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

Basic WRF-WRFDA cycling

A

aFalcione

New member
Even after following the instructions of the User guide, chapter 6, I've been facing trouble in setting up a simple cycling assimilation. I am wondering whether I am doing something inherently wrong.


Here is the set up for a forecast starting on 2022-09-14_12:00 and ending on 2022-09-16_12:00:
  1. I create the initial conditions via WPS using GFS global data from the whole 48h period. So wrfinput and wrbdy contain 3h global data from 14_12 to 16_12.
  2. I use da_wrfvar.exe and da_update_bc.exe on wrfinput and wrfbdy to update the initial conditions.
  3. Using the outputs of 3) as inputs, I make a 3h-forecast from 14_12 to 14_15.
  4. I use da_wrfvar.exe and and da_update_bc.exe on wrfout_d01_2022-09-14_15:00:00.
  5. I use the outputs of 4) as inputs for another 3h-forecast, from 14_15 to 14_18.
  6. And so on.
The result is very different from a unique 48h-long "Control" forecast that I made for the same period. I am sure that the problem is not linked to the observations that I am assimilating because I also tried to turn of the assimilation of each kind of observations in &wrfvar4, but still the result is very different from the control forecast.

What am I doing wrong?

Thank you!!
PS: this is with WRFv3.8.1
 
Last edited:
Hello, I have the same problem as you. Has your problem been solved? And what method did you use?
Thank you!
 
wangjh said:
Hello, I have the same problem as you. Has your problem been solved? And what method did you use?
Thank you!
Click to expand...
Unluckily I did not manage to fix it, nor I did continue working on the cycling.
But if you find your error I'd be happy to know it!
 
Hello,

It seems that you are trying to set up a cycling assimilation in WRF, but you are experiencing significant differences between the cycling assimilation and the control forecast. You have followed the instructions in the User guide, chapter 6, and turned off the assimilation of observations, but the problem persists.

There could be several reasons why your cycling assimilation is not working properly. Here are a few things you could check:

Check your domain configuration: Make sure that your domain configuration is consistent between the cycling assimilation and the control forecast. If the domain configurations are different, this could be a source of discrepancy between the two forecasts.

Check your assimilation settings: Make sure that your assimilation settings are the same between the cycling assimilation and the control forecast. Check the namelist settings for &da, &obsproc, &wrfvar, and &namelist_wps. Even small differences in the settings could lead to significant differences in the forecast.

Check your observation data: Make sure that your observation data is the same between the cycling assimilation and the control forecast. Even small differences in the observation data could lead to significant differences in the forecast. My Sutter Online
 
Last edited:
You must log in or register to reply here.
Top