/lis/plugins/LIS_lsmda_pluginMod.F90 You could copy the NoahMP361 + ESA CCI SM DA obs entries and then enter them for NoahMP401: ! NoahMP-3.6 ESACCI soil moisture replace "LIS_noahmp36Id" with "LIS_noahmp401Id". Please try that to see if that fixes your problem. Thanks. |
Beta Was this translation helpful? Give feedback.
-
|
I'm going to retract my last comment, it looks like there is a bug in LIS. @Gayathridevi123, I notice that your The routine @Gayathridevi123, can you try switching your output naming style from "4 level hierarchy" to "3 level hierarchy"? |
Beta Was this translation helpful? Give feedback.
-
|
Thank you so much for your help, @bmcandr. For @Gayathridevi123: I just tested the master version of LIS that you have been using with the following exact private module setup: https://github.com/NASA-LIS/LISF/blob/master/env/discover/lisf_7_gnu_11.2.0_impi_2021.4.0 and I was able to successfully able to run and generate the DA OBS output files with no problem and no changes to the code. Also, I configured with the -1 debug flag option (when running the ./configure script), and it didn't capture any issues or problems with that recently. |
Beta Was this translation helpful? Give feedback.
-
|
Good catch, @bmcandr! Yes, you want to change: That should fix the issue within your lis.config file, @Gayathridevi123. |
Beta Was this translation helpful? Give feedback.
-
|
That sounds great! Definitely I will try and update it tomorrow. Thank you so much once again @karsenau @bmcandr, for your time, effort and support. |
Beta Was this translation helpful? Give feedback.
-
|
First of all, I am really thankful to both of you. I thought my compiler versions are creating the issue and I almost lost all hopes the day before yesterday. I sincerely appreciate your effort and support in solving the issues. I changed the Output naming style: "4 level hierarchy" to "3 level hierarchy". Now DAOBS directory is making and output is writing in the 201501 directory. EnKF directory is also created with output. But after that, unfortunately, it stopped by giving an error- Program received signal SIGSEGV: Segmentation fault - invalid memory reference. Backtrace for this error:
|
Beta Was this translation helpful? Give feedback.
-
|
Hi @Gayathridevi123: Glad you were able to get past the error you were getting. Thanks! |
Beta Was this translation helpful? Give feedback.
-
|
Hi @karsenau, I would like to know about the no of ensembles per tile. I'm just trying for only a 5 day simulation as a test case. So is there any threshold value for the no of ensembles? |
Beta Was this translation helpful? Give feedback.
-
|
Hi @Gayathridevi123: There are some papers out there about this to make sure you capture enough uncertainty and error in what you are perturbing. For testing purposes -- you can test with few as 5 or 6 members. We typically use 12 for testing, in case you want to run in the future with more members. I would recommend first testing with "coldstart" model state conditions for Noah3.6, since you need to run LDT to generate the ensemble-based restart files -- those files have to match the number of ensemble members and tiles of your LIS model run. |
Beta Was this translation helpful? Give feedback.
-
|
Hi @karsenau, Thank you for your reply regarding the ensemble member. As you suggested I tried
Program received signal SIGSEGV: Segmentation fault - invalid memory reference. Backtrace for this error:
|
Beta Was this translation helpful? Give feedback.
-
|
Hi @karsenau, Those errors were due to compiler issues now installed LISF with the Intel compiler and the problems are solved. Thank you for helping me in resolving the issues related to Data assimilation. Gayathri |
Beta Was this translation helpful? Give feedback.
-
|
Hi @Gayathridevi123: I am glad you were able to resolve the issues you were encountering and thank you for letting us know. |
Beta Was this translation helpful? Give feedback.
Hi @Gayathridevi123,
The error you encountered indicates that NoahMP401 version hasn't been set up with the ESA CCI SM DA obs. When I look in our make/default.cfg file, it looks like the plugins are actively there in the latest LIS code, but the connections need to be here, if using the master branch:
https://github.com/NASA-LIS/LISF/blob/master/lis/plugins/LIS_lsmda_pluginMod.F90
You could copy the NoahMP361 + ESA CCI SM DA obs entries and then enter them for NoahMP401:
! NoahMP-3.6 ESACCI soil moisture
call registerlsmdainit(trim(LIS_noahmp36Id)//"+"//&
trim(LIS_ESACCIsmobsId)//char(0),noahmp36_dasoilm_init)
call registerlsmdagetstatevar(trim(LIS_noahmp36Id)//"+"//&
trim(LIS_ESACCIsmobs…