-
Notifications
You must be signed in to change notification settings - Fork 4
/
grace_tr_correction.py
366 lines (258 loc) · 12.5 KB
/
grace_tr_correction.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
# -*- coding: utf-8 -*-
"""
Created on Tue Dec 19 09:47:04 2017
@author: bec
"""
from __future__ import print_function
from builtins import zip
from builtins import range
import os
import numpy as np
import matplotlib.pyplot as plt
import re
import pandas as pd
import scipy.optimize as optimization
from datetime import datetime
from scipy import interpolate
from calendar import monthrange as mr
from dateutil.relativedelta import relativedelta
from WA_Hyperloop import becgis
import WA_Hyperloop.get_dictionaries as gd
def get_ts_from_complete_data(complete_data, mask, keys, dates = None):
if keys == None:
keys = list(complete_data.keys())
common_dates = becgis.common_dates([complete_data[key][1] for key in keys])
becgis.assert_proj_res_ndv([complete_data[key][0] for key in keys])
MASK = becgis.open_as_array(mask, nan_values = True)
tss = dict()
for key in keys:
var_mm = np.array([])
for date in common_dates:
tif = complete_data[key][0][complete_data[key][1] == date][0]
DATA = becgis.open_as_array(tif, nan_values = True)
DATA[np.isnan(DATA)] = 0.0
DATA[np.isnan(MASK)] = np.nan
var_mm = np.append(var_mm, np.nanmean(DATA))
tss[key] = (common_dates, var_mm)
return tss
def get_ts_from_complete_data_spec(complete_data, lu_fh, keys, a, dates = None):
if keys == None:
keys = list(complete_data.keys())
common_dates = becgis.common_dates([complete_data[key][1] for key in keys])
becgis.assert_proj_res_ndv([complete_data[key][0] for key in keys])
MASK = becgis.open_as_array(lu_fh, nan_values = True)
lucs = lucs = gd.get_sheet4_6_classes()
gw_classes = list()
for subclass in ['Forests','Rainfed Crops','Shrubland','Forest Plantations']:
gw_classes += lucs[subclass]
mask_gw = np.logical_or.reduce([MASK == value for value in gw_classes])
tss = dict()
for key in keys:
var_mm = np.array([])
for date in common_dates:
tif = complete_data[key][0][complete_data[key][1] == date][0]
DATA = becgis.open_as_array(tif, nan_values = True)
DATA[np.isnan(DATA)] = 0.0
DATA[np.isnan(MASK)] = np.nan
alpha = np.ones(np.shape(DATA)) * a
alpha[mask_gw] = 0.0
var_mm = np.append(var_mm, np.nanmean(DATA * alpha))
tss[key] = (common_dates, var_mm)
return tss
def endofmonth(dates):
dts = np.array([datetime(dt.year, dt.month,
mr(dt.year, dt.month)[1]) for dt in dates])
return dts
def calc_var_correction(metadata, complete_data, output_dir,
formula = 'p-et-tr+supply_swa', plot = True,
slope = False, bounds = ([0.0, 0.0, 1.], [1., 1., 12.])):
bounds = np.array(bounds)
if plot:
output_dir = os.path.join(output_dir)
if not os.path.exists(output_dir):
os.makedirs(output_dir)
keys, operts = split_form(formula)
tss = get_ts_from_complete_data(complete_data, metadata['lu'], keys)
# Load the Grace timeseries.
grace = read_grace_csv(metadata['GRACE'])
# The monthly datasets are cumulatives, i.e. precipitation for "2003-03-01" shows the precitpiation between 03-01 and 03-31.
# Grace dates are interpolated and corrected here to match with those cumulatives. The new grace array will have the same length
# as the other timeseries (in tss), i.e. the array is padded with NaNs.
grace = interp_ts(grace, (endofmonth(tss[keys[0]][0]), -9999))
new_dates = np.array([datetime(dt.year, dt.month, 1) for dt in grace[0]])
grace = (new_dates, grace[1])
# Fit to storage values (slope = False) or to the trend in storage values (slope = True).
if slope:
grace = (grace[0], grace[1] - np.nanmean(grace[1]))
# Function to determine the balance
def func(x, alpha, beta, theta, slope = slope):
ds = np.zeros(x.size)
for key, oper in zip(keys, operts):
scalar_array = alpha * (np.cos((x - theta) * (np.pi / 6)) * 0.5 + 0.5) + (beta * (1 - alpha))
if key == keys[-1]:
new_data = tss[key][1][x] * scalar_array
else:
new_data = tss[key][1][x]
if oper == '+':
ds += new_data
elif oper == '-':
ds -= new_data
elif oper == '/':
ds /= new_data
elif oper == '*':
ds *= new_data
else:
raise ValueError('Unknown operator in formula')
if slope:
return np.cumsum(ds) - np.mean(np.cumsum(ds))
else:
return np.cumsum(ds)
# Check which Grace datapoints are missing and make a list of indice-numbers for which
# data is available.
msk = ~np.isnan(grace[1])
x = np.where(msk == True)[0].astype(int).tolist()
print("Starting alpha optimization")
# Initial values for alpha, beta and theta
p0 = (bounds[0] + bounds[1])/2.
# Find values for alpha, beta and theta so that the waterbalance matches as closely
# as possible with Grace.
a, b = optimization.curve_fit(func, x, grace[1][msk], p0 = p0 , bounds= bounds)
if plot:
var_name = keys[-1]
plt.figure(1, figsize = (10,6))
plt.clf()
plt.grid(b=True, which='Major', color='0.65',linestyle='--', zorder = 100)
ax = plt.gca()
plt.plot(*tss[var_name], label = "Total Supply", color = 'k')
plt.xlabel("Time [date]")
plt.ylabel("Total Supply [mm]")
ax.fill_between(*tss[var_name], color = '#6bb8cc', label = 'SW Supply')
ax.fill_between(*calc_gwsupply(tss[var_name], a), color = '#c48211', label = 'GW Supply')
ax.set_facecolor('lightgray')
plt.scatter(*tss[var_name], color = 'k')
plt.ylim([0, np.nanmax(tss[var_name][1]) * 1.2])
plt.xlim([tss[var_name][0][0],tss[var_name][0][-1]])
plt.title('Surface and Groundwater Supplies')
plt.legend()
plt.savefig(os.path.join(output_dir, metadata['name'], 'SWGW_Supply'))
plt.figure(2, figsize = (10,6))
plt.clf()
plt.grid(b=True, which='Major', color='0.65',linestyle='--', zorder = 0)
ax = plt.gca()
ax.set_facecolor('lightgray')
X = np.arange(tss[var_name][0].size)
scalar_array = a[0] * (np.cos((X - a[2]) * (np.pi / 6)) * 0.5 + 0.5) + (a[1] * (1 - a[0]))
plt.plot(tss[var_name][0], scalar_array, color = 'darkblue')
plt.scatter(tss[var_name][0], scalar_array, color = 'darkblue')
plt.ylim([0, 1])
plt.xlim([tss[var_name][0][0],tss[var_name][0][-1]])
plt.xlabel("Time [date]")
plt.ylabel(r"$Supply_{SW}\ /\ Supply_{tot}\ [-]$")
plt.suptitle(r"$Supply_{SW} = (\alpha \cdot [\cos(t \cdot \frac{2\pi}{12} - \theta) \cdot \frac{1}{2} + \frac{1}{2}] + \beta \cdot (1 - \alpha)) \cdot Supply_{tot}$" + '\n' + r"$\alpha = $" +
'{0:.2f}'.format(a[0]) + r', $\beta = $' +
'{0:.2f}'.format(a[1]) + r', $\theta = $' +
'{0:.2f}'.format(a[2]))
plt.savefig(os.path.join(output_dir, metadata['name'], 'SWGW_Ratio'))
plt.figure(3, figsize = (10,6))
plt.clf()
plt.grid(b=True, which='Major', color='0.65',linestyle='--', zorder = 0)
ax = plt.gca()
ax.set_facecolor('lightgray')
plt.plot(*grace, label = 'GRACE', color = 'r')
plt.plot(*calc_polyfit(grace, order = 1), color = 'r', linestyle = ':')
if metadata['lu_based_supply_split']:
plt.plot(grace[0], func(X, 0., 1., 1., slope = False), color = 'darkblue', label = 'WAplus (lu)')
plt.plot(*calc_polyfit((grace[0], func(X, 0., 1., 1., slope = False)), order = 1), color = 'darkblue', linestyle = ':')
plt.plot(grace[0], func(X, a[0], a[1], a[2], slope = False), color = 'k', label = 'WAplus (lu + grace)')
plt.plot(*calc_polyfit((grace[0], func(X, a[0], a[1], a[2], slope = False)), order = 1), color = 'k', linestyle = ':')
else:
plt.plot(grace[0], func(X, a[0], a[1], a[2], slope = False), color = 'k', label = 'WAplus (grace)')
plt.plot(*calc_polyfit((grace[0], func(X, a[0], a[1], a[2], slope = False)), order = 1), color = 'k', linestyle = ':')
plt.legend()
plt.xlim([grace[0][0], grace[0][-1]])
plt.ylabel('dS/dt [mm/month]')
plt.xlabel("Time [date]")
plt.savefig(os.path.join(output_dir, metadata['name'], 'dSdt_Grace'))
x0 = tss[keys[0]][0][0]
return a, x0
def calc_gwsupply(total_supply, params):
x = list(range(len(total_supply[0])))
scalar_array = params[0] * (np.cos((x - params[2]) * (np.pi / 6)) * 0.5 + 0.5) + (params[1] * (1 - params[0]))
gw_supply = (total_supply[0], total_supply[1] - (total_supply[1] * scalar_array))
return gw_supply
def correct_var(metadata, complete_data, output_dir, formula,
new_var, slope = False, bounds = (0, [1.0, 1., 12.])):
var = split_form(formula)[0][-1]
a, x0 = calc_var_correction(metadata, complete_data, output_dir,
formula = formula, slope = slope, plot = True, bounds = bounds)
for date, fn in zip(complete_data[var][1], complete_data[var][0]):
geo_info = becgis.get_geoinfo(fn)
data = becgis.open_as_array(fn, nan_values = True)
x = calc_delta_months(x0, date)
fraction = a[0] * (np.cos((x - a[2]) * (np.pi / 6)) * 0.5 + 0.5) + (a[1] * (1 - a[0]))
data *= fraction
folder = os.path.join(output_dir, metadata['name'],
'data', new_var)
if not os.path.exists(folder):
os.makedirs(folder)
bla = os.path.split(fn)[1].split('_')[-1]
filen = 'supply_sw_' + bla[0:6] + '.tif'
fn = os.path.join(folder, filen)
becgis.create_geotiff(fn, data, *geo_info)
meta = becgis.sort_files(folder, [-10,-6], month_position = [-6,-4])[0:2]
return a, meta
def calc_delta_months(x0, date):
if isinstance(x0, datetime):
x0 = x0.date()
if isinstance(date, datetime):
date = date.date()
x = 0
checker = x0
direction = {False: -1, True: 1}[date > x0]
while checker != date:
checker += relativedelta(months = 1) * direction
x += 1 * direction
return x
def toord(dates):
return np.array([dt.toordinal() for dt in dates[0]])
def interp_ts(source_ts, destin_ts):
f = interpolate.interp1d(toord(source_ts), source_ts[1],
bounds_error = False, fill_value = np.nan)
values = f(toord(destin_ts))
return (destin_ts[0], values)
def read_grace_csv(csv_file):
df = pd.read_csv(csv_file)
dt = [datetime.strptime(dt, '%Y-%m-%d').date() for dt in df['date'].values]
grace_mm = np.array(df['dS [mm]'].values)
return np.array(dt), grace_mm
def calc_polyfit(ts, order = 1):
dts_ordinal = toord(ts)
p = np.polyfit(dts_ordinal[~np.isnan(ts[1])],
ts[1][~np.isnan(ts[1])], order)
vals = np.polyval(p, dts_ordinal)
dts = [datetime.fromordinal(dt).date() for dt in dts_ordinal]
return (dts, vals)
def calc_form(tss, formula):
varias, operts = split_form(formula)
assert len(varias) == len(operts)
ds = np.zeros(tss[varias[0]][0].shape)
for vari, oper in zip(varias, operts):
if oper == '+':
ds += tss[vari][1]
elif oper == '-':
ds -= tss[vari][1]
elif oper == '/':
ds /= tss[vari][1]
elif oper == '*':
ds *= tss[vari][1]
else:
raise ValueError('Unknown operator in formula')
return (tss[varias[0]][0], ds)
def split_form(formula):
varias = re.split("\W", formula)
operts = re.split("[a-z]+", formula, flags=re.IGNORECASE)[1:-1]
while '_' in operts:
operts.remove('_')
if len(varias) > len(operts):
operts.insert(0, '+')
return varias, operts