the impact of e_we and e_sn on simulation

Any miscellaneous question/problem/comment related to WRF support that does not fit under any of the other category
Post Reply
sd16121
Posts: 75
Joined: Sat Feb 06, 2021 4:44 pm

the impact of e_we and e_sn on simulation

Post by sd16121 » Tue Mar 15, 2022 12:27 pm

Dear Office,

I wonder whether the grid number will influence the WRF simulation result:
**********
&domains
time_step = 30,
max_dom = 1,
e_we = 100,
e_sn = 100,
dx = 4500,
dy = 4500,
parent_grid_ratio = 1,
parent_time_step_ratio = 1,
/

&domains
time_step = 3,
max_dom = 3,
e_we = 18, 22, 22,
e_sn = 18, 22, 22,
dx = 13500, 4500, 1500,
dy = 13500, 4500, 1500,
parent_grid_ratio = 1, 3, 3,
parent_time_step_ratio = 1, 3, 3,
/
**********
See the two cases above, one is a single model domain, another is a nested model runs. If I compare the simulation results of 4500m resolution (domain 1 = 4500m) and 1500m resolution (domain 2 = 4500m), they should be same or different?

The attached files are my simulation results, you may refer to them.

Best regards,
Stella
Attachments
0.png

sd16121
Posts: 75
Joined: Sat Feb 06, 2021 4:44 pm

Re: the impact of e_we and e_sn on simulation

Post by sd16121 » Thu Mar 31, 2022 8:22 pm

Dear Colleagues,

If there are anyone who could help with this question? Many thanks.

Best wishes,
Stella

sd16121
Posts: 75
Joined: Sat Feb 06, 2021 4:44 pm

Re: the impact of e_we and e_sn on simulation

Post by sd16121 » Sun Apr 17, 2022 1:19 pm

Dear Colleagues,

If anyone could provide some suggestions about my question above? Your help will be highly appreciated.

Best,
Stella

sd16121
Posts: 75
Joined: Sat Feb 06, 2021 4:44 pm

Re: the impact of e_we and e_sn on simulation

Post by sd16121 » Sat May 07, 2022 4:22 pm

Dear Colleagues,

Anyone can help with this issue? Looking forward to hearing from you soon.

Best wishes,
Stella

kwerner
Posts: 2839
Joined: Wed Feb 14, 2018 9:21 pm

Re: the impact of e_we and e_sn on simulation

Post by kwerner » Wed May 11, 2022 11:14 pm

Hi Stella,
Again, I apologize for the delay in response. As this question was posted under the same forum category as your other question, the same issue occurred.

We would expect the result to differ in these two cases, because in the nested simulation, domain 1 is receiving initial and boundary conditions from the input/first-guess data, and then updating the boundary and initial conditions for domain 2, based on domain 1. Whereas if you only have the single domain case, there is no update to a nested domain.
NCAR/MMM

sd16121
Posts: 75
Joined: Sat Feb 06, 2021 4:44 pm

Re: the impact of e_we and e_sn on simulation

Post by sd16121 » Fri May 13, 2022 7:41 pm

Dear Colleague,

Thank you for your answer. So, did you mean that the two cases with same resolution (4500m) should have different simulation results?
single model domain: domain1=4500m
nested model runs: domain2=4500m (domain1=1500m)

By the way, I wonder whether the grid number (e_we and e_sn) will influence the WRF simulation result?

Best regards,
Stella

kwerner
Posts: 2839
Joined: Wed Feb 14, 2018 9:21 pm

Re: the impact of e_we and e_sn on simulation

Post by kwerner » Wed May 18, 2022 10:55 pm

Stella,
Yes, that is what I meant. The difference will occur because of the fact that domain 2 is influenced by domain 1. And yes, the size of the domain (e_we and e_sn) will also cause slightly different results because the boundary conditions will be located in a different place.
NCAR/MMM

Post Reply

Return to “Miscellaneous”