climatology: Hovmuller VS Line Plot

[Teddy Allen]

Dear GrADS users,
Last summer I received a host of helpful responses to a question concerning how to create a climatology to be visualized through a Hovmuller plot. I recently used that same information and applied the climatology to be visualized through a line plot time series for a different project concerning OLR. Unfortunately, the climatology that was used for the OLR line plot time series was not accurate. 
 
I confirmed that it was not accurate by simply displaying all OLR years on the same plot and it varied greatly from what the climatology produced.
 
Below is my script that attempts to create a line plot daily climatology time series of OLR from 1979-2007 at a specific location (the gridbox containing 18.75N,78W).
 
'sdfopen c:\testdata\netcdf\olr.day.mean.nc'
'set grid off'
'set grads off'
'set lat 18.75'
'set lon 282'
'set line 8 1 6'
'set cmark 0'
'set tlsupp year'
'set vrange 180 290'
**************************
* t=1 is Apr1 1979 * 
* t=10413 is Sept30 2007 *
**************************
'set t 1 182'
'define olrclim = ave(olr,t+0,t=10413,365)'
*this is for clim from Apr1-Sept30 1979-2007
*the dataset at t=1 begins on APR 1 1979
'd olrclim'
'draw title OLR climatology 1979-2007'
 
I was assuming that the below lines: 
'set t 1 182'
'define olrclim = ave(olr,t+0,t=10413,365)'
simply say to create 182 daily time steps, basically from Apr 1 to Sept30, and to calculate an average for every day from Apr - Sept. 30 within the entire 1979-2007 dataset.
 
Can anybody comment on the construction of the time series?
Thank you very much.
Sincerely,
Teddy Allen
 

web: http://personal.ecu.edu/tla0923/HOME.html
"To move quickly, go alone. To move far, go together."
African Proverb                                                                                                        *
 


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gradsusr.org/pipermail/gradsusr/attachments/20081031/f5ed4244/attachment.html