-
Notifications
You must be signed in to change notification settings - Fork 0
/
xr_functions.py
7831 lines (7714 loc) · 433 KB
/
xr_functions.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
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# import packages
import os
import shutil
import glob
import json
import traceback
import gzip
import itertools
import math
import random
import string
from datetime import datetime, date
from pathlib import Path
import netCDF4 as nc
import xarray as xr
import xesmf as xe
import rioxarray as rxr
import cftime as cft
import numpy as np
from numpy.polynomial.polynomial import polyfit
import pandas as pd
from pandas import option_context
from reportlab.lib import utils
from reportlab.lib.units import inch
from reportlab.pdfbase.pdfmetrics import stringWidth
from reportlab.lib.pagesizes import letter
from reportlab.pdfgen import canvas
from matplotlib import pyplot as plt
import matplotlib.cm as cm
import matplotlib as matplotlib
import matplotlib.path as mpath
from cycler import cycler
import cartopy.crs as ccrs
import cartopy
import seaborn as sns
import nc_time_axis
import docx
import textwrap
import dask.config
from numcodecs import Blosc, Zlib, Zstd
from functools import partial
from cycler import cycler
import zipfile
import scipy
from scipy.optimize import curve_fit
from plotnine import (
ggplot,
aes,
geom_point,
geom_line,
geom_errorbar,
geom_smooth,
geom_ribbon,
stat_smooth,
element_rect,
xlim,
ylim,
scale_fill_manual,
scale_color_manual,
scale_color_identity,
scale_x_continuous,
scale_y_continuous,
scale_x_datetime,
labs,
guides,
guide_legend,
theme,
theme_bw,
element_text,
element_line,
element_blank
)
from mizani.breaks import date_breaks
from mizani.formatters import date_format
###########################################################################################################################
# General
###########################################################################################################################
# check directory paths exist, if not error out
def dir_check(string):
if Path(string).is_dir():
return string
else:
raise NotADirectoryError(string)
# remove file if it already exists, otherwise skip
def rmv_file(string):
try:
Path(string).unlink()
except OSError:
pass
# remove entire directory
def rmv_dir(string):
try:
if Path(string).exists():
if os.path.islink(string):
os.unlink(string)
else:
shutil.rmtree(string)
except Exception as error:
print(error)
pass
# create rand string generator
def rand_str_gen(count=1000):
# create empty list for generated strings
generated_strs = set()
while True:
if len(generated_strs) == count:
return
# create random 16 digit string
candidate_str = ''.join(random.choices(string.ascii_letters + string.digits, k=16))
# check if in generated
if candidate_str not in generated_strs:
generated_strs.add(candidate_str)
yield candidate_str
# mv old files within /projects folder
def subdir_list(cur_dir):
try:
dir_list = [x[0] for x in os.walk(cur_dir, topdown=False)]
except Exception as error:
print(error)
return dir_list
# mv old files within /projects folder
def mv_subdir_list(cur_dir, new_dir):
try:
# call string generator
dir_list = subdir_list(cur_dir)
str_gen = rand_str_gen(len(dir_list)*2)
# loop through new project directory for list of sub dir
move_list = []
for i in dir_list:
# generate unique string and add to scratch dir
new_out = new_dir + next(str_gen)
# add to move list
move_list.append([i, new_out])
except Exception as error:
print(error)
return move_list
# mv old files within /projects folder
def mv_dir(cur_dir, new_dir):
try:
# make project directory folder
Path(new_dir).mkdir(parents=True, exist_ok=True)
# move regional folder to new project location
shutil.move(cur_dir, new_dir)
except Exception as error:
print(error)
pass
# read config file from sys.arg
def read_config(sys_argv_file):
with open(sys_argv_file) as f:
config = json.load(f)
return config
# function to extract position of all sublists within a list
def extract_sublist(lst, position):
return [item[position] for item in lst]
# convert RH to SH; RH (%; 0-100), tair (kelvin), pressure (Pa)
def specific_humidity(rh, tair, pres):
# constants
Lv = 2.5*10**6 # latent heat of vaporization
Rv = 461.0 # gas constant for water vapor
T0 = 273.15 # Temperature offset for Kelvin scale
es0 = 6.112 # saturation vapor pressure at 0 deg C
# calculations borrowed from Jing
pres_hPa = pres*0.01
x = (Lv/Rv)*((1.0/T0)-1.0/tair)
es = es0**x
sh = 0.622*(rh/100.0)*es/pres_hPa
return sh
# make graph circular
def add_circle_boundary(ax):
# Compute a circle in axes coordinates, which we can use as a boundary
# for the map. We can pan/zoom as much as we like - the boundary will be
# permanently circular.
theta = np.linspace(0, 2*np.pi, 100)
center, radius = [0.5, 0.5], 0.5
verts = np.vstack([np.sin(theta), np.cos(theta)]).T
circle = mpath.Path(verts * radius + center)
ax.set_boundary(circle, transform=ax.transAxes)
# functions to select seasons
def is_summer(month):
return (month >= 5) & (month <= 9)
def is_maes_summer(month):
return (month >= 6) & (month <= 8)
def is_winter(month):
return (month > 10) | (month < 4 )
# functions to select time since warming onset
def is_initial(year):
return (year <= 5)
def is_middle(year):
return (year > 5) & (year <= 10)
def is_longterm(year):
return (year > 10) & (year <= 15)
# calculate RSME across models for all grids
def root_mean_squared_error(dataset, var, agg_over):
ds_means = dataset[var].mean(dim=agg_over)
rsme = np.sqrt(((dataset[var] - ds_means)**2).mean(dim=agg_over))
return rsme
# shift array
def shift5(arr, num, fill_value=np.nan):
result = np.empty_like(arr)
if num > 0:
result[:num] = fill_value
result[num:] = arr[:-num]
elif num < 0:
result[num:] = fill_value
result[:num] = arr[-num:]
else:
result[:] = arr
return result
###########################################################################################################################
# subset/resamp/concat output files
###########################################################################################################################
# collect netcdf file names in each given input directory,
# create raw and subset file names based on config.json
def xr_files(config):
raw_file_info = []
cat_file_info = []
for f_dir in config['dir']['in']:
# create list of netcdfs in directory
files = sorted(glob.glob("{}*{}*.nc".format(f_dir,config['filename_glob_segment'])))
# create output directory from input basename and output directory
base_name = Path(f_dir).parents[config['parent_dir_num']].name
# create directory for subset/compressed files
Path(config['dir']['out'],base_name).mkdir(parents=True, exist_ok=True)
# create list of all files
files_for_cat = []
# debug print - print file names to file
with open(Path(config['dir']['out'], base_name,'debug_output.txt'),"a") as printfile:
for line in files:
printfile.write(line)
#loop through individual files
for f in files:
# create raw file info for subset/resample
sub_file = Path(config['dir']['out'],'tmp', Path(f).stem + '_tempsub.nc')
out_file = Path(config['dir']['out'], base_name, Path(f).name)
raw_file_info.append([f, sub_file, out_file])
# create list of subset/resampled filenames (not paths) for xr_mfopendata concatenation
cat_file_name = config['dir']['out'] + base_name + "/" + Path(f).name
files_for_cat.append(cat_file_name)
# make cat file names
cat_name = [Path(f_dir).parents[config['parent_dir_num']].name.split(config['filename_splitby_character'])[i] for i in config['filename_segement_select']]
cat_file = Path(config['dir']['out'],'concat_files','_'.join(cat_name)+"_concat.nc")
# add each line to subfile list
cat_file_info.append([files_for_cat, cat_file])
# return file info list
return [raw_file_info, cat_file_info]
# identify grids of certain parameters
def grid_info(config):
# open file of 1901 simulation with PCT info
init_file = Path(config['dir']['in'][0],config['initfile'])
csv_dir = config['dir']['out']+'concat_files'
# open netcdf file
with xr.open_dataset(init_file, engine=config['nc_read']['engine']) as ds_tmp:
ds = ds_tmp.load()
# select grid cells - here we subset by ALTMAX less than 5 meters
ds1 = ds[["ALTMAX","lat"]].sel(time=config['sim_start'])
ds2 = ds1.to_dataframe()
grids = ds2[(ds2.ALTMAX < 5) & (ds2.lat > 50)].index.get_level_values('lndgrid').unique()
df = pd.DataFrame(grids)
df.to_csv(Path(csv_dir,'grid_subset.csv'))
# output lat lon by lndgrid
sub_list = ["lon","lat"]
ds[sub_list].to_dataframe().to_csv(Path(csv_dir,'grid_gps.csv'))
sub_list = ["ALTMAX"]
ds[sub_list].sel(time=config['sim_start']).to_dataframe().to_csv(Path(csv_dir,'grid_init_altmax.csv'))
ds.close()
# open surfdata for sand,silt,organic
ds_sf = xr.open_dataset(Path(config['surfdata']),\
engine=config['nc_read']['engine'])
sub_list = ["PCT_SAND","PCT_CLAY"]
ds_sf[sub_list].to_dataframe().to_csv(Path(csv_dir,'grid_texture.csv'))
sub_list = ["ORGANIC"]
ds_sf[sub_list].to_dataframe().to_csv(Path(csv_dir,'grid_organic.csv'))
sub_list = ["PCT_NAT_PFT"]
ds_sf[sub_list].to_dataframe().to_csv(Path(csv_dir,'grid_pfts.csv'))
return grids
# copy file from slow disk archive to scratch for processing
def copy_subset_resamp(raw_f, config):
# start subsetting netcdfs
xr_subset(raw_f,config)
# open, subset, save new netcdf files with/without compression
def xr_subset(raw_f, config):
src_file = raw_f[0]
sub_file = raw_f[1]
# remove subset netcdf output file if it already exists
rmv_file(sub_file)
# call xarray to subset/compress individual file
with xr.open_dataset(src_file, engine=config['nc_read']['engine']) as ds_tmp:
ds = ds_tmp.load()
# set netcdf write characteristics for xarray.to_netcdf()
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'],_FillValue=config['nc_write']['fillvalue'])
# encoding
encoding = {var: comp for var in ds[config['var_sub']].data_vars}
# write netcdf
ds[config['var_sub']].to_netcdf(sub_file, mode="w", encoding=encoding, \
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
# call xr_resample when subset complete
xr_resample(raw_f, config)
# open, resample, save new netcdf files with/without compression
def xr_resample(raw_f, config):
sub_file = raw_f[1]
out_file = raw_f[2]
# remove subset netcdf output file if it already exists
rmv_file(out_file)
# call xarray to subset/compress individual file
with xr.open_dataset(sub_file, engine=config['nc_read']['engine']) as ds_tmp:
ds = ds_tmp.load()
# set netcdf write characteristics for xarray.to_netcdf()
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'],_FillValue=config['nc_write']['fillvalue'])
# encoding
encoding = {var: comp for var in ds.data_vars}
# write netcdf
ds.resample(time='1D').mean().to_netcdf(out_file, mode="w", encoding=encoding, \
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
# remove temp file after copying
rmv_file(sub_file)
# open, concat, save new netcdf files with/without compression
def xr_concat(sub_f ,config):
file_list = sub_f[0]
cat_file = sub_f[1]
# remove subset netcdf output file if it already exists
rmv_file(cat_file)
# call xarray to subset/compress individual file
with xr.open_mfdataset(file_list, parallel=True, engine=config['nc_read']['engine']) as ds_tmp:
ds = ds_tmp.load()
# set netcdf write characteristics for xarray.to_netcdf()
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'],_FillValue=config['nc_write']['fillvalue'])
# encoding
encoding = {var: comp for var in ds.data_vars}
# write netcdf
ds.to_netcdf(cat_file, mode="w", encoding=encoding, \
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
###########################################################################################################################
# Adjust surface datasets
###########################################################################################################################
# functions to be applied using xarray.apply_ufunc in xr_adjust_surf()
# the apply_ufunc accesses each pft vector along lsmlat/lsmlon (each cell)
def pft_notree_adjust(da):
# copy 1D array of length 15 for each pft (ds is read-only when passed in as a numpy.ndarray)
da_tmp = da.copy()
# define vectors for subsetting natpft vector
tree = [1,2,3,4,5,6,7,8]
bsg = [0,11,12]
# calculate tree percentage sum, bareground/actic shrub/arctic grass (bsg) percentage sums
tree_sum = da[tree].sum()
bsg_sum = da[bsg].sum()
# divide tree sum by three, add to bsg, round to 1 decimal place
third = tree_sum // 3
da_tmp[0] = round(da[0] + third,1)
da_tmp[11] = round(da[11] + third,1)
da_tmp[12] = round(da[12] + third,1)
# set all tree values to zero
da_tmp[tree] = 0.0
# check for 100 percent, if not subtract difference from bare ground
if (da_tmp.sum() == 100):
return da_tmp
elif (da_tmp.sum() > 100.0):
da_tmp[0] = da_tmp[0] - abs(da_tmp.sum() - 100.0)
return da_tmp
elif (da_tmp.sum() < 100.0):
da_tmp[0] = da_tmp[0] + abs(da_tmp.sum() - 100.0)
return da_tmp
def pft_c3arcticgrass_adjust(da):
# copy 1D array of length 15 for CLM5 pfts (ds is read-only when passed in as a numpy.ndarray)
da_tmp = da.copy()
# list of PFT positions that are not C3 arctic grass to make zero
other_pfts = [0,1,2,3,4,5,6,7,8,9,10,11,13,14]
# set all other pfts to zero
da_tmp[other_pfts] = 0.0
# set to all arctic grass PFT
da_tmp[12] = 100.0
# return adjusted PFT array
return da_tmp
def natveg_adjust(da):
# change all netveg cell percentages to 100%
da = 100.0 # when no other dimension present ds is simply numpy.float64 reassigned without brackets
return da
def urban_adjust(da):
# copy 1D array of length three containing percentage for each urban level
da_tmp = da.copy()
# create list of position in array to assign zeros
urban_levels = [0,1,2]
# change all urban percentages to zero
da_tmp[urban_levels] = 0.0
return da_tmp
def zero_adjust(da):
# change all cell values to 0% for other PCT_
da = 0.0 # when no other dimension present ds is simple numpy.float64 reassigned without brackets
return da
# adjust surface data using xarray.apply_ufunc
def xr_adjust_surf(f, config):
# parse model name and file name from xr_surf.py dask client.submit call
model_name = f[0]
file_name = f[1]
# create file paths from json surfdata
in_file = Path(config['dir']['in'], file_name)
file_stem = Path(file_name).stem
out_file = Path(config['dir']['out'], file_stem + config['file_ending'] + ".nc")
# open netcdf surface dataset file
with xr.open_dataset(in_file, engine=config['nc_read']['engine']) as ds_tmp:
ds = ds_tmp.load()
# use xarray.apply_ufunc to acces/edit each cell value (or cell array if another dimension like pft is present)
try: # try statement to capture errors during application of xarray.apply_ufunc
for i in config['vars'][model_name]: # loop through adjusted vars from json config file
var = i[0] # first position from json file is variable name
adj_func = globals().get(i[1]) # second position from json file is the function name to grab from globals
core_dim = i[2] # third position from json file is a list of core dimensions
var_attr = ds[var].attrs # make copy of dataarray's attributes
# apply adj_func using xarray.apply_ufunc
ds[var] = xr.apply_ufunc( # replace netcdf variable with adjusted variable
adj_func, # function to apply across cells
ds[var], # dataset to vectorize
input_core_dims=[core_dim], # core dim of analysis, empty if no 3rd dimenion of data within cell
output_core_dims=[core_dim], # returned dim is same as input
vectorize=True, # loop over other dimensions of multi-dimensional array (here lat/lon)
)
if core_dim:
ds[var] = ds[var].transpose(core_dim[0], ...) # return order if core_dim was used and subsequently placed first
for i in var_attr:
ds[var].attrs[i] = var_attr[i] # restore lost attributes from apply_ufunc
except Exception: # write errors to output text file if they occur
with open(Path(config['dir']['out'], file_stem + config['file_ending'] + '_output.txt'), 'a') as f:
traceback.print_exc(file=f)
# set netcdf write characteristics for xarray.to_netcdf()
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'], _FillValue=None) #config['nc_write']['fillvalue'])
# set encoding for each variable
encoding = {var: comp for var in ds.data_vars}
# write netcdf
ds.to_netcdf(out_file, mode="w", encoding=encoding, \
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
###########################################################################################################################
# unzip .nc.gz and .zip files
###########################################################################################################################
# create list of files to unzip that is passed to dask cluster
def gunzip_file_list(config):
# define empty list to hold file paths
file_info = []
# loop through all directories listed in config_unzip
for f_dir in config['dir']['in']:
# create list of netcdfs in each directory with .gz ending
files = sorted(glob.glob("{}*.gz".format(f_dir)))
# find name of directory
base_name = Path(f_dir).name
# create directory for unzipped files
Path(config['dir']['out'],base_name).mkdir(parents=True, exist_ok=True)
#loop through individual files
for f in files:
# create tmp file name
tmp_file = Path(config['dir']['out'],'temporary_files', Path(f).name)
# create unziped file name
unzipped_file = Path(config['dir']['out'], base_name, Path(f).stem)
# update unzipped file name to remove odd naming from CEDA for 2010-2021 in all temp folders
unzipped_file = Path(str(unzipped_file).replace('crujra.v2.3.1.','crujra.v2.3.'))
# update new dswrf to correct file name
unzipped_file = Path(str(unzipped_file).replace('crujra.v2.4.','crujra.v2.3.'))
# append file into
file_info.append([f, tmp_file, unzipped_file])
# return file info list
return file_info
# create list of files to unzip that is passed to dask cluster
def gunzip_list(config):
# define empty list to hold file paths
file_info = []
# loop through all directories listed in config_unzip
for f_dir in config['dir']['in']:
# create list of netcdfs in each directory with .gz ending
files = sorted(glob.glob("{}*.gz".format(f_dir)))
#loop through individual files
for f in files:
# append file into
file_info.append([f, config['dir']['out']])
# return file info list
return file_info
# copy to tmp from project folder and unzip to scratch
def ungunzip_file(f, config):
# pull file, tmp file, and unzipped file paths from input
src_file = f[0]
unzip_dir = f[1]
unzip_file = Path(unzip_dir+str(Path(src_file).stem))
# unzip temp file to final scratch destination
with gzip.open(src_file, 'rb') as f_in:
with open(unzip_file, 'wb') as f_out:
shutil.copyfileobj(f_in, f_out)
# copy to tmp from project folder and unzip to scratch
def copy_gunzip(f, config):
# pull file, tmp file, and unzipped file paths from input
src_file = f[0]
tmp_file = f[1]
unzip_file = f[2]
# copy from slow project disk to scratch
shutil.copy(src_file, tmp_file)
# unzip temp file to final scratch destination
with gzip.open(tmp_file, 'rb') as f_in:
with open(unzip_file, 'wb') as f_out:
shutil.copyfileobj(f_in, f_out)
with open(Path(config['site_dir'] + 'debug_gzip.txt'), 'a') as f:
for line in file_info:
f.write(line + '\n')
def unzip_dir_list(config):
# define empty list to hold file paths
file_info = []
# loop through all directories listed in config_unzip
for f_dir in config['dir']['in']:
# create list of netcdfs in each directory with .gz ending
files = sorted(glob.glob("{}*.zip".format(f_dir)))
#loop through individual files
for f in files:
# check pull filename from globbed path to use as directory for unzipped files
if '1a' in f:
dir_name = 'Baseline_1901-2000'
elif '1b' in f:
dir_name = 'Baseline_2000-2021'
elif '1c' in f:
dir_name = 'OTC'
elif '1d' in f:
dir_name = 'Snow_fence'
else:
dir_name = Path(f).stem
# create unziped file location
unzipped_dir = Path(config['dir']['out']+dir_name)
# append to list of zip files to unzip
file_info.append([f, unzipped_dir])
# return file info list
return file_info
def unzip_dir(f, config):
# pull zipped file and destination file paths from input
src_zip = f[0]
unzip_dir = f[1]
# make directory for output files
Path(unzip_dir).mkdir(parents=True, exist_ok=True)
# process each individual file within zipped namelist
with zipfile.ZipFile(src_zip) as zipped_file:
for member in zipped_file.namelist():
filename = os.path.basename(member)
# skip directories
if not filename:
continue
# copy file (taken from zipfile's extract)
source = zipped_file.open(member)
target = open(os.path.join(unzip_dir, filename), "wb")
with source, target:
shutil.copyfileobj(source, target)
###########################################################################################################################
# Climate Biascorrection
###########################################################################################################################
# create full list of crujra reanalysis files (clm interpolated during transient simulation, all variables, by year)
def full_reanalysis_list(config_file):
# load site's config information
config = read_config(config_file)
# remove previous copy of crujra folder
rmv_dir(config['site_dir'])
# remake directory for subset files
Path(config['site_dir']+"sub/").mkdir(parents=True, exist_ok=True)
# debug print the config file info
with open(Path(config['site_dir'] + 'debug_filelist.txt'), 'a') as f:
for line in config:
f.write(line + ': ' + str(config[line]) + '\n')
# read all CRUJRA input file names from reanalysis directory
reanalysis_files = sorted(glob.glob("{}*.nc".format(config['reanalysis_dir'])))
# loop through files in reanalysis archive linking config, in, out files
file_info = []
for line in reanalysis_files:
in_file = line
sub_file = config['site_dir'] + "sub/" + str(Path(line).name)
file_info.append([config_file, in_file, sub_file])
return file_info
# subset original dswrf CRUJRAv2.3_updated files (to check on diurnal phase issue)
def raw_dswrf_list(config_file):
# load site's config information
config = read_config(config_file)
# remake directory for subset files
Path(config['site_dir']+"raw_dswrf_sub/").mkdir(parents=True, exist_ok=True)
# read all CRUJRA input file names from reanalysis directory
reanalysis_files = sorted(glob.glob("{}*.nc".format('/scratch/jw2636/wrpmip/CRUJRAv2.3_unzipped/data/dswrf_updated/')))
# loop through files in reanalysis archive
file_info = []
for line in reanalysis_files:
in_file = line
sub_file = config['site_dir'] + "raw_dswrf_sub/" + str(Path(line).name)
file_info.append([config_file, in_file, sub_file])
return file_info
# read in raw dswrf data, scale, apply mask, subsample to check on time issue
def subset_raw_dswrf(f):
# parse input strings
config = read_config(f[0])
f_in = Path(f[1])
f_sub = Path(f[2])
# open half degree CLM domain file for the landmask
with xr.open_dataset(Path(config['clm_landmask']), engine=config['nc_read']['engine'], decode_cf=True, use_cftime=True) as ds_tmp:
ds_mask = ds_tmp.load()
# rename dimensions
ds_mask = ds_mask.rename_dims({'nj': 'lat', 'ni': 'lon'})
# move and rotate lon data from 'xc' dataarray into dataset coordinate
ds_mask = ds_mask.assign_coords({'lon': (((ds_mask.xc.values[0] + 180) % 360) - 180)})
# move and subset lat data from 'yc' dataarray into dataset coordinate
ds_mask = ds_mask.assign_coords({'lat': extract_sublist(ds_mask.yc.values, 0)})
# sort the lon coornidate into acending order for xarray to work
ds_mask = ds_mask.sortby('lon')
# change land mask of 1/0s into boolean for subsetting
ds_mask['mask'] = ds_mask['mask'].astype(bool)
# open crujra_dswrf file to apply mask
with xr.open_dataset(f_in, engine=config['nc_read']['engine'], decode_cf=True, use_cftime=True) as ds_tmp:
ds = ds_tmp.load()
# convert dswrf from j/m2 to w/m2 by dividing by the seconds in 6hrs
ds['dswrf'] = ds['dswrf']/21600
# use where and drop=True to remove masked (ocean) cells
ds = ds.where(ds_mask['mask'], drop=True)
# subset nearest point based on lat/lon coordinate index
ds = ds.sel(lon=config['lon'], lat=config['lat'], method='nearest')
## set encoding for netcdfs
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'], _FillValue=None) #config['nc_write']['fillvalue'])
# set encoding for each variable
encoding = {var: comp for var in ds.data_vars}
# save netcdf to file
ds.to_netcdf(f_sub, mode="w", encoding=encoding, \
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
# glob subset raw dswrf files into timeseries to plot against obs and clm output crurja data
# this showed that both the clm output and raw dswrf files were offset by GMT hours from UTC 0
def concat_raw_dswrf(config_file):
# read config file
config = read_config(config_file)
# create name for full crujra site file
cru_file = Path(config['site_dir'] + "dswrf_" + config['site_name'] + "_dat.nc")
# glob all files in raw folder
f = sorted(glob.glob("{}*.nc".format(config['site_dir'] + 'raw_dswrf_sub/')))
# read in all files
with xr.open_mfdataset(f, parallel=True, engine=config['nc_read']['engine'], decode_cf=True, use_cftime=True) as ds_tmp:
ds = ds_tmp.load()
## set netcdf write characteristics for xarray.to_netcdf()
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'],_FillValue=None) #config['nc_write']['fillvalue'])
# encoding
encoding = {var: comp for var in ds.data_vars}
# write netcdf
ds.to_netcdf(cru_file, mode="w", encoding=encoding, \
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
# copy crujra file (clm hourly output), subset to only climate forcing and select site grid cell
# also shift crujra time index to GMT time at site so that 1901-01-01 00:00:00 actually means midnight
def subset_reanalysis(f):
# parse input strings
config = read_config(f[0])
f_in = Path(f[1])
f_sub = Path(f[2])
# open netcdf file
with xr.open_dataset(f_in, engine=config['nc_read']['engine'], decode_cf=True, use_cftime=True) as ds_tmp:
ds = ds_tmp.load()
# list of variables to keep
keep = ['FSDS','FLDS','PBOT','RAIN','SNOW','QBOT','TBOT','WIND']
# pull grid form config file
grid = config['lndgrid'] - 1
# subset to variables of interest and grid of interest for each site
ds = ds[keep].isel(lndgrid=grid)
# add snow and rain to make total precip
ds['PRECIP'] = ds['RAIN'] + ds['SNOW']
# shift the index by offset from GMT time at location
ds.coords['time'] = ds.indexes['time'].round('h') #.shift(config['cru_GMT_adj'], 'H')
## set encoding for netcdfs
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'], _FillValue=None) #config['nc_write']['fillvalue'])
# set encoding for each variable
encoding = {var: comp for var in ds[keep].data_vars}
# write netcdf
ds.to_netcdf(f_sub, mode="w", encoding=encoding, \
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
# glob files for site file of all years of subsampled/time shifted cru climate
# function is applied while makeing the full file list below that is reduced to only years that align with data
def concat_crujra(f, config):
# create name for full crujra site file
cru_file = Path(config['site_dir'] + "CRUJRA_" + config['site_name'] + "_allyears.nc")
# read in all files
with xr.open_mfdataset(f, parallel=True, engine=config['nc_read']['engine'], decode_cf=True, use_cftime=True) as ds_tmp:
ds = ds_tmp.load()
## set netcdf write characteristics for xarray.to_netcdf()
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'],_FillValue=None) #config['nc_write']['fillvalue'])
# encoding
encoding = {var: comp for var in ds.data_vars}
# write netcdf
ds.to_netcdf(cru_file, mode="w", encoding=encoding, \
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
# create list of cru climate files/years that align with observational data to compare
def cru_sitesubset_list(config_file):
# load site's config information
config = read_config(config_file)
# list of files in sub folder
file_list = sorted(glob.glob("{}*.nc".format(config['site_dir']+"sub/")))
# call function to concat all crujra files subset to site's gridcell
concat_crujra(file_list, config)
# create list of years present from file names
year_list = []
for item in file_list:
year_list.append(item.split('.')[-2].split('-')[0])
# subset year list to years where data exists
year_bool = [((int(i) >= config['year_start'])&(int(i) <= config['year_end'])) for i in year_list]
year_list = list(itertools.compress(year_list, year_bool))
# loop through files in reanalysis folder for data years
file_info = []
for year in year_list:
in_file = glob.glob("{}*{}*.nc".format(config['site_dir']+"sub/", year))[0]
file_info.append(in_file)
return [config_file, file_info]
# concatenate cru climate years that align with obs data using list created above
def concat_cru_sitesubset(f):
# read in config file for site
config = read_config(f[0])
cat_file = config['site_dir'] + "CRUJRA_" + config['site_name'] + "_dat.nc"
# open netcdf file
with xr.open_mfdataset(f[1], parallel=True, engine=config['nc_read']['engine'], decode_cf=True, use_cftime=True) as ds_tmp:
ds = ds_tmp.load()
# Final QQ for clm created hourly cru data
ds['FSDS'] = ds['FSDS'].where(ds['FSDS'] > 0, 0.0)
ds['FLDS'] = ds['FLDS'].where(ds['FLDS'] > 0, 0.0)
ds['RAIN'] = ds['RAIN'].where(ds['RAIN'] > 0, 0.0)
ds['SNOW'] = ds['SNOW'].where(ds['SNOW'] > 0, 0.0)
ds['QBOT'] = ds['QBOT'].where(ds['QBOT'] > 0, 0.0)
ds['WIND'] = ds['WIND'].where(ds['WIND'] > 0, 0.0)
## set netcdf write characteristics for xarray.to_netcdf()
comp = dict(zlib=config['nc_write']['zlib'], shuffle=config['nc_write']['shuffle'],\
complevel=config['nc_write']['complevel'],_FillValue=None) #config['nc_write']['fillvalue'])
# encoding
encoding = {var: comp for var in ds.data_vars}
# write netcdf
ds.to_netcdf(cat_file, mode="w", encoding=encoding,\
format=config['nc_write']['format'],\
engine=config['nc_write']['engine'])
# read in observational data, adjust all aspects of obs data as needed per site, output to netcdf
def combine_site_observations(config_file):
# read in config file for site
config = read_config(config_file)
try:
# read in columns of interest from first file
obs_data = pd.read_csv(config['obs']['f1']['name'], sep=config['obs']['f1']['sep'], \
index_col=False, engine='python', skiprows=config['obs']['f1']['skip_rows'], \
usecols=config['obs']['f1']['cols_old'])
except:
# if no data files listed end function
print_string = 'No observation data; combine_site_observations() skipped for ' + config['site_name']
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(print_string)
return
# enforce column order from subset procedure from usecols in read_csv
obs_data = obs_data[config['obs']['f1']['cols_old']]
# rename data columns to CLM standard
obs_data = obs_data.rename(columns=config['obs']['f1']['cols_new'])
# remove rows with no date/time of measurement
obs_data = obs_data.dropna(subset=config['obs']['f1']['datetime_cols'])
# print statement code to use when testing/adding new sites
with option_context('display.max_rows', 10, 'display.max_columns', 10):
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data.head(), file=f)
print(obs_data.dtypes, file=f)
# handle site specific idiosyncrasies
site = config['site_name']
match site:
case 'USA-EightMileLake':
# convert numerical timestamp to string for datetime.strptime
obs_data['time'] = obs_data['TIMESTAMP_START'].astype(str)
# remove -9999 values, as numbers are float values have to use np.isclose
#num_cols = obs_data.select_dtypes(np.number).columns
#obs_data[num_cols] = obs_data[num_cols].mask(np.isclose(obs_data[num_cols].values, -9999))
# change -9999 fill values to NAs
obs_data = obs_data.replace(-9999, np.NaN)
# convert TBOT from celsius to kelvin
obs_data.loc[:,'TBOT'] = obs_data['TBOT'] + 273.15
# Convert from kPa to Pa
obs_data.loc[:,'PBOT'] = obs_data['PBOT'] * 1000
# convert RH to SH
obs_data.loc[obs_data['RH'] < 0, 'RH'] = 0
obs_data.loc[obs_data['RH'] > 100, 'RH'] = 100
obs_data.loc[:,'QBOT'] = specific_humidity(obs_data['RH'], obs_data['TBOT'], obs_data['PBOT'])
obs_data = obs_data.drop(columns=['RH'])
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data, file=f)
case 'USA-Toolik':
# fix hourly timestep - cannot have 24 as hour value, only 0:23 for datetime.strptime
obs_data.loc[:,'hour'] = obs_data['hour'] - 100
obs_data['hour'] = obs_data['hour'].astype(int)
# combine date and hour columns for timestamp - need to pad hours with preceeding zeros
obs_data['time'] = obs_data['date'].astype(str) + " " + obs_data['hour'].astype(str).str.zfill(4)
with option_context('display.max_rows', 10, 'display.max_columns', 10):
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data.head(), file=f)
print(obs_data.dtypes, file=f)
# convert TBOT from celsius to kelvin
obs_data.loc[:,'TBOT'] = obs_data['TBOT'] + 273.15
# change mbar -> Pa
obs_data.loc[:,'PBOT'] = obs_data['PBOT'] * 100
obs_data.loc[obs_data['PBOT'] < 90000, 'PBOT'] = np.NaN
# convert RH to SH
obs_data.loc[obs_data['RH'] < 0, 'RH'] = 0
obs_data.loc[obs_data['RH'] > 100, 'RH'] = 100
obs_data['QBOT'] = specific_humidity(obs_data['RH'], obs_data['TBOT'], obs_data['PBOT'])
obs_data = obs_data.drop(columns=['RH'])
# precip
obs_data['PRECIP'] = obs_data['PRECIP']/3600
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data, file=f)
case 'SWE-Abisko':
# abiskos data is very messy and has all kinds of non-numeric character strings which confuses python
# python then turns all columns into objects (strings) which breaks all the math code
# to fix this I have to force all columns to numeric which makes all non-numbers into NaNs
cols_to_num = ['TBOT','FSDS','FLDS','PBOT','RH','WIND']
for col in cols_to_num:
obs_data[col] = pd.to_numeric(obs_data[col], errors='coerce')
# convert numerical timestamp to string for datetime.strptime
obs_data['time'] = obs_data['Timestamp (UTC)'].astype(str)
# remove -9999 values, as numbers are float values have to use np.isclose
#num_cols = obs_data.select_dtypes(np.number).columns
#obs_data[num_cols] = obs_data[num_cols].mask(np.isclose(obs_data[num_cols].values, -6999))
# change -6999 fill values to NAs
#obs_data = obs_data.replace(-6999, np.NaN)
# convert TBOT from celsius to kelvin
obs_data.loc[:,'TBOT'] = obs_data['TBOT'] + 273.15
# clean FLDS values
obs_data.loc[obs_data['FLDS'] < 50, 'FLDS'] = np.NaN
# Convert from mbar to Pa
obs_data.loc[:,'PBOT'] = obs_data['PBOT'] * 100
obs_data.loc[obs_data['PBOT'] < 95000, 'PBOT'] = np.NaN
# convert RH to SH
obs_data.loc[obs_data['RH'] < 0, 'RH'] = 0
obs_data.loc[obs_data['RH'] > 100, 'RH'] = 100
obs_data.loc[:,'QBOT'] = specific_humidity(obs_data['RH'], obs_data['TBOT'], obs_data['PBOT'])
obs_data = obs_data.drop(columns=['RH'])
# precip
obs_data['PRECIP'] = obs_data['PRECIP']/3600
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data, file=f)
print(obs_data.dtypes, file=f)
case 'RUS-Seida':
# make time column from datetime string
obs_data['time'] = obs_data['datetimeGMT3'].astype(str) # + " " + obs_data['hourGMT3'].astype(str).str.zfill(2)
# drop old time columns
obs_data = obs_data.drop(columns=['datetimeGMT3'])
# set to datetime pandas
obs_data.loc[:,'time'] = pd.to_datetime(obs_data['time'], format="%Y-%m-%d %H:%M")
# set index to timestamp
obs_data = obs_data.set_index('time')
# resample the sub-hourly data to hourly averages
obs_data = obs_data.resample('1h').mean()
# change index to str with correct format
obs_data.index = obs_data.index.strftime('%Y-%m-%d %H:%M:%S')
# convert numerical timestamp to string for datetime.strptime to integrate back into set coding below
obs_data = obs_data.reset_index()
obs_data['time'] = obs_data['time'].astype(str)
# convert TBOT from celsius to kelvin
obs_data.loc[:,'TBOT'] = obs_data['TBOT'] + 273.15
# only given par - will scale up based on a par = 0.46 shortwave (shortwave = par/0.46)
obs_data.loc[:,'FSDS'] = (obs_data['PAR']/4.57)/0.46
obs_data = obs_data.drop(columns = ['PAR'])
# convert rh to sh
#obs_data.loc[:,'QBOT'] = specific_humidity(obs_data['RH'], obs_data['TBOT'], obs_data['PBOT'])
obs_data = obs_data.drop(columns=['RH','WIND'])
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data, file=f)
# pressure, windspeed seem to be in same units as clm reprocessed crujra
case 'CAN-DaringLake':
# read in second dataset
obs_data2 = pd.read_csv(config['obs']['f2']['name'], sep=config['obs']['f2']['sep'], \
index_col=False, engine='python', skiprows=config['obs']['f2']['skip_rows'], \
usecols=config['obs']['f2']['cols_old'])
# enforce column order from subset procedure from usecols in read_csv
obs_data2 = obs_data2[config['obs']['f2']['cols_old']]
# rename data columns to CLM standard
obs_data2 = obs_data2.rename(columns=config['obs']['f2']['cols_new'])
# remove rows with no date/time of measurement
obs_data2 = obs_data2.dropna(subset=config['obs']['f2']['datetime_cols'])
# concat pandas dataframes
obs_data = pd.concat([obs_data, obs_data2], ignore_index=True)
# convert numerical timestamp to string for datetime.strptime
obs_data.loc[:,'Hour'] = obs_data['Hour'].astype(int) - 100
obs_data.loc[:,'Date'] = obs_data['Year'].astype(str) + "-" + obs_data['Month'].astype(str) + "-" + obs_data['Day'].astype(str)
obs_data.loc[:,'time'] = obs_data['Date'].astype(str) + " " + obs_data['Hour'].astype(str).str.zfill(4)
obs_data = obs_data.drop(columns = ['Date'])
# convert TBOT from celsius to kelvin
obs_data.loc[:,'TBOT'] = obs_data['TBOT'] + 273.15
# No pressure given so I'll calculate air pressure given a rough reference (400m ~ 96357Pa) and plug in sites TBOT
# into barometric pressure equation to estimate the pressure a few meters up at site elevation (424m)
# basically adding temperature variability to reference pressure through barometric pressure function
# I'm only doing this because I need pressure to convert RH to SH
#g0 = 9.80665 # gavitational constat in m/s2
#M0 = 0.0289644 # molar mass of air kg/mol
#R0 = 8.3144598 # universal gas constant - J/(mol K)
#hb = 0 # reference level, here just below sites elevation
#Pb = 101325 # estimated reference pressure at 400 meters and 0 degre C
#obs_data.loc[:,'PBOT'] = Pb*np.exp((-g0*M0*(424-hb))/(R0*obs_data['TBOT']))
#obs_data.loc[obs_data['PBOT'] < 90000, 'PBOT'] = np.NaN
# convert RH to SH
#obs_data.loc[:,'QBOT'] = specific_humidity(obs_data['RH'], obs_data['TBOT'], obs_data['PBOT'])
# precip
#obs_data['PRECIP'] = obs_data['PRECIP']/3600
# remove uneeded columns
obs_data = obs_data.drop(columns=['RH','PRECIP'])
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data, file=f)
case 'USA-Utqiagvik':
# subset to BD for Barrow in strSitCom
obs_data = obs_data.loc[obs_data['SITE'] == 'BD']
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data, file=f)
# read other files and concat to first
for extra_file in config['obs']['f1']['extended_files']:
obs_data2 = pd.read_csv(extra_file, index_col=False, engine='python', skiprows=config['obs']['f1']['skip_rows'], \
usecols=config['obs']['f1']['cols_old'], sep=config['obs']['f1']['sep'])
# enforce column order from subset procedure from usecols in read_csv
obs_data2 = obs_data2[config['obs']['f1']['cols_old']]
# rename data columns to CLM standard
obs_data2 = obs_data2.rename(columns=config['obs']['f1']['cols_new'])
# remove rows with no date/time of measurement
obs_data2 = obs_data2.dropna(subset=config['obs']['f1']['datetime_cols'])
# select Barrow site
obs_data2 = obs_data2.loc[obs_data2['SITE'] == 'BD']
# concat file
obs_data = pd.concat([obs_data, obs_data2], ignore_index=True)
# drop site column
obs_data = obs_data.drop(columns = ['SITE'])
# set time from timestamp
obs_data['time'] = obs_data['strAlaska'].astype(str)
# remove -9999 values, as numbers are float values have to use np.isclose
num_cols = obs_data.select_dtypes(np.number).columns
obs_data[num_cols] = obs_data[num_cols].mask(np.isclose(obs_data[num_cols].values, -999.9))
# remove zero wind speeds, select 2014 forward
obs_data.loc[obs_data['WIND'] <= 0, 'WIND'] = np.nan
obs_data.loc[obs_data.time < '2014-01-01', 'WIND'] = np.nan
# scale celsius to kelvin
obs_data.loc[:,'TBOT'] = obs_data['TBOT'] + 273.15
# precip
obs_data['PRECIP'] = obs_data['PRECIP']/3600
# calculate SWIN from PAR
#obs_data.loc[:,'FSDS'] = obs_data['PAR']/2.1#4.57)/0.46
obs_data = obs_data.drop(columns = ['PAR','PRECIP'])
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data, file=f)
case 'USA-Atqasuk':
# subset to BD for Barrow in strSitCom
obs_data = obs_data.loc[obs_data['SITE'] == 'AD']
with open(Path(config['site_dir'] + 'debug_obs.txt'), 'a') as f:
print(obs_data, file=f)
# read other files and concat to first
for extra_file in config['obs']['f1']['extended_files']:
obs_data2 = pd.read_csv(extra_file, index_col=False, engine='python', skiprows=config['obs']['f1']['skip_rows'], \
usecols=config['obs']['f1']['cols_old'], sep=config['obs']['f1']['sep'])