LAKEPARM and routing-parameter changes do not affect CHANOBS

22 views
Skip to first unread message

Negusu Tarekegn

unread,
Jul 24, 2026, 4:27:26 PMJul 24
to wrf-hydro_users

Dear WRF-Hydro users,

I am running offline WRF-Hydro v5.3 with reach-based routing for the Lake Tana, Upper Blue Nile Basin.

At the outlet of Lake Tana, the simulated wet-season peak occurs about 45 days earlier than observed and is substantially overestimated. The attached figure compares the observed discharge with the control run and several sensitivity experiments from January 1, 2006, to May 31, 2007.

Among the experiments, the three most recent were:

TEST08: Spatial soil and groundwater parameters improved the tributary simulations and slightly improved the lake outlet.

TEST09: Changes to LAKEPARM.nc produced CHANOBS results identical to TEST08.

TEST10: Large changes to routing parameters in Route_Link.nc, intended to delay the outlet flow, also produced results identical to TEST08.

I confirmed that selected CHANOBS_DOMAIN files from TEST08, TEST09, and TEST10 were byte-for-byte identical.

Could you please clarify:

  1. Does nlakes = 0 mean that LAKEPARM.nc is inactive?
  2. Are MusK and MusX used with channel_option = 2?
  3. How can Lake Tana be activated as a lake or storage component?
  4. Should LAKEOUT files be produced when lake routing is active with reach-based routing?
  5. Is lake_option required in WRF-Hydro v5.3? Adding it to hydro.namelist caused a namelist read error.
  6. What is the recommended way to represent Lake Tana storage and delayed release?

Any guidance would be greatly appreciated.

Best regards,

Negusu

 

image.png

Arezoo RafieeiNasab

unread,
Jul 24, 2026, 4:55:52 PMJul 24
to wrf-hyd...@ucar.edu
Hi Negsu, 

Would you be able to share the namelists with us, as well as the Routelink and LAKEPARM files? I would like to see the options activated before answering some of your questions. 

Thanks!
Arezoo

--
You received this message because you are subscribed to the Google Groups "wrf-hydro_users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wrf-hydro_use...@ucar.edu.
To view this discussion visit https://groups.google.com/a/ucar.edu/d/msgid/wrf-hydro_users/CACGuQy4dToSdBeGnd75%3DnH7bsd5YaVx%3DdKDu8vdLwVE7y8pSjg%40mail.gmail.com.


--
-------------------------------------------------------------------------------------------------
My working day may not be your working day. Please do not feel obliged to reply to this email outside of your normal working hours.
-------------------------------------------------------------------------------------------------
Arezoo Rafieei Nasab, Ph.D.
Scientist V
NSF NCAR Research Applications Laboratory

Negusu Tarekegn

unread,
Jul 24, 2026, 5:44:24 PMJul 24
to wrf-hyd...@ucar.edu

Hi Arezoo,

Thank you for your quick and timely response.

I have attached the requested files: the namelists, Route_Link.nc, and LAKEPARM.nc.

I would greatly appreciate your guidance after you check the activated options and file settings.

Best regards,
Negusu




files.zip

Arezoo RafieeiNasab

unread,
Jul 24, 2026, 10:44:00 PMJul 24
to wrf-hyd...@ucar.edu
Hi Negsu,

I took a look at the files you shared and have a follow up question about the lake area. It seems the lake area is set to 2974659776.06591 sqkm, which is a very large number. Could you double check this number? I Googled the lake surface area and it seems the actual area is 1,000 times smaller. Also the weir length is larger than the dam length, which does not seem correct to me. 
Here is my two cents on the questions you asked. 

  1. Does nlakes = 0 mean that LAKEPARM.nc is inactive?
I have not seen this error. If you want me to test your domain, please share the full domain directory so I can test it. But this seems to indicate that WRF-Hydro somehow did not see any lakes, but the file you shared with me shows the lake in the Routelink and also in LAKEPARM.The nc file contains the lake information. So I am not sure why it is not being read. Could we do a test? Please set the outlake  = 1. The model will output the LAKEOUT files if it models the lake correctly. 

  1. Are MusK and MusX used with channel_option = 2
I do not think these are used in the MC methodology, they are calculated based on other properties. If you want to modify a parameter to delay the flow, Manning's coefficient would be the one to adjust. However, I am not sure how much impact it will have. I haven't tested it myself, but this is a parameter that users often calibrate.

  1. How can Lake Tana be activated as a lake or storage component?
What you have in the hydro.namelist is usually what we want to set which is define the LAKEPARM file, indicating that there is a lake available. 

  1. Should LAKEOUT files be produced when lake routing is active with reach-based routing?
Regardless of the routing option, if you set outlake = 1, it should output LAKEOUT files. 

  1. Is lake_option required in WRF-Hydro v5.3? Adding it to hydro.namelist caused a namelist read error.
lake_option is a new feature and added in the latest release (https://github.com/NCAR/wrf_hydro_nwm_public/releases/tag/v5.4.0), therefore if you are using an earlier version it will give you an error. 

  1. What is the recommended way to represent Lake Tana storage and delayed release?
Let's circle back to this question after we figure out the earlier issues. Please carefully review the numbers specified in the LAKEPARM.nc file, turn on the outelake and let me know what you find. 

Thanks!
Arezoo


Arezoo RafieeiNasab

unread,
Jul 29, 2026, 11:46:03 AMJul 29
to wrf-hyd...@ucar.edu
Hi Negsu, 

I checked the domain you sent me, and it works if you upgrade the WRF-Hydro version to 5.4 (https://github.com/NCAR/wrf_hydro_nwm_public/releases/tag/v5.4.0). This is a more recent release and has some bug fixes for lake modules also which seems to resolve the issue you were facing. So I would recommend upgrading to 5.4. I did not change anything in the domains or namelists, and it ran for me fine. 

As for the lake area, I checked the WRF-Hydro code and it should be in square kilometers, not square meters. I think the documentation contains an error. I also checked some of my files, and they are also in square km. After you upgrade the WRF-Hydro model version, please adjust the lake area so we can see some meaningful release. 

Please let me know if this works for you or if you have any remaining questions. 
Thanks!
Arezoo

Negusu Tarekegn

unread,
Jul 29, 2026, 12:27:24 PMJul 29
to wrf-hyd...@ucar.edu

Hi Arezoo,

Thank you very much for testing the domain and identifying the issue. This is extremely helpful.

I will upgrade WRF-Hydro to version 5.4, keep the domain files and namelists unchanged, and correct the lake area to km².

Thank you also for checking the source code and confirming that the lake area should be provided in square kilometres. I will share the results after completing the new test and let you know if any issues remain.

Best regards,
Negusu


Reply all
Reply to author
Forward
0 new messages