Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

WRF crashes unexpectedly when conducting urban meteorological simulations using LCZ.

N

nanyin

New member
Hello, I encountered this message while using W2W. When I was conducting urban meteorological simulations, WRF exited after only running for a few minutes. I wonder if this is related to the setting of my num_urban_nbui value. I found the Registry.EM_COMMON file and came across this information: rconfig integer num_urban_nbui derived 1 1 irh.

I modified it to: rconfig integer num_urban_nbui derived 1 10 irh, and then recompiled WRF. Unfortunately, W2W still prompts "Set nbui_max to 10 during compilation, in order to optimize memory storage" and WRF still exits after running for a few minutes just like before.

I've tried using WRF versions 4.5.2 and 4.3.1, but the issue persists. Are there any other solutions? Thank you!
 
Can you clarify what is W2W?

Please send me your namelist.wps and namelist.input to take a look. Thanks.
 
Thank you for your reply. W2W is a tool that combines LCZ maps with WPS processing. Detailed information about W2W is available on this website: Client Challenge.
 

Attachments

  • namelist.wps.txt
    1.3 KB · Views: 1
  • namelist.input.txt
    3.8 KB · Views: 2
num_urban_nbui is a derived variable and you don't need to explicitly specify it.

Your case failed possibly due to other reasons. In your namelist.input. some options are not set cirrectly. Please change as follows:

cu_physics = 3, 0 , 0,

sf_urban_physics = 3, 3, 3,

Let me know whether you can run successfully. if not, please send me the error messages shown in your rsl files.

By the way, WPS provides LCZ data. I wonder why you created it using W2W, ---- this could be a potential issue, too.
 
Thank you so much for your reply—your help means a lot to me! I tried your suggestions and successfully generated the output. However, there are a large number of "USING DEFAULT URBAN MORPHOLOGY" messages in the rsl.error file. At the end of the file, it shows "wrf: SUCCESS COMPLETE WRF", which seems to indicate that the urban module has not been loaded correctly.

I suspect the issue might be with the W2W tool I used. The reason I turned to W2W is that I don’t know how to load LCZ maps into WRF, and I haven’t been able to find a complete workflow for using LCZ maps. Currently, I’m using WPS 4.5 and WRF 4.5.2. I’ve seen some websites mention that operations like linking files and modifying configuration files are required, but I’m not sure where to start with these steps.

Could you please share a detailed, step-by-step workflow for loading LCZ maps into WRF? I would be extremely grateful for your guidance!
 

Attachments

  • rsl.error.0000.txt
    1.1 MB · Views: 0
Generally speaking, the warning mesage doesn't really matter. It tells that some information may be missing in the input data and WRF simply uses the default value to fill the missing value, --- this is acceptable and should be fine.

LCZ map data is available for download from WRF website. Please see Static Data Downloads

To use this dataset, please set

geog_data_res = ‘cglc_modis_lcz + default'

Hope this is helpful for you. Let me know if you have any issues using LCZ as input.
 
Ming Chen said:
Generally speaking, the warning mesage doesn't really matter. It tells that some information may be missing in the input data and WRF simply uses the default value to fill the missing value, --- this is acceptable and should be fine.

LCZ map data is available for download from WRF website. Please see Static Data Downloads

To use this dataset, please set

geog_data_res = ‘cglc_modis_lcz + default'

Hope this is helpful for you. Let me know if you have any issues using LCZ as input.
Click to expand...
Hello! I modified geog_data_res = 'cglc_modis_lcz + default', properly extracted the LCZ compressed package, and imported it into WRF. The initial run went smoothly. I set the simulation time to 12 hours, but halfway through the run, the following error occurred. Could you please help me figure out what caused this?

Here is the ending part of rsl.error.0000:
d01 2020-07-01_14:03:18 16 points exceeded v_cfl = 2 in domain d01 at time 2020-07-01_14:03:18 hours
d01 2020-07-01_14:03:18 Max W: 128 42 31 W: 24.03 w-cfl: 2.79 dETA: 0.01
d01 2020-07-01_14:03:18 15 points exceeded v_cfl = 2 in domain d01 at time 2020-07-01_14:03:18 hours
d01 2020-07-01_14:03:18 Max W: 128 42 32 W: 6.21 w-cfl: 2.83 dETA: 0.01
d01 2020-07-01_14:03:18 11 points exceeded v_cfl = 2 in domain d01 at time 2020-07-01_14:03:18 hours
d01 2020-07-01_14:03:18 Max W: 128 42 33 W: 17.28 w-cfl: 2.41 dETA: 0.01
Timing for main: time 2020-07-01_14:03:24 on domain 3: 4.36805 elapsed seconds
Timing for main: time 2020-07-01_14:03:30 on domain 3: 4.50340 elapsed seconds
Timing for main: time 2020-07-01_14:03:36 on domain 3: 4.46337 elapsed seconds
Timing for main: time 2020-07-01_14:03:36 on domain 2: 19.99683 elapsed seconds
Timing for main: time 2020-07-01_14:03:42 on domain 3: 4.35114 elapsed seconds
Timing for main: time 2020-07-01_14:03:48 on domain 3: 4.44320 elapsed seconds
Timing for main: time 2020-07-01_14:03:54 on domain 3: 4.48397 elapsed seconds
Timing for main: time 2020-07-01_14:03:54 on domain 2: 19.91335 elapsed seconds
Timing for main: time 2020-07-01_14:04:00 on domain 3: 4.47731 elapsed seconds
Timing for main: time 2020-07-01_14:04:06 on domain 3: 4.60983 elapsed seconds
Timing for main: time 2020-07-01_14:04:12 on domain 3: 4.48074 elapsed seconds
Timing for main: time 2020-07-01_14:04:12 on domain 2: 20.29238 elapsed seconds
Timing for main: time 2020-07-01_14:04:12 on domain 1: 63.23141 elapsed seconds
d01 2020-07-01_14:04:12 6 points exceeded v_cfl = 2 in domain d01 at time 2020-07-01_14:04:12 hours
d01 2020-07-01_14:04:12 Max W: 128 42 28 W: 13.34 w-cfl: 2.33 dETA: 0.02
d01 2020-07-01_14:04:12 1 points exceeded v_cfl = 2 in domain d01 at time 2020-07-01_14:04:12 hours
d01 2020-07-01_14:04:12 Max W: 129 42 30 W: 19.93 w-cfl: 2.05 dETA: 0.02
d01 2020-07-01_14:04:12 14 points exceeded v_cfl = 2 in domain d01 at time 2020-07-01_14:04:12 hours
d01 2020-07-01_14:04:12 Max W: 129 42 30 W: 16.29 w-cfl: 2.53 dETA: 0.02

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0 0x770c9b423e59 in ???
#1 0x770c9b422e75 in ???
#2 0x770c9ae4532f in ???
at ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
#3 0x59f50180f28f in ???
#4 0x59f50182be98 in ???
#5 0x59f50182fa7b in ???
#6 0x59f501246861 in ???
#7 0x59f50135bc4e in ???
#8 0x59f500daac78 in ???
#9 0x59f500c3ffe3 in ???
#10 0x59f4ffc81cae in ???
#11 0x59f4ffc08bc7 in ???
#12 0x59f4ffc085fe in ???
#13 0x770c9ae2a1c9 in __libc_start_call_main
at ../sysdeps/nptl/libc_start_call_main.h:58
#14 0x770c9ae2a28a in __libc_start_main_impl
at ../csu/libc-start.c:360
#15 0x59f4ffc08634 in ???
#16 0xffffffffffffffff in ???
 
Hi
The CFL violation indicates the model is numerically unstable. Please set the options below:
(1) set a larger epssm value (e.g., epssm = 0.9)
(2) turn on w_damping (w_damping =1 )
(3) reduce time step to 4 x Delx

Then rerun your case. Please let me know if you still have issues.
 
You must log in or register to reply here.
Top