forked from Bis-Bala/agdc_tidal
-
Notifications
You must be signed in to change notification settings - Fork 0
/
geomed_wf.py
326 lines (308 loc) · 16.1 KB
/
geomed_wf.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
from __future__ import print_function
from __future__ import absolute_import
from __future__ import division
import click
import functools
import sys
import os
import csv
import numpy as np
import rasterio
import datetime as DT
import xarray as xr
from datetime import datetime
from itertools import product
import logging
import logging.handlers as lh
import dask.array as da
import datacube.api
import fnmatch
import copy
from collections import defaultdict
from dateutil.relativedelta import relativedelta
from dateutil.rrule import rrule, YEARLY
from datacube.ui import click as ui
# from datacube.ui.expression import parse_expressions
from enum import Enum
from pathlib import Path
from datacube.storage.storage import write_dataset_to_netcdf
from rasterio.enums import ColorInterp
from gqa_filter import list_gqa_filtered_cells, get_gqa
from dateutil.rrule import rrule, YEARLY
from otps import TimePoint
from otps.predict_wrapper import predict_tide
from datacube.api.grid_workflow import Tile
from datacube_stats import statistics
from math import cos, asin, sqrt
import hdmedians as hd
logging.basicConfig()
_log = logging.getLogger('agdc-temporal-geomedian-test')
_log.setLevel(logging.INFO)
#: pylint: disable=invalid-name
required_option = functools.partial(click.option, required=True)
MY_GEO = {}
MY_DATE = {}
DEFAULT_PROFILE = {
'blockxsize': 256,
'blockysize': 256,
'compress': 'lzw',
'driver': 'GTiff',
'interleave': 'band',
'nodata': -999,
'tiled': True}
#: pylint: disable=too-many-arguments
@ui.cli.command(name='tidal-temporal-range')
@ui.global_cli_options
# @ui.executor_cli_options
# @click.command()
@required_option('--epoch', 'epoch', default=2, type=int, help='epoch like 2 5 10')
@required_option('--lon_range', 'lon_range', type=str, required=True, help='like (130.01, 130.052) under quote')
@required_option('--lat_range', 'lat_range', type=str, required=True, help='like (-13.023,-13.052) under quote')
@required_option('--year_range', 'year_range', type=str, required=True, help='2010-2017 i.e 2010-01-01 to 2017-01-01')
@click.option('--tide_post', 'tide_post', type=str, default='',
help='pick up tide post from epoch_tide_post_model.csv in current directory using Haversin algorithm for a closest cluster or provide from google map like (130.0123, -11.01)')
@click.option('--per', 'per', default=10, type=int, help='10 25 50 for low tide/high tide 10/10 25/25 50/50' )
@click.option('--season', 'season', default='dummy', type=str, help='summer winter autumn spring')
@click.option('--ls7fl', default=True, is_flag=True, help='To include all LS7 data set it to False')
@click.option('--debug', default=False, is_flag=True, help='Build in debug mode to get details of tide height within time range')
# @ui.parsed_search_expressions
# @ui.pass_index(app_name='agdc-tidal-analysis-app')
def main(epoch, lon_range, lat_range, year_range, tide_post, per, season, ls7fl, debug):
# dc = datacube.Datacube(app="tidal-range")
products = ['ls5_nbar_albers', 'ls7_nbar_albers', 'ls8_nbar_albers']
dc=datacube.Datacube(app='tidal_temporal_test')
td_info = MyTide(dc, lon_range, lat_range, products, epoch, year_range, tide_post, per, season, ls7fl, debug)
print ("Input date range " + year_range )
for (acq_min, acq_max) in td_info.get_epochs():
if season == "dummy":
print ("running task for epoch " + str(acq_min) + " TO " + str(acq_max) + " on percentile " + str(per
) + " tide post " + tide_post + " for lon/lat range " + lon_range + lat_range + " epoch " + str(epoch))
else:
print ("running task for epoch " + str(acq_min) + " TO " + str(acq_max) + " on percentile " + str(per
) + " tide post " + tide_post + " for lon/lat range " + lon_range + lat_range + " epoch " + str(epoch
) + " for season " + season)
td_info.tidal_task(acq_min, acq_max)
def pq_fuser(dest, src):
valid_bit = 8
valid_val = (1 << valid_bit)
no_data_dest_mask = ~(dest & valid_val).astype(bool)
np.copyto(dest, src, where=no_data_dest_mask)
both_data_mask = (valid_val & dest & src).astype(bool)
np.copyto(dest, src & dest, where=both_data_mask)
class MyTide():
def __init__(self, dc, lon_range, lat_range, products, epoch, year_range, tide_post, per, season, ls7fl, debug):
self.dc = dc
self.lon = eval(lon_range)
self.lat = eval(lat_range)
self.products = products
self.epoch = epoch
self.start_epoch = datetime.strptime(year_range.split('-')[0] +"-01-01", "%Y-%m-%d").date()
self.end_epoch = datetime.strptime(year_range.split('-')[1]+"-01-01", "%Y-%m-%d").date()
self.tide_post = tide_post
self.per = per
self.season = season
self.ls7fl = ls7fl
self.debug = debug
def get_epochs(self):
for dt in rrule(YEARLY, interval=self.epoch, dtstart=self.start_epoch, until=self.end_epoch):
if dt.date() >= self.end_epoch:
print ("CALCULATION finished and data available in MY_GEO and MY_DATE dictionary ")
return
acq_min = dt.date()
acq_max = acq_min + relativedelta(years=self.epoch, days=-1)
acq_min = max(self.start_epoch, acq_min)
acq_max = min(self.end_epoch, acq_max)
yield acq_min, acq_max
def distance(self, lat1, lon1, lat2, lon2):
p = 0.017453292519943295
a = 0.5 - cos((lat2 - lat1) * p)/2 + cos(lat1 * p) * cos(lat2 * p) * (1 - cos((lon2 - lon1) * p)) / 2
return 12742 * asin(sqrt(a))
def extract_otps_range(self, date_list):
# open the otps lon/lat file
tp = list()
ln = 0
la = 0
if self.tide_post:
ln = eval(self.tide_post)[0]
la = eval(self.tide_post)[1]
else:
print ("reading from tidal model file and using Haversine algorithm to extract shortest distance")
from operator import itemgetter
# first find centroid of lat lon range
la = (self.lat[0] + self.lat[1])/2
ln = (self.lon[0] + self.lon[1])/2
rdlist = list()
fname = "./epoch_tide_post_model.csv"
try:
with open (fname, 'rt') as f:
reader = csv.reader(f, delimiter=',')
for rd in reader:
rdlist.append((rd[0], rd[1], rd[2],
self.distance(la, ln, float(rd[1]), float(rd[0]))))
rdlist = sorted(rdlist, key=itemgetter(3))
print ( "Found tide post coordinates,depth and shortest distance", rdlist[0] )
la = float(rdlist[0][1])
ln = float(rdlist[0][0])
except IOError as e:
print ("Unable to open file: " +str(fname))
sys.exit()
for dt in date_list:
tp.append(TimePoint(ln, la, dt))
tides = predict_tide(tp)
if len(tides) == 0:
print ("No tide height observed from OTPS model within lat/lon range")
sys.exit()
print ("received from predict tides ", str(datetime.now()))
date_low = list()
date_high = list()
tide_dict = dict()
for tt in tides:
tide_dict[datetime.strptime(tt.timepoint.timestamp.isoformat()[0:19], "%Y-%m-%dT%H:%M:%S")] = tt.tide_m
tide_dict = sorted(tide_dict.items(), key=lambda x: x[1])
# lowest = round(float(self.per)*len(tide_dict)/100)
dr = float(tide_dict[len(tide_dict)-1][1]) - float(tide_dict[0][1])
lmr = float(tide_dict[0][1]) + dr*float(self.per)*0.01 # low tide max range
hlr = float(tide_dict[len(tide_dict)-1][1]) - dr*float(self.per)*0.01 # low tide max range
date_low = [x for x in tide_dict if x[1] <= lmr]
date_high = [x for x in tide_dict if x[1] > hlr]
# date_high = tide_dict[-int(lowest):]
print ("lowest tides range and number " + str(date_low[0][1]) + "," + str(date_low[len(date_low)-1][1])
+ " " + str(len(date_low)))
print ("highest tides range and number " + str(date_high[0][1]) + "," + str(date_high[len(date_high)-1][1])
+ " " + str(len(date_high)))
or_date_low = copy.deepcopy(date_low)
or_date_high = copy.deepcopy(date_high)
if self.debug:
print ("lowest tides list ", [[datetime.strftime(date[0], "%Y-%m-%d"), date[1]] for date in date_low])
print ("")
print ("highest tides list", [[datetime.strftime(date[0], "%Y-%m-%d"), date[1]] for date in date_high])
print ("")
print ("ALL TIDES LIST", [[datetime.strftime(date[0], "%Y-%m-%d"), date[1]] for date in tide_dict])
date_low = [dd[0] for dd in date_low]
date_high = [dd[0] for dd in date_high]
return date_low, date_high, or_date_low, or_date_high
def build_my_dataset(self, acq_min, acq_max):
nbar_data = None
dt5 = "2011-12-01"
dtt7 = "1999-07-01"
dt7 = "2003-03-01"
dt8 = "2013-04-01"
ed = acq_max
sd = acq_min
for i, st in enumerate(self.products):
prod = None
acq_max = ed
acq_min = sd
print (" doing for sensor", st )
if st == 'ls5_nbar_albers' and acq_max > datetime.strptime(dt5, "%Y-%m-%d").date() and \
acq_min > datetime.strptime(dt5, "%Y-%m-%d").date():
print ("LS5 post 2011 Dec data is not exist")
continue
elif st == 'ls5_nbar_albers' and acq_max > datetime.strptime(dt5, "%Y-%m-%d").date() and \
acq_min < datetime.strptime(dt5, "%Y-%m-%d").date():
acq_max = datetime.strptime(dt5, "%Y-%m-%d").date()
print (" epoch end date is reset for LS5 2011/12/01")
if st == 'ls7_nbar_albers' and self.ls7fl and acq_max > datetime.strptime(dt7, "%Y-%m-%d").date() and \
acq_min > datetime.strptime(dt7, "%Y-%m-%d").date():
print ("LS7 post 2003 March data is not included")
continue
elif st == 'ls7_nbar_albers' and self.ls7fl and acq_max > datetime.strptime(dt7, "%Y-%m-%d").date() and \
acq_min < datetime.strptime(dt7, "%Y-%m-%d").date():
acq_max = datetime.strptime(dt7, "%Y-%m-%d").date()
print (" epoch end date is reset for LS7 2003/03/01")
if st == 'ls7_nbar_albers' and acq_max < datetime.strptime(dtt7, "%Y-%m-%d").date() and \
acq_min < datetime.strptime(dtt7, "%Y-%m-%d").date():
continue
if st == 'ls8_nbar_albers' and acq_max < datetime.strptime(dt8, "%Y-%m-%d").date() and \
acq_min < datetime.strptime(dt8, "%Y-%m-%d").date():
continue
elif st == 'ls8_nbar_albers' and acq_max > datetime.strptime(dt8, "%Y-%m-%d").date() and \
acq_min < datetime.strptime(dt8, "%Y-%m-%d").date():
acq_min = datetime.strptime(dt8, "%Y-%m-%d").date()
if st == 'ls5_nbar_albers':
prod = 'ls5_pq_albers'
elif st == 'ls7_nbar_albers':
prod = 'ls7_pq_albers'
else:
prod = 'ls8_pq_albers'
# add extra day to the maximum range to include the last day in the search
# end_ep = acq_max + relativedelta(days=1)
indexers = {'time':(acq_min, acq_max), 'x':(str(self.lon[0]), str(self.lon[1])),
'y':(str(self.lat[0]), str(self.lat[1])), 'group_by':'solar_day'}
pq = self.dc.load(product=prod, fuse_func=pq_fuser, **indexers)
if st == 'ls5_nbar_albers' and len(pq) == 0:
print ("No LS5 data found")
continue
if nbar_data is not None and st == 'ls7_nbar_albers' and len(pq) == 0:
print ("No LS7 data found")
continue
indexers = {'time':(acq_min, acq_max), 'x':(str(self.lon[0]), str(self.lon[1])), 'y':(str(self.lat[0]), str(self.lat[1])),
'measurements':['blue', 'green', 'red', 'nir', 'swir1', 'swir2'], 'group_by':'solar_day'}
mask_clear = pq['pixelquality'] & 15871 == 15871
if nbar_data is not None:
new_data = self.dc.load(product=st, **indexers)
new_data = new_data.where(mask_clear)
nbar_data = xr.concat([nbar_data, new_data], dim='time')
else:
nbar_data = self.dc.load(product=st, **indexers)
nbar_data = nbar_data.where(mask_clear)
# if season then filtered only season data
if self.season.upper() == 'WINTER':
nbar_data = nbar_data.isel(time=nbar_data.groupby('time.season').groups['JJA'])
elif self.season.upper() == 'SUMMER':
nbar_data = nbar_data.isel(time=nbar_data.groupby('time.season').groups['DJF'])
elif self.season.upper() == 'SPRING':
nbar_data = nbar_data.isel(time=nbar_data.groupby('time.season').groups['SON'])
elif self.season.upper() == 'AUTUMN':
nbar_data = nbar_data.isel(time=nbar_data.groupby('time.season').groups['MAM'])
# filtered out lowest and highest date range
date_list = nbar_data.time.values.astype('M8[s]').astype('O').tolist()
date_low, date_high, or_date_low, or_date_high = self.extract_otps_range(date_list)
date_low = [s.strftime("%Y-%m-%d %H:%M:%S") for s in date_low]
date_high = [s.strftime("%Y-%m-%d %H:%M:%S") for s in date_high]
date_all = [s.strftime("%Y-%m-%d %H:%M:%S") for s in date_list]
low_match = [i for i, item in enumerate(date_all) if item in date_low]
high_match = [i for i, item in enumerate(date_all) if item in date_high]
nbar_low = nbar_data.isel(time=low_match)
nbar_high = nbar_data.isel(time=high_match)
print (" loaded nbar data with low " + str(datetime.now().time()))
return nbar_low, nbar_high, or_date_low, or_date_high
def tidal_task(self, acq_min, acq_max):
# gather latest datasets as per product names
ds_low, ds_high, date_low, date_high = self.build_my_dataset(acq_min, acq_max)
# calculate medoid
# For a slice of 1000:1000 for entire time seried do like
# smallds = ds_high.isel(x=slice(None, None, 4), y=slice(None, None, 4))
key = ''
print ("creating GEOMEDIAN for lower range " + str(datetime.now().time()))
med_low = statistics.combined_var_reduction(ds_low, hd.nangeomedian)
print ("creating GEOMEDIAN for higher range " + str(datetime.now().time()))
med_high = statistics.combined_var_reduction(ds_high, hd.nangeomedian)
key = ''
if self.season.upper() != 'DUMMY':
key = "GEO_" + str(acq_min) + "_" + str(acq_max) + "_" + self.season.upper() + "_LOW"
else:
key = "GEO_" + str(acq_min) + "_" + str(acq_max) + "_LOW"
MY_GEO[key] = copy.deepcopy(med_low)
MY_DATE[key] = copy.deepcopy(date_low)
key = ''
if self.season.upper() != 'DUMMY':
key = "GEO_" + str(acq_min) + "_" + str(acq_max) + "_" + self.season.upper() + "_HIGH"
else:
key = "GEO_" + str(acq_min) + "_" + str(acq_max) + "_HIGH"
MY_GEO[key] = copy.deepcopy(med_high)
MY_DATE[key] = copy.deepcopy(date_high)
if (acq_max == self.end_epoch):
print (" calculation finished and data is available in MY_GEO and MY_DATE dictionaries ",
str(datetime.now().time()))
return
if __name__ == '__main__':
'''
The program gets all LANDSAT datasets excluding post March 2003 LS7 datasets and applied cloud free pq data.
It accepts lon and lat ranges and optional tide_post. If it is not available, it pulls tidal post model data and
extracts epoch/seasonal data and calculates geomedian and low high data dictionaries for six bands and output to a
MY_GEO and MY_DATE.
MY_GEO dictionary has LOW and HIGH xarray geomedian datasets and MY_DATE has low high tidal ranges and number of datasets.
It needs small spatial range to cover wide time range.
'''
main()