Hello all,
I am simulating the Typhoon MANGKHUT in 2018 (from 20180900 to 20180917). I have set up three domains with resolutions of 12 km, 4 km, and 1.3 km and I am using a two-way nested setup, using spectral nudging, with one wrfinput file.
My issue is that the model crashed with a segmentation fault at **20180916T16:00:00** (Also attached namelist.input).The error information which I got is as follows:
==== backtrace (tid: 170675) ====
0 0x0000000002a21c52 module_sf_sfclayrev_mp_psim_stable_() ???:0
1 0x0000000002a1db1a module_sf_sfclayrev_mp_sfclayrev1d_() ???:0
2 0x0000000002a1b6ef module_sf_sfclayrev_mp_sfclayrev_() ???:0
3 0x0000000001c318e0 module_surface_driver_mp_surface_driver_() ???:0
4 0x00000000014df0e4 module_first_rk_step_part1_mp_first_rk_step_part1_() ???:0
5 0x0000000001d7a041 solve_em_() ???:0
6 0x0000000001b165e4 solve_interface_() ???:0
7 0x0000000001701aeb module_integrate_mp_integrate_() ???:0
8 0x0000000001702108 module_integrate_mp_integrate_() ???:0
9 0x0000000001702108 module_integrate_mp_integrate_() ???:0
10 0x00000000017003c1 module_wrf_top_mp_wrf_run_() ???:0
11 0x00000000004201a4 uwin_interface_wrf_mp_wrf_component_run_() ???:0
12 0x00000000005238c9 ESMCI::FTable::callVFuncPtr() ???:0
13 0x00000000005269a5 ESMCI_FTableCallEntryPointVMHop() ???:0
14 0x00000000008f09e5 ESMCI::VM::enter() ???:0
15 0x0000000000524788 c_esmc_ftablecallentrypointvm_() ???:0
16 0x0000000000ab448d esmf_compmod_mp_esmf_compexecute_() ???:0
17 0x0000000000c32f7f esmf_gridcompmod_mp_esmf_gridcomprun_() ???:0
18 0x00000000004164de MAIN__() ???:0
19 0x00000000004148ce main() ???:0
20 0x00000000000223d5 __libc_start_main() ???:0
21 0x00000000004147d9 _start() ???:0
=================================
Since the typhoon had already made landfall by this time, a huge negative (-993.397 W/m2) sensible heat flux was found in an urban grid at 20180916T16:00:00 (see figure), At that time, I was using sf_urban_physics = 0.
What I already tested:
(1) I checked other users' error logs ((RESOLVED) WRF Single-layer, UCM crashed) and tried modifying the TS_SCHEME in URBPARM_LCZ.TBL and URBPARM.TBL from 1 (4-layer model) to 2 (force-restore method), but it didn’t resolve the issue.
(2) Then, I changed sf_urban_physics = 1, 1, 0, but the model still crashed at 20180916T16:00:00 with the same error.
(3) However, when I set sf_urban_physics = 1, 1, 1, the model ran successfully without crashing, but I encountered the following warning:
**"Nest movement is disabled because of namelist settings. UCMs/Noah LSM cannot be specified with moving nests. Movement disabled."**
### Follow-up Questions:
(1) How should I resolve this issue?
(2) If setting sf_urban_physics = 1, 1, 1 disables nest movement, even though the model completes successfully, does this mean the results from the moving nested domain are incorrect?
I am simulating the Typhoon MANGKHUT in 2018 (from 20180900 to 20180917). I have set up three domains with resolutions of 12 km, 4 km, and 1.3 km and I am using a two-way nested setup, using spectral nudging, with one wrfinput file.
My issue is that the model crashed with a segmentation fault at **20180916T16:00:00** (Also attached namelist.input).The error information which I got is as follows:
==== backtrace (tid: 170675) ====
0 0x0000000002a21c52 module_sf_sfclayrev_mp_psim_stable_() ???:0
1 0x0000000002a1db1a module_sf_sfclayrev_mp_sfclayrev1d_() ???:0
2 0x0000000002a1b6ef module_sf_sfclayrev_mp_sfclayrev_() ???:0
3 0x0000000001c318e0 module_surface_driver_mp_surface_driver_() ???:0
4 0x00000000014df0e4 module_first_rk_step_part1_mp_first_rk_step_part1_() ???:0
5 0x0000000001d7a041 solve_em_() ???:0
6 0x0000000001b165e4 solve_interface_() ???:0
7 0x0000000001701aeb module_integrate_mp_integrate_() ???:0
8 0x0000000001702108 module_integrate_mp_integrate_() ???:0
9 0x0000000001702108 module_integrate_mp_integrate_() ???:0
10 0x00000000017003c1 module_wrf_top_mp_wrf_run_() ???:0
11 0x00000000004201a4 uwin_interface_wrf_mp_wrf_component_run_() ???:0
12 0x00000000005238c9 ESMCI::FTable::callVFuncPtr() ???:0
13 0x00000000005269a5 ESMCI_FTableCallEntryPointVMHop() ???:0
14 0x00000000008f09e5 ESMCI::VM::enter() ???:0
15 0x0000000000524788 c_esmc_ftablecallentrypointvm_() ???:0
16 0x0000000000ab448d esmf_compmod_mp_esmf_compexecute_() ???:0
17 0x0000000000c32f7f esmf_gridcompmod_mp_esmf_gridcomprun_() ???:0
18 0x00000000004164de MAIN__() ???:0
19 0x00000000004148ce main() ???:0
20 0x00000000000223d5 __libc_start_main() ???:0
21 0x00000000004147d9 _start() ???:0
=================================
Since the typhoon had already made landfall by this time, a huge negative (-993.397 W/m2) sensible heat flux was found in an urban grid at 20180916T16:00:00 (see figure), At that time, I was using sf_urban_physics = 0.
What I already tested:
(1) I checked other users' error logs ((RESOLVED) WRF Single-layer, UCM crashed) and tried modifying the TS_SCHEME in URBPARM_LCZ.TBL and URBPARM.TBL from 1 (4-layer model) to 2 (force-restore method), but it didn’t resolve the issue.
(2) Then, I changed sf_urban_physics = 1, 1, 0, but the model still crashed at 20180916T16:00:00 with the same error.
(3) However, when I set sf_urban_physics = 1, 1, 1, the model ran successfully without crashing, but I encountered the following warning:
**"Nest movement is disabled because of namelist settings. UCMs/Noah LSM cannot be specified with moving nests. Movement disabled."**
### Follow-up Questions:
(1) How should I resolve this issue?
(2) If setting sf_urban_physics = 1, 1, 1 disables nest movement, even though the model completes successfully, does this mean the results from the moving nested domain are incorrect?
Last edited: