Re: How to define grided variable from point data

[<Mick.Spillane@xxxxxxxx>]

On Mon, 15 Nov 2004, Yogesh K. Tiwari wrote:
> I have the point data like :
> title         I        J      K        L
> CO2_sat      --       --      --    1:1860206
> lat          --       --      --    1:1860206
> lon          --       --      --    1:1860206
> year         --       --      --    1:1860206
>   mo          --       --      --    1:1860206
> day          --       --      --    1:1860206
> hour         --       --      --    1:1860206
> minut        --       --      --    1:1860206
> here L is number of index.
> ++++ I want to compute the average column burdon of
>       CO2_sat as a function of latitude, long, time.

Hi Yogesh,
	While there would be ways based on the various "scat2grid"
functions, a simple minded approach in which a series of masks are
used to isolate regions within the region, may be appropriate.
	Separate zonal, meridional, and time averages are easy to
illustrate, as in the demo below, but 2-d or 3-d masks would be
possible with nested "repeat" loops.
	For example the zonal averages at 85S, 75S, ... 85N in a
globe-spanning dataset could be found using:

let lat0=-90 ; let dlat=10
sp touch zoneav.d          ! prepare file to accept values
repeat/k=1:18 (go zoneav)  ! put results into the file using append
def axis/y=85S:85N:10/units=latitude ylat  ! recover results
def grid/y=ylat glat
file/form=free/g=glat/var=mylat,myav,npts zoneav.d
plot myav

where you have previously created the following script.

!********************* zoneav.jnl ***************************
let lat1=(`k`-1)*dlat + lat0
let lat2=lat1+dlat
let co2mask=if(lat ge lat1 and lat lt lat2)then co2_sat
list/nohead/app/form=(f5.1,e12.4,f8.0)/file=zoneav.d \
                  lat1+dlat/2,co2mask[l=@ave],co2mask[l=@ngd]
!************************************************************