|
1 ;******************************************************** |
|
2 ;using model biome vlass |
|
3 ; |
|
4 ; required command line input parameters: |
|
5 ; ncl 'model_name="10cn" model_grid="T42" dirm="/.../ film="..."' 01.npp.ncl |
|
6 ; |
|
7 ; histogram normalized by rain and compute correleration |
|
8 ;************************************************************** |
|
9 load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl" |
|
10 load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl" |
|
11 load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/contributed.ncl" |
|
12 ;************************************************************** |
|
13 procedure set_line(lines:string,nline:integer,newlines:string) |
|
14 begin |
|
15 ; add line to ascci/html file |
|
16 |
|
17 nnewlines = dimsizes(newlines) |
|
18 if(nline+nnewlines-1.ge.dimsizes(lines)) |
|
19 print("set_line: bad index, not setting anything.") |
|
20 return |
|
21 end if |
|
22 lines(nline:nline+nnewlines-1) = newlines |
|
23 ; print ("lines = " + lines(nline:nline+nnewlines-1)) |
|
24 nline = nline + nnewlines |
|
25 return |
|
26 end |
|
27 ;************************************************************** |
|
28 ; Main code. |
|
29 begin |
|
30 |
|
31 plot_type = "ps" |
|
32 plot_type_new = "png" |
|
33 |
|
34 ;--------------------------------------------------- |
|
35 ; model name and grid |
|
36 |
|
37 model_grid = "T42" |
|
38 |
|
39 model_name = "cn" |
|
40 model_name1 = "i01.06cn" |
|
41 model_name2 = "i01.10cn" |
|
42 |
|
43 ;--------------------------------------------------- |
|
44 ; get biome data: model |
|
45 |
|
46 biome_name_mod = "Model PFT Class" |
|
47 |
|
48 dirm = "/fis/cgd/cseg/people/jeff/clamp_data/model/" |
|
49 film = "class_pft_"+model_grid+".nc" |
|
50 fm = addfile(dirm+film,"r") |
|
51 |
|
52 classmod = fm->CLASS_PFT |
|
53 |
|
54 delete (fm) |
|
55 |
|
56 ; model data has 17 land-type classes |
|
57 |
|
58 nclass_mod = 17 |
|
59 |
|
60 ;-------------------------------------------------- |
|
61 ; get model data: landmask, landfrac and area |
|
62 |
|
63 dirm = "/fis/cgd/cseg/people/jeff/surface_data/" |
|
64 film = "lnd_T42.nc" |
|
65 fm = addfile (dirm+film,"r") |
|
66 |
|
67 landmask = fm->landmask |
|
68 landfrac = fm->landfrac |
|
69 area = fm->area |
|
70 |
|
71 delete (fm) |
|
72 |
|
73 ; change area from km**2 to m**2 |
|
74 area = area * 1.e6 |
|
75 |
|
76 ;---------------------------------------------------- |
|
77 ; read data: time series, model |
|
78 |
|
79 dirm = "/fis/cgd/cseg/people/jeff/clamp_data/model/" |
|
80 film = model_name2 + "_Fire_C_1979-2004_monthly.nc" |
|
81 fm = addfile (dirm+film,"r") |
|
82 |
|
83 data_mod = fm->COL_FIRE_CLOSS(18:25,:,:,:) |
|
84 |
|
85 delete (fm) |
|
86 |
|
87 ; Units for these variables are: |
|
88 ; g C/m^2/s |
|
89 |
|
90 ; change unit to g C/m^2/month |
|
91 |
|
92 nsec_per_month = 60*60*24*30 |
|
93 |
|
94 data_mod = data_mod * nsec_per_month |
|
95 |
|
96 data_mod@unit = "gC/m2/month" |
|
97 ;---------------------------------------------------- |
|
98 ; read data: time series, observed |
|
99 |
|
100 dirm = "/fis/cgd/cseg/people/jeff/fire_data/ob/GFEDv2_C/" |
|
101 film = "Fire_C_1997-2006_monthly_"+ model_grid+".nc" |
|
102 fm = addfile (dirm+film,"r") |
|
103 |
|
104 data_ob = fm->FIRE_C(0:7,:,:,:) |
|
105 |
|
106 delete (fm) |
|
107 |
|
108 ob_name = "GFEDv2" |
|
109 |
|
110 ; Units for these variables are: |
|
111 ; g C/m^2/month |
|
112 |
|
113 ;--------------------------------------------------- |
|
114 ; take into account landfrac |
|
115 |
|
116 area = area * landfrac |
|
117 data_mod = data_mod * conform(data_mod, landfrac, (/2,3/)) |
|
118 data_ob = data_ob * conform(data_ob, landfrac, (/2,3/)) |
|
119 |
|
120 delete (landfrac) |
|
121 |
|
122 ;------------------------------------------------------------- |
|
123 ; html table1 data |
|
124 |
|
125 ; column (not including header column) |
|
126 |
|
127 col_head = (/"Model Fire_Flux (PgC/yr)" \ |
|
128 ,"Observed Fire_Flux (PgC/yr)" \ |
|
129 ,"Correlation Coefficient" \ |
|
130 ,"Ratio model/observed" \ |
|
131 ,"M_score" \ |
|
132 ,"Timeseries plot" \ |
|
133 /) |
|
134 |
|
135 ncol = dimsizes(col_head) |
|
136 |
|
137 ; row (not including header row) |
|
138 |
|
139 ; using model biome class: |
|
140 row_head = (/"Not Vegetated" \ |
|
141 ,"Needleleaf Evergreen Temperate Tree" \ |
|
142 ,"Needleleaf Evergreen Boreal Tree" \ |
|
143 ; ,"Needleleaf Deciduous Boreal Tree" \ |
|
144 ,"Broadleaf Evergreen Tropical Tree" \ |
|
145 ,"Broadleaf Evergreen Temperate Tree" \ |
|
146 ,"Broadleaf Deciduous Tropical Tree" \ |
|
147 ,"Broadleaf Deciduous Temperate Tree" \ |
|
148 ; ,"Broadleaf Deciduous Boreal Tree" \ |
|
149 ; ,"Broadleaf Evergreen Shrub" \ |
|
150 ,"Broadleaf Deciduous Temperate Shrub" \ |
|
151 ,"Broadleaf Deciduous Boreal Shrub" \ |
|
152 ,"C3 Arctic Grass" \ |
|
153 ,"C3 Non-Arctic Grass" \ |
|
154 ,"C4 Grass" \ |
|
155 ,"Corn" \ |
|
156 ; ,"Wheat" \ |
|
157 ,"All Biome" \ |
|
158 /) |
|
159 nrow = dimsizes(row_head) |
|
160 |
|
161 ; arrays to be passed to table. |
|
162 text = new ((/nrow, ncol/),string ) |
|
163 |
|
164 ;***************************************************************** |
|
165 ; (A) get time-mean |
|
166 ;***************************************************************** |
|
167 |
|
168 x = dim_avg_Wrap(data_mod(lat|:,lon|:,month|:,year|:)) |
|
169 data_mod_m = dim_avg_Wrap( x(lat|:,lon|:,month|:)) |
|
170 delete (x) |
|
171 |
|
172 x = dim_avg_Wrap( data_ob(lat|:,lon|:,month|:,year|:)) |
|
173 data_ob_m = dim_avg_Wrap( x(lat|:,lon|:,month|:)) |
|
174 delete (x) |
|
175 |
|
176 ;---------------------------------------------------- |
|
177 ; compute correlation coef |
|
178 |
|
179 landmask_1d = ndtooned(landmask) |
|
180 data_mod_1d = ndtooned(data_mod_m) |
|
181 data_ob_1d = ndtooned(data_ob_m) |
|
182 |
|
183 good = ind(landmask_1d .gt. 0.) |
|
184 ; print (dimsizes(good)) |
|
185 |
|
186 cc = esccr(data_mod_1d(good),data_ob_1d(good),0) |
|
187 ; print (cc) |
|
188 |
|
189 delete (landmask_1d) |
|
190 delete (data_mod_1d) |
|
191 delete (data_ob_1d) |
|
192 delete (good) |
|
193 |
|
194 ;---------------------------------------------------- |
|
195 ; compute M_global |
|
196 |
|
197 score_max = 1. |
|
198 |
|
199 Mscore = cc * cc * score_max |
|
200 |
|
201 M_global = sprintf("%.2f", Mscore) |
|
202 |
|
203 ;---------------------------------------------------- |
|
204 ; global res |
|
205 |
|
206 resg = True ; Use plot options |
|
207 resg@cnFillOn = True ; Turn on color fill |
|
208 resg@gsnSpreadColors = True ; use full colormap |
|
209 resg@cnLinesOn = False ; Turn off contourn lines |
|
210 resg@mpFillOn = False ; Turn off map fill |
|
211 resg@cnLevelSelectionMode = "ManualLevels" ; Manual contour invtervals |
|
212 |
|
213 ;---------------------------------------------------- |
|
214 ; global contour: model vs ob |
|
215 |
|
216 plot_name = "global_model_vs_ob" |
|
217 |
|
218 wks = gsn_open_wks (plot_type,plot_name) |
|
219 gsn_define_colormap(wks,"gui_default") |
|
220 |
|
221 plot=new(3,graphic) ; create graphic array |
|
222 |
|
223 resg@gsnFrame = False ; Do not draw plot |
|
224 resg@gsnDraw = False ; Do not advance frame |
|
225 |
|
226 ;---------------------- |
|
227 ; plot correlation coef |
|
228 |
|
229 gRes = True |
|
230 gRes@txFontHeightF = 0.02 |
|
231 gRes@txAngleF = 90 |
|
232 |
|
233 correlation_text = "(correlation coef = "+sprintf("%.2f", cc)+")" |
|
234 |
|
235 gsn_text_ndc(wks,correlation_text,0.20,0.50,gRes) |
|
236 |
|
237 ;----------------------- |
|
238 ; plot ob |
|
239 |
|
240 data_ob_m = where(landmask .gt. 0., data_ob_m, data_ob_m@_FillValue) |
|
241 |
|
242 title = ob_name |
|
243 resg@tiMainString = title |
|
244 |
|
245 resg@cnMinLevelValF = 1. |
|
246 resg@cnMaxLevelValF = 10. |
|
247 resg@cnLevelSpacingF = 1. |
|
248 |
|
249 plot(0) = gsn_csm_contour_map_ce(wks,data_ob_m,resg) |
|
250 |
|
251 ;----------------------- |
|
252 ; plot model |
|
253 |
|
254 data_mod_m = where(landmask .gt. 0., data_mod_m, data_mod_m@_FillValue) |
|
255 |
|
256 title = "Model "+ model_name |
|
257 resg@tiMainString = title |
|
258 |
|
259 resg@cnMinLevelValF = 1. |
|
260 resg@cnMaxLevelValF = 10. |
|
261 resg@cnLevelSpacingF = 1. |
|
262 |
|
263 plot(1) = gsn_csm_contour_map_ce(wks,data_mod_m,resg) |
|
264 |
|
265 ;----------------------- |
|
266 ; plot model-ob |
|
267 |
|
268 resg@cnMinLevelValF = -8. |
|
269 resg@cnMaxLevelValF = 2. |
|
270 resg@cnLevelSpacingF = 1. |
|
271 |
|
272 zz = data_ob_m |
|
273 zz = data_mod_m - data_ob_m |
|
274 title = "Model_"+model_name+" - Observed" |
|
275 resg@tiMainString = title |
|
276 |
|
277 plot(2) = gsn_csm_contour_map_ce(wks,zz,resg) |
|
278 |
|
279 ; plot panel |
|
280 |
|
281 pres = True ; panel plot mods desired |
|
282 pres@gsnMaximize = True ; fill the page |
|
283 |
|
284 gsn_panel(wks,plot,(/3,1/),pres) ; create panel plot |
|
285 |
|
286 ; system("convert "+plot_name+"."+plot_type+" "+plot_name+"."+plot_type_new+";"+ \ |
|
287 ; "rm "+plot_name+"."+plot_type) |
|
288 |
|
289 clear (wks) |
|
290 delete (plot) |
|
291 |
|
292 delete (data_ob_m) |
|
293 delete (data_mod_m) |
|
294 delete (zz) |
|
295 |
|
296 resg@gsnFrame = True ; Do advance frame |
|
297 resg@gsnDraw = True ; Do draw plot |
|
298 |
|
299 ;******************************************************************* |
|
300 ; (B) Time series : per biome |
|
301 ;******************************************************************* |
|
302 |
|
303 data_n = 2 |
|
304 |
|
305 dsizes = dimsizes(data_mod) |
|
306 nyear = dsizes(0) |
|
307 nmonth = dsizes(1) |
|
308 ntime = nyear * nmonth |
|
309 |
|
310 year_start = 1997 |
|
311 year_end = 2004 |
|
312 |
|
313 ;------------------------------------------- |
|
314 ; Calculate "nice" bins for binning the data |
|
315 |
|
316 ; using model biome class |
|
317 nclass = nclass_mod |
|
318 |
|
319 range = fspan(0,nclass,nclass+1) |
|
320 |
|
321 ; print (range) |
|
322 ; Use this range information to grab all the values in a |
|
323 ; particular range, and then take an average. |
|
324 |
|
325 nx = dimsizes(range) - 1 |
|
326 |
|
327 ;------------------------------------------- |
|
328 ; put data into bins |
|
329 |
|
330 ; using observed biome class |
|
331 ; base = ndtooned(classob) |
|
332 ; using model biome class |
|
333 base = ndtooned(classmod) |
|
334 |
|
335 ; output |
|
336 |
|
337 area_bin = new((/nx/),float) |
|
338 yvalues = new((/ntime,data_n,nx/),float) |
|
339 |
|
340 ; Loop through each range, using base. |
|
341 |
|
342 do i=0,nx-1 |
|
343 |
|
344 if (i.ne.(nx-1)) then |
|
345 idx = ind((base.ge.range(i)).and.(base.lt.range(i+1))) |
|
346 else |
|
347 idx = ind(base.ge.range(i)) |
|
348 end if |
|
349 ;--------------------- |
|
350 ; for area |
|
351 |
|
352 data = ndtooned(area) |
|
353 |
|
354 if (.not.any(ismissing(idx))) then |
|
355 area_bin(i) = sum(data(idx)) |
|
356 else |
|
357 area_bin(i) = area_bin@_FillValue |
|
358 end if |
|
359 |
|
360 ;############################################################# |
|
361 ; using model biome class: |
|
362 ; set the following 4 classes to _FillValue: |
|
363 ; (3)Needleleaf Deciduous Boreal Tree, |
|
364 ; (8)Broadleaf Deciduous Boreal Tree, |
|
365 ; (9)Broadleaf Evergreen Shrub, |
|
366 ; (16)Wheat |
|
367 |
|
368 if (i.eq.3 .or. i.eq.8 .or. i.eq.9 .or. i.eq.16) then |
|
369 area_bin(i) = area_bin@_FillValue |
|
370 end if |
|
371 ;############################################################# |
|
372 |
|
373 delete (data) |
|
374 |
|
375 ;--------------------- |
|
376 ; for data_mod and data_ob |
|
377 |
|
378 do n = 0,data_n-1 |
|
379 |
|
380 t = -1 |
|
381 do m = 0,nyear-1 |
|
382 do k = 0,nmonth-1 |
|
383 |
|
384 t = t + 1 |
|
385 |
|
386 if (n.eq.0) then |
|
387 data = ndtooned(data_ob(m,k,:,:)) |
|
388 end if |
|
389 |
|
390 if (n.eq.1) then |
|
391 data = ndtooned(data_mod(m,k,:,:)) |
|
392 end if |
|
393 |
|
394 ; Calculate average |
|
395 |
|
396 if (.not.any(ismissing(idx))) then |
|
397 yvalues(t,n,i) = avg(data(idx)) |
|
398 else |
|
399 yvalues(t,n,i) = yvalues@_FillValue |
|
400 end if |
|
401 |
|
402 ;############################################################# |
|
403 ; using model biome class: |
|
404 ; set the following 4 classes to _FillValue: |
|
405 ; (3)Needleleaf Deciduous Boreal Tree, |
|
406 ; (8)Broadleaf Deciduous Boreal Tree, |
|
407 ; (9)Broadleaf Evergreen Shrub, |
|
408 ; (16)Wheat |
|
409 |
|
410 if (i.eq.3 .or. i.eq.8 .or. i.eq.9 .or. i.eq.16) then |
|
411 yvalues(t,n,i) = yvalues@_FillValue |
|
412 end if |
|
413 ;############################################################# |
|
414 |
|
415 end do |
|
416 end do |
|
417 |
|
418 delete(data) |
|
419 end do |
|
420 |
|
421 delete(idx) |
|
422 end do |
|
423 |
|
424 delete (base) |
|
425 delete (data_mod) |
|
426 delete (data_ob) |
|
427 |
|
428 ;---------------------------------------------------------------- |
|
429 ; get area_good |
|
430 |
|
431 good = ind(.not.ismissing(area_bin)) |
|
432 |
|
433 area_g = area_bin(good) |
|
434 |
|
435 n_biome = dimsizes(good) |
|
436 |
|
437 ;---------------------------------------------------------------- |
|
438 ; data for tseries plot |
|
439 |
|
440 yvalues_g = new((/ntime,data_n,n_biome/),float) |
|
441 |
|
442 yvalues_g@units = "TgC/month" |
|
443 |
|
444 ; change unit to Tg C/month |
|
445 ; change unit from g to Tg (Tera gram) |
|
446 factor_unit = 1.e-12 |
|
447 |
|
448 yvalues_g = yvalues(:,:,good) * conform(yvalues_g,area_g,2) * factor_unit |
|
449 |
|
450 ;------------------------------------------------------------------- |
|
451 ; general settings for line plot |
|
452 |
|
453 res = True |
|
454 res@xyDashPatterns = (/0,0/) ; make lines solid |
|
455 res@xyLineThicknesses = (/2.0,2.0/) ; make lines thicker |
|
456 res@xyLineColors = (/"blue","red"/) ; line color |
|
457 |
|
458 res@trXMinF = year_start |
|
459 res@trXMaxF = year_end + 1 |
|
460 |
|
461 res@vpHeightF = 0.4 ; change aspect ratio of plot |
|
462 ; res@vpWidthF = 0.8 |
|
463 res@vpWidthF = 0.75 |
|
464 |
|
465 res@tiMainFontHeightF = 0.025 ; size of title |
|
466 |
|
467 res@tmXBFormat = "f" ; not to add trailing zeros |
|
468 |
|
469 ; res@gsnMaximize = True |
|
470 |
|
471 ;---------------------------------------------- |
|
472 ; Add a boxed legend using the simple method |
|
473 |
|
474 res@pmLegendDisplayMode = "Always" |
|
475 ; res@pmLegendWidthF = 0.1 |
|
476 res@pmLegendWidthF = 0.08 |
|
477 res@pmLegendHeightF = 0.06 |
|
478 res@pmLegendOrthogonalPosF = -1.17 |
|
479 ; res@pmLegendOrthogonalPosF = -1.00 ;(downward) |
|
480 ; res@pmLegendOrthogonalPosF = -0.30 ;(downward) |
|
481 |
|
482 ; res@pmLegendParallelPosF = 0.18 |
|
483 res@pmLegendParallelPosF = 0.23 ;(rightward) |
|
484 res@pmLegendParallelPosF = 0.73 ;(rightward) |
|
485 res@pmLegendParallelPosF = 0.83 ;(rightward) |
|
486 |
|
487 ; res@lgPerimOn = False |
|
488 res@lgLabelFontHeightF = 0.015 |
|
489 res@xyExplicitLegendLabels = (/"observed",model_name/) |
|
490 |
|
491 ;******************************************************************* |
|
492 ; (A) time series plot: monthly ( 2 lines per plot) |
|
493 ;******************************************************************* |
|
494 |
|
495 ; x-axis in time series plot |
|
496 |
|
497 timeI = new((/ntime/),integer) |
|
498 timeF = new((/ntime/),float) |
|
499 timeI = ispan(1,ntime,1) |
|
500 timeF = year_start + (timeI-1)/12. |
|
501 timeF@long_name = "year" |
|
502 |
|
503 plot_data = new((/2,ntime/),float) |
|
504 plot_data@long_name = "TgC/month" |
|
505 |
|
506 ;---------------------------------------------- |
|
507 ; time series : per biome |
|
508 |
|
509 do m = 0, n_biome-1 |
|
510 |
|
511 plot_name = "monthly_biome_"+ m |
|
512 |
|
513 wks = gsn_open_wks (plot_type,plot_name) |
|
514 |
|
515 title = "Fire : "+ row_head(m) |
|
516 res@tiMainString = title |
|
517 |
|
518 plot_data(0,:) = yvalues_g(:,0,m) |
|
519 plot_data(1,:) = yvalues_g(:,1,m) |
|
520 |
|
521 plot = gsn_csm_xy(wks,timeF,plot_data,res) |
|
522 |
|
523 ; system("convert "+plot_name+"."+plot_type+" "+plot_name+"."+plot_type_new+";"+ \ |
|
524 ; "rm "+plot_name+"."+plot_type) |
|
525 |
|
526 clear (wks) |
|
527 delete (plot) |
|
528 |
|
529 end do |
|
530 |
|
531 ;-------------------------------------------- |
|
532 ; time series: global |
|
533 |
|
534 plot_name = "monthly_global" |
|
535 |
|
536 wks = gsn_open_wks (plot_type,plot_name) |
|
537 |
|
538 title = "Fire : "+ row_head(n_biome) |
|
539 res@tiMainString = title |
|
540 |
|
541 do k = 0,ntime-1 |
|
542 plot_data(0,k) = sum(yvalues_g(k,0,:)) |
|
543 plot_data(1,k) = sum(yvalues_g(k,1,:)) |
|
544 end do |
|
545 |
|
546 plot = gsn_csm_xy(wks,timeF,plot_data,res) |
|
547 |
|
548 ; system("convert "+plot_name+"."+plot_type+" "+plot_name+"."+plot_type_new+";"+ \ |
|
549 ; "rm "+plot_name+"."+plot_type) |
|
550 |
|
551 clear (wks) |
|
552 delete (plot) |
|
553 end |
|
554 |