[gradsusr] draw shaded color in whole area from station data

[Davide Sacchetti]

It is not an easy task
any approach you use you make errors because you have to predict
information you actually have only on stations. Then the approach you
may use is not unique. I haven't the truth ... When gridding
precipitation data on regional scale I do like this:
      * regrid precipitation data on a thin grid using oabin function
      * propagate precipitation in a uniform way inside a circle with a
        radius depending on accumulation time
      * where circle are overlapping perform a weighted mean, based on
        1/distance^2 from stations (I wrote a grads extension for this
        purpose, I can send it if you are interested)

changing type of data or scale may lead to substantial change in the
method ...

I dont' know if this may help or not ...
bye bye
Davide


On Wed, 2012-12-12 at 00:34 -0800, Chinh ta huu wrote:
> Hi,
> 
> - If we use cress man, it is required historical observed data for
> correction, Do you have other way to do that, please recommend me
> 
> Thank you very much
> Chinh
> 
> -- 
> (Mr.) Ta Huu Chinh
> National, Weather Forecasting Center, Ha Noi, Viet Nam. 
> No: 4th, Dang Thai Than, Hoan Kiem, Ha Noi
> Tel: 84-903-298-856
> Office: 84-43-824-4917
> Email: chinhth2010 at gmail.com
> Website: http://vn.myblog.yahoo.com/chinhth2005/
> Website: http://bestknowledge-chinhth.blogspot.com/
> Website: http://anhcuoidep.vn
> 
> _______________________________________________
> gradsusr mailing list
> gradsusr at gradsusr.org
> http://gradsusr.org/mailman/listinfo/gradsusr

-- 
--> Attenzione cambio indirizzo: davide.sacchetti at arpal.gov.it
--
Davide Sacchetti
Centro Funzionale Meteo Idrologico di Protezione Civile della Regione Liguria
ARPAL Unità Tecnica Complessa di livello Regionale
V.le Brigare Partigiane 2 16121 Genova (I)
tel: +39 010 6437535                    fax: +39 010 6437520
mail: davide.sacchetti at arpal.gov.it     web: www.meteoliguria.it