[gradsusr] daily anomaly

Fri Sep 7 17:40:41 EDT 2012

Hi Sujata,

Actually your problem is solved by computing the climatology with few harmonics as described: 

You have 39 years of data. You can create a 39 year-long climatology with few harmonics as we described in the article. In this way you do not need to worry about leap years and missing values. And the daily climatological cycle is much smoother than computed in the usual way (see Fig.6). 

The implementation is simple, at every grid point, estimate 39-year long climatology by equation (3) of the article. Substitute P=365.25, H=4, N=total number of days in 39 years. 

Once you have produced climatology of 39 years you automatically can produce the 39 year-long anomalies. Usually 4 harmonics are good, but you can find a better number by re-producing Fig.2 for your data.

I see Muthu’s response in creating separate climatology for leap years and in this way also you have to construct 39 year-long climatology (I referred this as simple averaging in the article). In this way you only have 10 data points for Feb 29, whereas for all other days you have 39 data points with the assumption that you have no missing values. Anyway, in the article you will find an in-depth analysis about the issues we face while estimating the climatology by simple averaging. 

Hope this helps and let me know if you have any questions.

Best regards,
Bala
-------------- next part --------------
Sujata,
Create two different climatology data sets, one for 365 days and another
for leap year with 366 days.
Set up repeated soft links for each year of data set respectively to
non-leap years and leap years
in analysis period (very easy to do, couple of lines of unix scripts).
Create a single grads ctl file
for all these climo years. Open up your total field control field and this
climo control file, then you are in business. We did like this at CPC more
than two decades ago and we continue to do like this now as we work with
daily cdas fields. I hope this helps. Please email me separately if you
need further assistance. Good day!
Muthu

On Fri, Sep 7, 2012 at 2:52 AM, Sujata Mandke <amin at tropmet.res.in> wrote:

>   Respected Narapusetty Sir,
>
>      Thanks for prompt reply and willingness to provide
>    further assistance in my problem. I will go thorugh your
>    article and then try to solve the problem.
>
>          However, you had solved the problem by
> calculating climatology. I am having longterm daily
>  climatology dataset, so there is no need to calculate climatology.
> Regards
> Sujata Mandke
> IITM,PUNE-8,INDIA.
>
> ----- Original Message -----
> From: Balachandrudu Narapusetty <bnarapus at gmu.edu>
> To: GrADS Users Forum <gradsusr at gradsusr.org>
> Cc: gradsusr at gradsusr.org
> Sent: Thu, 06 Sep 2012 20:10:31 +0530 (IST)
> Subject: Re: [gradsusr] daily anomaly
>
> Hi Dr. Sujata,
>
> I am happy to see somebody has same problem that I had some time ago. I
> solved it by estimating the climatology (and thereafter anomalies) using
> few harmonics.
>
> I am sure this article is useful in dealing with your problem:
>
> Narapusetty, Balachandrudu, Timothy DelSole, Michael K. Tippett, 2009:
> Optimal Estimation of the Climatological Mean. J. Climate, 22, 4845–4859.
> doi: http://dx.doi.org/10.1175/2009JCLI2944.1
>
> I am happy to provide further assistance if you have questions.
>
> Best regards,
> Bala
>
>
> ----- Original Message -----
> From: Sujata Mandke <amin at tropmet.res.in>
> Date: Thursday, September 6, 2012 7:24 am
> Subject: [gradsusr] daily anomaly
>
> > Dear Grads Users
> >    I wish to calculate daily OLR anomaly from
> >  the following datasets.
> >    I have daily OLR data (netcdf) for the period
> >   1Jun1974-31May2012. Also long term daily OLR
> >   climatology (netcdf) data for 365 days.
> >
> >   As such writting script for calculating daily
> >  anomaly is very simple. But how to tackle following
> >    two problems while calculating anomaly
> >
> > (i) Actual data is on Gregorian(real) calendar
> >    and thus has data for 366 days in leap years.
> > (ii) What to do with missing values?
> >
> > Does anybody has grads script to calculate daily
> > anomaly by taking care of above two problems.
> > Regards
> > Dr. Sujata Mandke,Scientist,
> > IITM,PUNE-8,INDIA.
> > _______________________________________________
> > gradsusr mailing list
> > gradsusr at gradsusr.org
> > http://gradsusr.org/mailman/listinfo/gradsusr
> >
>
> _______________________________________________
> gradsusr mailing list
> gradsusr at gradsusr.org
> http://gradsusr.org/mailman/listinfo/gradsusr
>
>
> _______________________________________________
> gradsusr mailing list
> gradsusr at gradsusr.org
> http://gradsusr.org/mailman/listinfo/gradsusr
>

-- 

------------------------------------------------------------------
Dr.Muthuvel Chelliah, eMail:Muthuvel.Chelliah at noaa.gov
Rm.3008, Climate Prediction Center, Ph:(301) 683-3393
NCEP/NWS/NOAA/U.S.Dept.of Commerce,
5830, University Research Court, NCWCP
College Park, MD 20740.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gradsusr.org/pipermail/gradsusr/attachments/20120907/27cefb57/attachment-0003.html 
-------------- next part --------------
_______________________________________________
gradsusr mailing list
gradsusr at gradsusr.org
http://gradsusr.org/mailman/listinfo/gradsusr