-
Notifications
You must be signed in to change notification settings - Fork 3
/
noahmp_ldasout2cf.py
executable file
·226 lines (212 loc) · 8.96 KB
/
noahmp_ldasout2cf.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
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# convert NoahMP outputs to CF-compatible files
import sys
import os
import glob
import datetime
import argparse
import dateutil.parser
import numpy as np
import netCDF4 as nc
np.seterr(invalid='ignore')
TDIM = 'time'
TVAR = 'TIMES'
timeunits = 'hours since 1900-01-01'
XDIM = 'west_east'
XVAR = 'WEST_EAST'
YDIM = 'south_north'
YVAR = 'SOUTH_NORTH'
ACCVARS = ['ACSNOW', 'ACSNOM', 'SFCRNOFF', 'UGDRNOFF']
VALIDMIN = -1e10
def datetime4name(filename):
timestr = os.path.basename(filename).split('.')[0]
return datetime.datetime.strptime(timestr, '%Y%m%d%H')
def source_info(datadir, begtime, endtime):
allfiles = sorted(glob.glob(os.path.join(datadir, '*.LDASOUT_DOMAIN1')))
files = [x for x in allfiles
if datetime4name(x) >= begtime and datetime4name(x) < endtime]
timestep = 0
integrity = False
if len(files) > 1:
timesteps = set()
for ifile in range(1,len(files)):
timesteps.add((datetime4name(files[ifile]) - datetime4name(files[ifile-1])).total_seconds())
timestep = timesteps.pop()
if len(timesteps) == 0 \
and ((datetime4name(files[0]) - begtime).total_seconds() < timestep) \
and ((endtime - datetime4name(files[-1])).total_seconds() <= timestep):
integrity = True
return files, timestep, integrity
def define_output(wrfinput, fi, fo):
# inquire spatial dimension from wrfinput
with nc.Dataset(wrfinput, 'r') as fwrf:
if fwrf.MAP_PROJ == 6:
lat2d = np.squeeze(fwrf.variables['XLAT'][:])
lon2d = np.squeeze(fwrf.variables['XLONG'][:])
lat = lat2d[:,0]
lon = lon2d[0,:]
else:
lat = None
lon = None
# create Dimension
dimset = set()
for var in fi.variables:
if var.upper() in [TVAR, XVAR, YVAR]:
continue
dimset.update((x.lower() for x in fi.variables[var].dimensions))
for dim in dimset:
if dim == TDIM:
fo.createDimension(dim, None)
else:
fo.createDimension(dim, len(fi.dimensions[dim]))
# create vars
fo.createVariable(XDIM, 'f', (XDIM,), zlib=True, complevel=6)
fo.variables[XDIM].standard_name = 'longitude'.encode('ascii')
fo.variables[XDIM].units = 'degree_east'.encode('ascii')
fo.variables[XDIM].axis = 'X'.encode('ascii')
fo.variables[XDIM][:] = lon
fo.createVariable(YDIM, 'f', (YDIM,), zlib=True, complevel=6)
fo.variables[YDIM].standard_name = 'latitude'.encode('ascii')
fo.variables[YDIM].units = 'degree_north'.encode('ascii')
fo.variables[YDIM].axis = 'Y'.encode('ascii')
fo.variables[YDIM][:] = lat
for var in fi.variables:
if var.upper() == TVAR:
fo.createVariable(TDIM, 'f8', (TDIM,), zlib=True, complevel=6)
fo.variables[TDIM].standard_name = 'time'.encode('ascii')
fo.variables[TDIM].units = timeunits.encode('ascii')
fo.variables[TDIM].calendar = 'standard'.encode('ascii')
fo.variables[TDIM].axis = 'T'.encode('ascii')
elif var.upper() in set([XVAR, YVAR]):
pass
elif var.upper() not in ACCVARS:
dtype = fi.variables[var].dtype
dims = [x.lower() for x in fi.variables[var].dimensions]
dims_n = []
if TDIM in dims:
dims.remove(TDIM)
dims_n.append(TDIM)
if YDIM in dims:
dims.remove(YDIM)
dims_n.append(YDIM)
if XDIM in dims:
dims.remove(XDIM)
dims_n.append(XDIM)
if len(dims) > 1:
print("don't support multiple z-axis: " + str(dims) + ' of ' + var)
sys.exit(1)
elif len(dims) == 1:
zdim = dims[0]
dims_n.insert(1, zdim)
if np.issubdtype(dtype, np.float):
fo.createVariable(var, dtype, dims_n,
zlib=True, complevel=6, fill_value=np.nan)
else:
fo.createVariable(var, dtype, dims_n, zlib=True, complevel=6)
for att in fi.variables[var].ncattrs():
attval = fi.variables[var].getncattr(att)
if isinstance(fi.variables[var].getncattr(att), str):
attval = attval.encode('ascii')
fo.variables[var].setncattr(att, attval)
else: # ACCVARS
dtype = fi.variables[var].dtype
dims = [x.lower() for x in fi.variables[var].dimensions]
if np.issubdtype(dtype, np.float):
fo.createVariable(var, dtype, dims_n,
zlib=True, complevel=6, fill_value=np.nan)
else:
fo.createVariable(var, dtype, dims_n, zlib=True, complevel=6)
for att in fi.variables[var].ncattrs():
attval = fi.variables[var].getncattr(att)
if att.lower() == 'units':
attval = (attval + ' s-1').encode('ascii')
elif att.lower() == 'description':
attval = attval.lower().replace('accumulated ', '').replace('accumulatetd ','').encode('ascii')
elif isinstance(fi.variables[var].getncattr(att), str):
attval = attval.encode('ascii')
fo.variables[var].setncattr(att, attval)
for att in fi.ncattrs():
if isinstance(fi.getncattr(att), str):
fo.setncattr(att, fi.getncattr(att).encode('ascii'))
else:
fo.setncattr(att, fi.getncattr(att))
return
def copy_var(fi, fo, var, ind):
if var.lower() in [TDIM, XDIM, YDIM] \
or var.upper() in [TVAR, XVAR, YVAR] \
or var.upper() in ACCVARS:
return
# mask invalid value
fi.variables[var].set_auto_maskandscale(False)
fo.variables[var].set_auto_maskandscale(False)
v = fi.variables[var][:]
if np.issubdtype(fi.variables[var].dtype, np.float):
v[v <= VALIDMIN] = np.nan
# swap dimension
dims = [x.lower() for x in fi.variables[var].dimensions]
zdim = set(dims) - set([TDIM, XDIM, YDIM])
if len(zdim) == 1:
zdim = zdim.pop()
v = np.swapaxes(v, dims.index(zdim), 1)
fo.variables[var][ind,...] = v
return
def acc2flx(fic, fip, fo, var, ind, ts):
if var.upper() not in ACCVARS:
return
fic.variables[var].set_auto_maskandscale(False)
fip.variables[var].set_auto_maskandscale(False)
fo.variables[var].set_auto_maskandscale(False)
vc = fic.variables[var][:]
vp = fip.variables[var][:]
vc[vc <= VALIDMIN] = np.nan
vp[vp <= VALIDMIN] = np.nan
fo.variables[var][ind,...] = (vc - vp) / ts
return
def main(wrfinput, datadir, outfile, begtime, endtime, partially=False):
files, timestep, integrity = source_info(datadir, begtime, endtime)
if (not integrity) and (not partially):
print('not enough files (try --partially)')
sys.exit(1)
with nc.Dataset(outfile, 'w') as fo:
for ifile in range(len(files)):
print(files[ifile])
with nc.Dataset(files[ifile], 'r') as fi:
if ifile == 0:
define_output(wrfinput, fi, fo)
fo.variables[TDIM][ifile] = nc.date2num(datetime4name(files[ifile]),
fo.variables[TDIM].units)
for var in fi.variables:
if var.upper() in ACCVARS:
continue
copy_var(fi, fo, var, ifile)
print('acc to flux')
for ifile in reversed(range(1,len(files))):
with nc.Dataset(files[ifile], 'r') as fic, \
nc.Dataset(files[ifile-1], 'r') as fip:
for var in ACCVARS:
acc2flx(fic, fip, fo, var, ifile, timestep)
startfile = os.path.join(datadir,
(datetime4name(files[0]) - datetime.timedelta(seconds=timestep)).strftime('%Y%m%d%H') + '.LDASOUT_DOMAIN1')
if os.path.exists(startfile):
with nc.Dataset(files[0], 'r') as fic, \
nc.Dataset(startfile, 'r') as fip:
for var in ACCVARS:
acc2flx(fic, fip, fo, var, 0, timestep)
else:
for var in ACCVARS:
fo.variables[var][0,...] = fo.variables[var][1,...]
return
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='convert NoahMP outputs to single CF-compatible file')
parser.add_argument('wrfinput')
parser.add_argument('datadir', help='root directory of raw NoahMP outputs')
parser.add_argument('outfile', help='CF-compaible output file')
parser.add_argument('begtime', help='inclusive')
parser.add_argument('endtime', help='exclusive')
parser.add_argument('--partially', action='store_true')
args = parser.parse_args()
main(args.wrfinput, args.datadir, args.outfile,
dateutil.parser.parse(args.begtime),
dateutil.parser.parse(args.endtime),
args.partially)