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

Questions about topography sensitivity experiment

This post was from a previous version of the WRF&MPAS-A Support Forum. New replies have been disabled and if you have follow up questions related to this post, then please start a new thread from the forum home page.

Jacky Wei

New member
Hi, all!
I am trying to do some sensitivity experiments using WRF ARW v4.0. The three-level nesting domain is attached here ("wps_show_dom.pdf").

Question 1: I would like to examine the role of high topography (~700 m) in domain d03, but domains d01 and d02 also include topography. So, do I need to remove or reduce the topography in only d03 or all domains (d01-d03)? I have this question because I worry that the "topographic effect" from the domains d01 and d02 may also exert an influence on the innermost domain d03.

Question 2: Now, assuming that question 1 is well addressed, for example, only terrain height ("HGT_M" in geo_em_d0* file) in domain d03 is changed (such as reduced to one-third of the original value). The trouble is that although "HGT_M" is reduced, the perturbation induced by the high topography is still there in other dynamical and thermodynamical variables such as wind, pressure, humidity, temperature, etc. My idea is to fill in these variables over the region where the topography is reduced using the averaged value of the surrounding oceans or surface level. So, is it the right way?

Question 3: Now, question 2 also see its solution. However, the boundary conditions (BIs) that are generally derived from the NCEP GFS will also bring in the topographic influence. Do I also need to remove this topographic influence from BIs? For example, using the mean value of the entire simulation period to replace the corresponding value in BIs.

I have checked some references and tutorials and it seems that the only thing they suggested to do is to change the variable "HGT_M". But, as I have mentioned above, I am not sure if other factors will cause a topographic effect either that influences the domain d03.

Thank you very much for your answers!

Jacky Wei


  • wps_show_dom.pdf
    86.9 KB · Views: 51
Are you going to run 1-way nesting or two-way nesting case? The terrain height modification should depend on what style of nesting you want to implement.
If you want to run two-way nesting, then terrain height changes in D03 will feed back to D02 and then to D01. I would suggest that you change the terrain height in all the three domains, just to make sure it is consistency between them.

For your questions (2) and (3), personally I don't think your solutions are practical. It is hard to manually adjust meteorological elements, especially when these elements are used to drive LAM model. Note that this is just my personal opinion, which may not be correct. Probably you have a good approach to do such adjustments.
Dear Ming:
I deeply thank your reply!
For questions 2 and 3, I was not going to adjust those variables manually during the runtime of WRF. I only want to revise them in the initial and boundary conditions. Actually, those suggestions from questions 2 and 3 were from my supervisor. I don't think it is reasonable to make adjustments in initial and boundary conditions, because these adjustments will largely exceed and cover the "pure effect" of topography (have been proved from results of my new experiments). Since my original goal is to explore the role of topography, therefore I should only consider revising the "HGT_M" in domains d01-d03, as also suggested by you.
You suggested that for a two-way nesting run, revising topography in domain 01 is also needed. So, what's the solution of a one-way nesting run? I have compared the one-way and two-way nesting runs, and they actually produced an almost identical simulation.
Looking forward to your reply! Thank you so much!
Best regards,
If you run 1-way eating, and you only want to look at the impact of terrain change in your child domain, you can modify HGT in the child domain. In this case, the forcing from the parent domain is still realistic and the impact of HGT changes are only shown in your child domain output.