[Thread Prev][Thread Next][Index]

Re: [ferret_users] Re: finding monthly maximum in daily data [regriding problem]



Hi Peter,
The monthly axis you defined is a regularly-spaced axis, with grid cells approximately one month in size, but the months will be centered approximately on the first of each month. Do you want an axis with the first cell representing all of January 1961 then all of February, and so on?  If so, then you would come closer to that if you did not use the /edges qualifier.  But to really do it right, try this: Look up "monthly axis, creating" in the Users Guide index.  It'll take you to the section on DEFINE AXIS/EDGES, which has an example showing how to define a set of values representing the start of each month, and then using /EDGES to create a monthly axis where the edges are the boundaries between months and the coordinate value is in the middle.

As to the results you're getting, when you specify both a transformation, like X=@AVE,and a regridding operation, which in your case is GT=tmonthly2@MAX, together in the same _expression_, Ferret does the regridding first. As you found there's a difference between the maximum in each month of all the data in the x and y range, vs the maximum in each month of the data after averaging in x and y.  A person might want either result, so it's important to nail down in what order you want operations you want to be done, as you did in making two separate definitions.

Ansley



Peter Szabo wrote:
Hi ferreters,

one comment:
It looks proper, when i used any of the axes (now i used:
def axis/t="15-JAN-1961":"15-DEC-2000"/np=480/edges tmonthly2) and

let prex=pre[x=@ave,y=@ave]
let premmax=prex[gt=tmonthly2@max]


instead of

let premmax=pre[x=@ave,y=@ave,gt=tmonthly2@max]

Sorry, i should have mentioned the "x=@ave,y=@ave" operations.

The given values seem to be correct (the maximum value on the attached GIF are closer to the original GIF which was posted in my latter letter), but i still dont know if the used time-axis is appropriate or not.
And i dont know why i cant use the x=@ave, y=@ave and the gt=tmonthly2@max regriding operations together.

Peter

On Tue, Nov 25, 2008 at 2:52 PM, Peter Szabo <szabpet83@xxxxxxxxx> wrote:
Dear Ferret Users,

Im totally confused about creating appropriate time-axes: i would like to get monthly maximum precipitation from daily data set.
Here are some details of my original data set:

yes? sh axis/t=1-jan-1961:3-jan-1961 tax
 name       axis              # pts   start                end
 TAX       TIME             14610 r   01-JAN-1961 00:00    31-DEC-2000 00:00
T0 = 15-JAN-1901 00:00:00
   Axis span (to cell edges) = 350640
 
       L     T                   TBOX      TBOXLO                TSTEP (HOURS)
       1>  01-JAN-1961 00:00:00  24        31-DEC-1960 12:00:00   525624
       2>  02-JAN-1961 00:00:00  24        01-JAN-1961 12:00:00   525648
       3>  03-JAN-1961 00:00:00  24        02-JAN-1961 12:00:00   525672

When i try to find the monthly maximum value through regriding to monthly data with the below command, i just cant have the correct values.

def axis/t="15-JAN-1961":"15-DEC-2000"/np=480 tmonthly
let premmax=pre[gt=tmonthly@max]

I also tried creating the time axis with "/edges" and some others also.

def axis/t="15-JAN-1961":"15-DEC-2000"/np=480/edges tmonthly
def axis/t="1-JAN-1961:00:00":"31-DEC-2000:00:00"/t0="1-JAN-1961:00:00"/np=480/edges/units=days tmonthly
etc.

None of them gave what i expected.
I enclose the original daily data set (HU_predaily.gif) and one of the plots after regriding (HU_premax_tmonthly.gif). The maximum on the original plot is 33.75 and on the others they are always bigger (around 37).

Can anyone help me?

Thanks in advance,
Peter Szabo





[Thread Prev][Thread Next][Index]

Contact Us
Dept of Commerce / NOAA / OAR / PMEL / TMAP

Privacy Policy | Disclaimer | Accessibility Statement