[ncl-talk] Significance test for multi-model ensemble in panel plot

[Deniz Bozkurt]

Dear NCL-talk,

Attached is the low-resolution panel plot showing MSLP trend for reanalysis and 41 individual CMIP5 models as well as the ensemble mean of 41 models (last member of the panel plot). I am able to draw each plot and then hatch grid points having significant trends for reanalysis and 41 individual models with these commands:

nplots = 43
addmark = new(nplots,graphic)

do i=0,nplots-1
plots(i) = gsn_csm_contour_map(wks,slope2d(case|i,latitude|:,longitude|:),res) ;case is the number of processed data 
txres               = True
txres at txFontHeightF = 0.06
dum = gsn_add_text(wks,plots(i),allnames(i),-90,-90,txres) ;add text label for each plot
end do

do i = 0,nplots-2
mkres = True
mkres at gsMarkerIndex = 2
mkres at gsMarkerSizeF = 0.0006;0.0035
mkres at gsMarkerThicknessF = 0.007;0.025
mkres at gsMarkerColor = "grey20"
z_signif_1d = ndtooned(z_signif_mask(i,:,:))
ind_zsig = ind(z_signif_1d.ge.prob_t)
print (dimsizes(ind_zsig))
addmark(i) = gsn_add_polymarker(wks,plots(i),glon(ind_zsig),glat(ind_zsig),mkres)
delete(ind_zsig)
end do


My question is that how can I hatch the grids where e.g. 50% of the individual models have significant trends in the ensemble mean plot? As I delete “ind_zsig” variable (indice of significance test) end of each model process, I am stuck to create a loop for an array counting the number of models with significant trends in each grid. 

I hope I made a clear explanation.

Thanks in advance,
Deniz

--
Deniz Bozkurt, PhD
Post-doctoral Researcher
Center for Climate and Resilience Research
Department of Geophysics, University of Chile
Blanco Encalada 2002, Santiago, Chile
E-mail: dbozkurt at dgf.uchile.cl <mailto:dbozkurt at dgf.uchile.cl>, deniboz at gmail.com <mailto:deniboz at gmail.com>Web: http://dgf.uchile.cl/~dbozkurt/ <http://dgf.uchile.cl/~dbozkurt/>



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20161214/ef57f986/attachment-0001.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: CMIP5_1980_2005_mslp_trendDJF.png
Type: image/png
Size: 1010351 bytes
Desc: not available
Url : http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20161214/ef57f986/attachment-0001.png