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

Strange time dimensions and values of wrf_out files

X

xjhhh48

New member
Hi all,

I am running the WRF-Noah MP model in East Asia during 2021.6-2021.8 without any errors reported. However, I found some strange things in my wrf.output files:
(1) Strange time dimension. Generally, my daily wrfout files have a dimension of 159*159*1 (My study area covers 159 grid cells). But an unusual length of time dimension (i.e., 159*159*59) occurred for specific dates (e.g., output on 12th July and others), with the size of the output file (~9 Mb) almost twice of previous outputs (~5 Mb). I checked that file and found the values of the dimension 2-59 are invalid, I only need the 1st-order results (159*159*1).
(2) Lower results. Since the WRF model is forced using ERA5 single level and pressure level data, I compared the WRF-based soil moisture and skin temperature with raw ERA5 results. I found a consistent underestimation existing in WRF results, which is forced by ERA5. I know errors are very common in such simulations, but I am not sure if this phenomenon is reasonable. Please see the figures below:
T.jpgSMS.jpg
(3) The deep soil moisture (variable: SMCWTD) modeled by Noah MP is always zeros on all days during the 3-month simulation in summer. Is it normal? How to see it?
Could you please give me some suggestions about how to solve/explain these issues? Any feedback would be deeply appreciated.
The namelist.wps and namelist.input files are attached.

Best,
Jinghua
 

Attachments

  • namelist.wps
    1.5 KB · Views: 2
  • namelist.input
    3.9 KB · Views: 9
I am perplexed by the model behavior to write multiple output to a single file when you specify
frames_per_outfile = 1, 1,
Honestly I have no explanation for this issue and I have never seen such issue before.

As for your 2nd question, would you please checked the ERA5 data and compare with wrfout? I am suspicious this is originated form ERA5. Plesse let me now if I am wrong.
 
Last edited:
Ming Chen said:
I am perplexed by the model behavior to write multiple output to a single file when you specify
frames_per_outfile = 1, 1,
Honestly I have no explanation for this issue and I have never seen such issue before.

As for your 2nd question, elf you please checked the ERA5 data and compare with wrfout? I am suspicious this is originated form ERA5. Plesse let me now if I am wrong.
Click to expand...
Thank you for your reply.
1. Yes, and this issue can occur very occasionally without any pattern being detected. It happens in about 10 days randomly during the 3-month simulation.
2. The inserted two figures are the comparisons between ERA5 and ERA5-derived WRF outputs. The difference is too large I think, even for the variable skin temperature.
 
I run a few tests but cannot repeat the wrfout issue. Are there any special pattern (e.g., at some specific times after the initiation of WRF run) or circumstances (e.g., restart run, specific physics options, etc.) for this issue to occur in your runs?
Please continue to pay attention to this issue and let us know if you find any hints that may lead to this weird model behavior. Thanks in advance.
 
You must log in or register to reply here.
Top