EOF and weight

Hi NCL users,

Below is a piece of NCL script from "EOF_1.ncl", in which the "weight" is
calculated as:

wSLP = SLP*conform(SLP, clat, 1)The weight is understandable for "area
weighted average", but for EOF analysis, the weight application here is just
making the value in high latitude relatively smaller, instead of reducing the
number of grids accordingly (the typical way for EOF analysis). I may
misunderstand it. Can anyone explain a bit more about the EOF and weight?

Thanks a lot!
Mark

; =================================================================
; create weights: sqrt(cos(lat)) [or sqrt(gw) ]
; =================================================================
  rad = 4.*atan(1.)/180.
  clat = f->lat
  clat = sqrt( cos(rad*clat) ) ; gw for gaussian grid

; =================================================================
; weight all observations
; =================================================================
  wSLP = SLP ; copy meta data
  wSLP = SLP*conform(SLP, clat, 1)
  wSLP@long_name = "Wgt: "+wSLP@long_name

; =================================================================
; Reorder (lat,lon,time) the *weighted* input data
; Access the area of interest via coordinate subscripting
; =================================================================
  x = wSLP({lat|latS:latN},{lon|lonL:lonR},time|:)

  eof = eofunc_Wrap(x, neof, optEOF)

      
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Tue Mar 29 10:57:36 2011