Hello, i want to use Nudging method.

[kobl1201]

Hello

when i use ROMS Model, I want to use Nudging method.

Actually, i want to output data for u,v velocity.

Therefore, i made initial data over 1 year, and then want to nudge interval 1 day.

How can i do??

I saw TNUG and M3NUDG. How can i write down this parameter for nudging interval 1 day.

thank you

[kate]

Could you be more specific? Are you nudging just at the boundaries as part of the boundary condition or do you want to nudge over the full 3-D fields? In either case, you need to provide the values to nudge to, for the former these are boundary files, for the latter they are climatology files.

You mention outputting daily. That's another question entirely - you set NHIS or NAVG, depending.

[kobl1201]

I want to be nuding for initail data.
When i run the model, i need to the climatology file, boundary file and initial file.
Among three data, if i wanto to nudge for 3-D fields, which files should i change??
I think i need to change initial data by making 1 year data (interval 1 day), and M3NUDG value will be changed 1.0d0.
I will outputs the file to the average day interval.
Is it right??
if you don'y understand i mean, could you tell me again, please.

[kate]

Still not clear, sorry.

I want to be nuding for initail data.

This doesn't make sense. The initial conditions are just that - conditions at the start of the run.

I would start without nudging to the 3-D fields and see how bad the boundaries look using just the boundary conditions. Nudging to 3-D fields is a crude tool. If you decide that you do need 3-D nudging, you can use a spatially varying nudging coefficient to apply it only where necessary (like in a band along the boundaries).

If you use nudging as part of your boundary condition, then TNUDG and M3NUDG will be used, as well as OBCFAC (at least for the RadNud I use). I have TNUDG=360, OBCFAC=120.

Having daily boundary files does not require you to set the nudging time to one day. ROMS will query your boundary file and find out what records are available. Having daily climatology files makes for massive file preparation - I try to get away with monthly climatologies if I need them.

[kobl1201]

thank your reply.

I read your answer ,and suddenly i have other question.

What does Nudging mean??

I have understood when the model is running, it mean chganging initial data.

i mean if the model calculate 2012.01.01, intial data apply data of 2012.01.01 time.

And if model calculate 2012.01.02, the inital data apply 2012.01.02 data, when i want to apply nudge 1 day interval.

Is it right for nudging meaning??? i think i don't understand nudging meaning. Could you explain it??

Thank you again.

[kate]

If you already have the full 3-D fields for every day, what is the purpose of running the model? What are you hoping to learn if you force the model towards these 3-D fields you already have?

Perhaps it's easier to show you the math?

Code: Select all

!
!  Nudge towards tracer climatology.
!
        IF (LtracerCLM(itrc,ng).and.LnudgeTCLM(itrc,ng)) THEN 
          DO k=1,N(ng)
            DO j=JstrR,JendR
              DO i=IstrR,IendR
                t(i,j,k,nnew,itrc)=t(i,j,k,nnew,itrc)+                  &    
     &                             dt(ng)*                              &    
     &                             CLIMA(ng)%Tnudgcof(i,j,k,ic)*        &    
     &                             (CLIMA(ng)%tclm(i,j,k,ic)-           &    
     &                              t(i,j,k,nnew,itrc))
              END DO
            END DO
          END DO
        END IF

Note that CLIMA(ng)%Tnudgcof(i,j,k,ic) here is a function of space, also that its units are one over seconds. TNUDG in the input file is in days, but it gets set to 1/TNUDG before it gets used.

[kobl1201]

Thank you.

i will tell you my purpose.

I was running the model set HYCOM model for initail data. And then i checked the results.

The results for currents was bad comapred with observation data.

So i changed initial data. and the initial data apply each 1 day for 1 year.

and i will be running model and compare OBS data.

Here is my question.

How can i apply the initial data each 1 day??

And this is analytical method or not?

[kate]

Is it data assimilation of the observational data that you are planning to do? Or is it that you want to nudge to the daily HYCOM model output?

The initial conditions are only for the first timestep of a model run. It sounds like you want to restart the model every day from new initial conditions.

A model run with strong nudging to the daily HYCOM fields will look a lot like the HYCOM run. Would the HYCOM run alone be good enough for you? If so, you don't need to run the model yourself.

[kobl1201]

i will make new initial data.

This data was made a method. Firstly, i was running the Roms model set output from HYCOM model.

And then, I compared the results, which averaged daily, with observation. However, the results showed too bad.

So i thought that the results influenced used obsercational data. ( it might be concept of data assimilation of the observational data.)

Finally, i made new initail data each daily for 1 year time , and i want to be rerunning ROMS model by applying the new intial condition each daily.

This is my planning.

Among my planning, i can't understand how apply the new inital condition each 1 day.

P.S. DO you have ROMS model user guide?? Actually, I am so stuffy with studying the model.

[kate]

Do you know about the ROMS wiki?There is stuff there about data assimilation under Tutorials. I have never done data assimilation, though - it's complicated.

[kobl1201]

i want to do data assimilation of the observational data.

How can i do??? Actually, i found Roms wiki. But i couldn't understand it.

i just changed M3NUDGE=366.0d0 -> M3NUDGE=1.0d0 or can i need other work??

my climatology data was combined only temp and salinty.