/
canesm_LE_regress.py
376 lines (274 loc) · 12.5 KB
/
canesm_LE_regress.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
import loadLE as le
import cccmaplots as cplt
import cccmautils as cutl
import canesm_LE_general as leg
import loadmodeldata as lmd
import pandas as pd
import constants as con
local=True
timeselc='1979-01-01,1989-12-31'
timeselp='2002-01-01,2012-12-31'
timeselall = '1979-01-01,2012-12-31'
ftype='fullts' # 'fullclimo' or 'climo' or 'fullts'
# this is not implemented here. but would do regressions in time on LE ensemble avg
ensmean=False
seasp='DJF' # season of spatial field
sear='DJF' #'DJF' # season of regional avgs
# spatial field1 in color
leconvsp=1
fieldsp='tas'; ncfieldsp='tas'; compsp='Amon';
cminsp=-1; cmaxsp=1 # for colors
cminspa=-1; cmaxspa=1 # for colors for AGCM
# spatial field2 in contours
leconvsp2=1
fieldsp2='zg50000.00'; ncfieldsp2='zg'; compsp2='Amon'
cminsp2=-10; cmaxsp2=10 # to calc contour interval
cminsp2a=-10; cmaxsp2a=10 # to calc contour interval for AGCM
# regional avg field 1
leconvr=-1; aconvr=-1 # this way, sea ice loss is linked with positive changes elsewhere
#fieldr='sic'; ncfieldr='sic'; compr='OImon'; regionr='bksmori';
fieldr='zg50000.00'; ncfieldr='zg'; compr='Amon'; regionr='bksmori'; leconvr=1; aconvr=1 # @@@@
# regional avg field 2
#leconvr2=-1; aconvr2=-1 # so cooling=high heights
#fieldr2='tas'; ncfieldr2='tas'; compr2='Amon'; regionr2='eurasiamori';
leconvr2=-1; aconvr2=-1 # so loss=high heights
fieldr2='sic'; ncfieldr2='sic'; compr2='OImon'; regionr2='bksmori';
fdictsp = {'field': fieldsp, 'ncfield': ncfieldsp, 'comp': compsp}
fdictsp2 = {'field': fieldsp2, 'ncfield': ncfieldsp2, 'comp': compsp2}
fdictr = {'field': fieldr+regionr, 'ncfield': ncfieldr, 'comp': compr}
fdictr2 = {'field': fieldr2+regionr2, 'ncfield': ncfieldr2, 'comp': compr2}
lat=le.get_lat(local=local)
lon=le.get_lon(local=local)
nlat=len(lat); nlon=len(lon)
def load_field(fdict,casename,timesel,seas,ftype='fullts',conv=1,local=False,verb=False):
"""
returns [numens x space.flat] or [numens]
"""
ledat = le.load_LEdata(fdict,casename,timesel=timesel,
rettype='ndarray',conv=conv,ftype=ftype,local=local,verb=verb)
print '@@@ ledat.shape ' + str(ledat.shape) # why does the 3d data have space flattened already...
# time needs to be first dimension
try:
if ledat.ndim==2:
ledat = ledat.T
elif ledat.ndim==3:
ledat = np.transpose(ledat,(1,0,2))
else:
print 'Loaded data is not 2 or 3 dimensions. Do not understand.'
raise Exception
except:
raise
lesea = cutl.seasonalize_monthlyts(ledat,season=seas).mean(axis=0) # numens x space.flat
return lesea
def load_agcmfield(field,sims,seas,conv=1,region=None,subsampyrs=11,styears=None):
""" loads subsampled agcm data from specified simulations
number of total samples will be determined by length of all sim data
and numyrs (e.g. (ntime / numyrs)*numsims)
returns subsamp
styears
"""
threed=False
simconv=1
if field=='tas': simfield='st'; simncfield='ST'
elif field=='zg50000.00': simfield='gz50000'; simncfield='PHI'; simconv=1/con.get_g()
elif field=='sia': simfield='sicn'; simncfield='SICN'; print '@@ danger, sia actually sicn average'
elif field=='sic': simfield='sicn'; simncfield='SICN'
else: print 'cannot addsims for ' + field
if region!=None:
simflddf = pd.DataFrame(lmd.loaddata((simfield,),sims,ncfields=(simncfield,), timefreq=seas,
region=region))*simconv
else:
# assume threed b/c no region given.
threed=True
simflddf = lmd.loaddata((simfield,),sims,ncfields=(simncfield,), timefreq=seas,
region=region,rettype='ndarray')*simconv
subsamp,styearsss = leg.subsamp_sims(simflddf,numyrs=subsampyrs,styears=styears,threed=threed)
return subsamp,styearsss
def slopemap(inr,insp,dims):
"""
inr is 1D [time or numens]
insp is 2D [time or numens x space.flat]
dims are a tuple of dims to reshape space to (nlat,nlon)
returns slopemap [dims]
"""
slope,intercept = np.polyfit(inr,insp, 1)
slopemap = slope.reshape(dims)
return slopemap
#casenames=('historical','historicalNat','historicalMisc')
casenames=('historical',)
for casename in casenames:
# # SPATIAL DATA
print 'SPATIAL'
lecseasp = load_field(fdictsp,casename,timeselc,seasp,ftype=ftype,conv=leconvsp)
lepseasp = load_field(fdictsp,casename,timeselp,seasp,ftype=ftype,conv=leconvsp)
leseasp = lepseasp-lecseasp
lecseasp2 = load_field(fdictsp2,casename,timeselc,seasp,ftype=ftype,conv=leconvsp2)
lepseasp2 = load_field(fdictsp2,casename,timeselp,seasp,ftype=ftype,conv=leconvsp2)
leseasp2 = lepseasp2-lecseasp2
# # 1D DATA
print '1D'
lecsear = load_field(fdictr,casename,timeselc,sear,ftype=ftype,conv=leconvr)
lepsear = load_field(fdictr,casename,timeselp,sear,ftype=ftype,conv=leconvr)
lesear = (lepsear-lecsear)/(lepsear-lecsear).std()
lecsear2 = load_field(fdictr2,casename,timeselc,sear,ftype=ftype,conv=leconvr2)
lepsear2 = load_field(fdictr2,casename,timeselp,sear,ftype=ftype,conv=leconvr2)
lesear2 = (lepsear2-lecsear2)/(lepsear2-lecsear2).std()
if casename!='historical': # really just has to be not the first loop thru
tmp = np.vstack((tmp,leseasp))
tmp2 = np.vstack((tmp2,leseasp2))
tmpr = np.hstack((tmpr,lesear))
tmpr2 = np.hstack((tmpr2,lesear2))
else:
tmp = leseasp
tmp2 = leseasp2
tmpr = lesear
tmpr2 = lesear2
leseasp=tmp
leseasp2=tmp2
lesear=tmpr
lesear2=tmpr2
# calc regression slopes
fldsponfldr = slopemap(lesear,leseasp,(nlat,nlon)) # SAT regress on regSIC
fldsp2onfldr = slopemap(lesear,leseasp2,(nlat,nlon)) # Z500 regress on regSIC
fldsponfldr2 = slopemap(lesear2,leseasp,(nlat,nlon)) # SAT regress on regSAT
fldsp2onfldr2 = slopemap(lesear2,leseasp2,(nlat,nlon)) # Z500 regress on regSAT
# === AGCM ==========
simsE=('E1','E2','E3','E4','E5');
sims=('R1','R2','R3','R4','R5');
asseasp,styears = load_agcmfield(fieldsp,sims,seasp) # already anomalies
asseasp2,styears = load_agcmfield(fieldsp2,sims,seasp,styears=styears) # already anomalies
assear,styears = load_agcmfield(fieldr,sims,sear,styears=styears,region=regionr) # already anomalies
assear2,styears = load_agcmfield(fieldr2,sims,sear,styears=styears,region=regionr2) # already anomalies
assear=assear / assear.std()
assear2=assear2 / assear2.std()
# Make panel (d) with E sims:
aesseasp,styearse = load_agcmfield(fieldsp,simsE,seasp)
aesseasp2,styearse = load_agcmfield(fieldsp2,simsE,seasp,styears=styearse)
aessear2,styearse = load_agcmfield(fieldr2,simsE,sear,styears=styearse,region=regionr2)
aessear2=aessear2 / aessear2.std()
(anens,anlat,anlon)=asseasp.shape
rshape=(anens,anlat*anlon)
# calc regression slopes: multiply regional avgs by -1 to get colors/signs right.
# this was accounted for in LE data upon loading.
asponfldr = slopemap(assear*aconvr,asseasp.reshape(rshape),(anlat,anlon)) # SAT regress on regSIC
asp2onfldr = slopemap(assear*aconvr,asseasp2.reshape(rshape),(anlat,anlon)) # Z500 regress on regSIC
asponfldr2 = slopemap(assear2*aconvr2,asseasp.reshape(rshape),(anlat,anlon)) # SAT regress on regSAT
asp2onfldr2 = slopemap(assear2*aconvr2,asseasp2.reshape(rshape),(anlat,anlon)) # Z500 regress on regSAT
aesponfldr2 = slopemap(aessear2*aconvr2,aesseasp.reshape(rshape),(anlat,anlon)) # SAT regress on regSAT
aesp2onfldr2 = slopemap(aessear2*aconvr2,aesseasp2.reshape(rshape),(anlat,anlon)) # Z500 regress on regSAT
# ====================== FIGURES ===============
printtofile=False
lons, lats = np.meshgrid(lon,lat)
cmlen=15.
incr = (cmaxsp2-cminsp2) / (cmlen)
conts = np.arange(cminsp2,cmaxsp2+incr,incr)
ttl1=seasp + ' regress on ' + sear + ' BKS Z500'
ttl2=seasp + ' regress on ' + sear + ' Eur SAT'
#ttl1=ttl2=''
fig,axs=plt.subplots(1,2)
fig.set_size_inches(10,5)
fig.subplots_adjust(wspace=0.05)
ax=axs[0]
bm,pc=cplt.kemmap(fldsponfldr,lat,lon,ptype='nheur',axis=ax,cmin=cminsp,cmax=cmaxsp,
title=ttl1,suppcb=True,
panellab='a',lcol='0.2')
bm.contour(lons,lats,fldsp2onfldr,levels=conts,
colors='0.5',linewidths=1,latlon=True)
ax=axs[1]
bm,pc=cplt.kemmap(fldsponfldr2,lat,lon,ptype='nheur',axis=ax,cmin=cminsp,cmax=cmaxsp,
title=ttl2,suppcb=True,
panellab='b',lcol='0.2')
bm.contour(lons,lats,fldsp2onfldr2,levels=conts,
colors='0.5',linewidths=1,latlon=True)
cplt.add_colorbar(fig,pc,orientation='horizontal')
if printtofile:
fig.savefig(fieldsp + '_' + fieldsp2 + seasp + \
'_regresson_' + fieldr+regionr + '_' + fieldr2 + regionr2 + sear + '.pdf')
# ========== AGCM
alat=con.get_t63lat(); alon=con.get_t63lon()
alons, alats = np.meshgrid(alon,alat)
cmlen=15.
incra = (cmaxsp2a-cminsp2a) / (cmlen)
contsa = np.arange(cminsp2a,cmaxsp2a+incra,incra)
fig,axs=plt.subplots(1,2)
fig.set_size_inches(10,5)
fig.subplots_adjust(wspace=0.05)
ax=axs[0]
bm,pc=cplt.kemmap(asponfldr,alat,alon,ptype='nheur',axis=ax,cmin=cminspa,cmax=cmaxspa,
title=ttl1,suppcb=True,
panellab='a',lcol='0.2')
bm.contour(alons,alats,asp2onfldr,levels=contsa,
colors='0.5',linewidths=1,latlon=True)
ax=axs[1]
bm,pc=cplt.kemmap(asponfldr2,alat,alon,ptype='nheur',axis=ax,cmin=cminspa,cmax=cmaxspa,
title=ttl2,suppcb=True,
panellab='b',lcol='0.2')
bm.contour(alons,alats,asp2onfldr2,levels=contsa,
colors='0.5',linewidths=1,latlon=True)
cplt.add_colorbar(fig,pc,orientation='horizontal')
if printtofile:
fig.savefig(fieldsp + '_' + fieldsp2 + seasp + \
'_regresson_' + fieldr+regionr + '_' + fieldr2 + regionr2 + sear + '_AGCM.pdf')
# ========== 4 panel fig =========
fig,axs=plt.subplots(2,2)
fig.set_size_inches(10,10)
fig.subplots_adjust(wspace=0.05,hspace=0.05)
ax=axs[0,0]
bm,pc=cplt.kemmap(fldsponfldr,lat,lon,ptype='nheur',axis=ax,cmin=cminsp,cmax=cmaxsp,
title=ttl1,suppcb=True,
panellab='a',lcol='0.2')
bm.contour(lons,lats,fldsp2onfldr,levels=conts,
colors='0.5',linewidths=1,latlon=True)
ax.set_ylabel('CGCM')
ax=axs[0,1]
bm,pc=cplt.kemmap(fldsponfldr2,lat,lon,ptype='nheur',axis=ax,cmin=cminsp,cmax=cmaxsp,
title=ttl2,suppcb=True,
panellab='b',lcol='0.2')
bm.contour(lons,lats,fldsp2onfldr2,levels=conts,
colors='0.5',linewidths=1,latlon=True)
cplt.add_colorbar(fig,pc,orientation='horizontal')
ax=axs[1,0]
bm,pc=cplt.kemmap(asponfldr,alat,alon,ptype='nheur',axis=ax,cmin=cminspa,cmax=cmaxspa,
title='',suppcb=True,
panellab='c',lcol='0.2')
bm.contour(alons,alats,asp2onfldr,levels=contsa,
colors='0.5',linewidths=1,latlon=True)
ax.set_ylabel('AGCM var ICE')
ax=axs[1,1]
bm,pc=cplt.kemmap(asponfldr2,alat,alon,ptype='nheur',axis=ax,cmin=cminspa,cmax=cmaxspa,
title='',suppcb=True,
panellab='d',lcol='0.2')
bm.contour(alons,alats,asp2onfldr2,levels=contsa,
colors='0.5',linewidths=1,latlon=True)
cplt.add_colorbar(fig,pc,orientation='horizontal')
if printtofile:
fig.savefig(fieldsp + '_' + fieldsp2 + seasp + \
'_regresson_' + fieldr+regionr + '_' + fieldr2 +\
regionr2 + sear + '_CGCMAGCM.pdf')
# PANEL D with E sims & residual: --------------
fig,axs=plt.subplots(1,2)
fig.set_size_inches(10,5)
ax=axs[0]
#ax=axs[2,1]
#bbox=ax.get_position()
#ax=fig.add_axes([bbox.x0,bbox.y0-bbox.height,bbox.width,bbox.height])
bm,pc=cplt.kemmap(aesponfldr2,alat,alon,ptype='nheur',axis=ax,cmin=cminsp,cmax=cmaxsp,
title=ttl2,suppcb=True,
panellab='e',lcol='0.2')
bm.contour(alons,alats,aesp2onfldr2,levels=conts,
colors='0.5',linewidths=1,latlon=True)
#cplt.add_colorbar(fig,pc,orientation='horizontal')
ax.set_ylabel('AGCM const ICE')
#fig,axs=plt.subplots(1,1)
#fig.set_size_inches(5,5)
ax=axs[1]
bm,pc=cplt.kemmap(asponfldr2-aesponfldr2,alat,alon,ptype='nheur',axis=ax,cmin=cminsp,cmax=cmaxsp,
title='residual: var - const',suppcb=True,
panellab='f',lcol='0.2')
bm.contour(alons,alats,asp2onfldr2-aesp2onfldr2,levels=conts,
colors='0.5',linewidths=1,latlon=True)
cplt.add_colorbar(fig,pc,orientation='horizontal')
if printtofile:
fig.savefig(fieldsp + '_' + fieldsp2 + seasp + \
'_regresson_' + fieldr+regionr + '_' + fieldr2 +\
regionr2 + sear + '_AGCM_E_R-Eresid.pdf')