/
find_emissions_that_caused_air_quality_change.py
207 lines (168 loc) · 8.08 KB
/
find_emissions_that_caused_air_quality_change.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
#!/usr/bin/env python3
import os
import time
import glob
import joblib
from itertools import islice
import pandas as pd
import numpy as np
import xarray as xr
import geopandas as gpd
import xesmf as xe
import dask.bag as db
from dask_jobqueue import SGECluster
from dask.distributed import Client
output = "PM2_5_DRY"
#output = "o3_6mDM8h_ppb"
year_range = '2015-2016'
reference_year = year_range.split('-')[0]
comparison_year = year_range.split('-')[1]
if output == "o3_6mDM8h_ppb":
emulator_output = "o3_6mDM8h"
else:
emulator_output = output
use_10percent_intervals = False
use_20percent_intervals = True
if use_10percent_intervals:
sub_folder = year_range
n_jobs = 20
walltime='01:00:00' # also change for .bash)
emission_configs = np.array(
np.meshgrid(
np.linspace(0.2, 1.3, 12),
np.linspace(0.2, 1.3, 12),
np.linspace(0.2, 1.3, 12),
np.linspace(0.2, 1.3, 12),
np.linspace(0.2, 1.3, 12),
)
).T.reshape(-1, 5)
elif use_20percent_intervals:
sub_folder = f'{year_range}_20percentintervals'
n_jobs = 5
walltime='00:05:00' # also change for .bash)
emission_configs = np.array(
np.meshgrid(
np.linspace(0.2, 1.2, 6),
np.linspace(0.2, 1.2, 6),
np.linspace(0.2, 1.2, 6),
np.linspace(0.2, 1.2, 6),
np.linspace(0.2, 1.2, 6),
)
).T.reshape(-1, 5)
df_obs = pd.read_csv(
f"/nobackup/earlacoa/machinelearning/data_annual/china_measurements_corrected_uptofeb2021/df_obs_o3_6mDM8h_ppb_PM2_5_DRY.csv",
index_col="datetime",
parse_dates=True,
)
# stations left
obs_files = glob.glob(f"/nobackup/earlacoa/machinelearning/data_annual/china_measurements_corrected_uptofeb2021/*.nc")
obs_files = [f"{obs_file[-8:-3]}" for obs_file in obs_files]
obs_files_completed = glob.glob(f"/nobackup/earlacoa/machinelearning/data_annual/find_emissions_that_match_change_air_quality/{sub_folder}_adjusted_scaled/*{output}*")
obs_files_completed = [f"{item[-12:-7]}" for item in obs_files_completed]
obs_files_remaining_set = set(obs_files) - set(obs_files_completed)
obs_files_remaining = [item for item in obs_files_remaining_set]
print(f"custom outputs remaining for {output}: {len(obs_files_remaining)}")
station_id = obs_files_remaining[0]
lat = df_obs.loc[df_obs.station_id == station_id].station_lat.unique()[0]
lon = df_obs.loc[df_obs.station_id == station_id].station_lon.unique()[0]
obs_change_abs = {}
obs_change_per = {}
baselines = {}
targets = {}
target_diffs = {}
change_per = 100 * ((
df_obs.loc[df_obs.station_id == station_id][output][comparison_year].values[0]
/ df_obs.loc[df_obs.station_id == station_id][output][reference_year].values[0]
) - 1)
change_abs = (
df_obs.loc[df_obs.station_id == station_id][output][comparison_year].values[0]
- df_obs.loc[df_obs.station_id == station_id][output][reference_year].values[0]
)
obs_change_abs.update({f"{station_id}_{output}": change_abs})
obs_change_per.update({f"{station_id}_{output}": change_per})
with xr.open_dataset(
f"/nobackup/earlacoa/machinelearning/data_annual/predictions/{emulator_output}_adjusted_scaled/ds_RES1.0_IND1.0_TRA1.0_AGR1.0_ENE1.0_{emulator_output}_popgrid_0.25deg_adjusted_scaled.nc"
)[emulator_output] as ds:
baseline = (
ds.sel(lat=lat, method="nearest").sel(lon=lon, method="nearest").values
)
baselines.update({f"{station_id}_{output}": baseline})
target_abs = baseline + change_abs
target_per = baseline * (1 + (change_per / 100))
target = np.mean([target_abs, target_per])
targets.update({f"{station_id}_{output}": target})
target_diffs.update({f"{station_id}_{output}": target - baseline})
def filter_emission_configs(emission_config):
station_diffs_abs = {}
station_diffs_per = {}
target_diffs_abs = {}
target_diffs_per = {}
inputs = emission_config.reshape(-1, 5)
filename = f"RES{inputs[0][0]:.1f}_IND{inputs[0][1]:.1f}_TRA{inputs[0][2]:.1f}_AGR{inputs[0][3]:.1f}_ENE{inputs[0][4]:.1f}"
with xr.open_dataset(
f"/nobackup/earlacoa/machinelearning/data_annual/predictions/{emulator_output}_adjusted_scaled/ds_{filename}_{emulator_output}_popgrid_0.25deg_adjusted_scaled.nc"
)[emulator_output] as ds:
prediction = (
ds.sel(lat=lat, method="nearest").sel(lon=lon, method="nearest").values
)
target_diff_abs = targets[f"{station_id}_{output}"] - prediction
target_diff_per = (100 * (prediction / targets[f"{station_id}_{output}"])) - 100
if abs(target_diff_per) < 1: # +/- 1% of target
target_diffs_abs.update({filename: target_diff_abs})
target_diffs_per.update({filename: target_diff_per})
station_diffs_abs.update({f"{station_id}_{output}": target_diffs_abs})
station_diffs_per.update({f"{station_id}_{output}": target_diffs_per})
return station_diffs_abs, station_diffs_per
def main():
# dask cluster and client
n_processes = 1
n_workers = n_processes * n_jobs
cluster = SGECluster(
interface="ib0",
walltime=walltime,
memory=f"32 G",
resource_spec=f"h_vmem=32G",
scheduler_options={
"dashboard_address": ":5761",
},
job_extra=[
"-cwd",
"-V",
f"-pe smp {n_processes}",
f"-l disk=32G",
],
local_directory=os.sep.join([os.environ.get("PWD"), "dask-find-emis-pm-space"]),
)
client = Client(cluster)
cluster.scale(jobs=n_jobs)
time_start = time.time()
# dask bag over emission_configs
print(f"predicting over {len(emission_configs)} emission configs for {station_id} ...")
bag_emission_configs = db.from_sequence(emission_configs, npartitions=n_workers)
results = bag_emission_configs.map(filter_emission_configs).compute()
station_diffs_abs = [result[0] for result in results]
station_diffs_per = [result[1] for result in results]
key = [key for key in baselines.keys()][0]
station_diffs_abs = [station_diff_abs for station_diff_abs in station_diffs_abs if len(station_diff_abs[key]) > 0]
station_diffs_per = [station_diff_per for station_diff_per in station_diffs_per if len(station_diff_per[key]) > 0]
merged_per = {}
for station_diff_per in station_diffs_per:
merged_per = {**merged_per, **station_diff_per[key]}
merged_abs = {}
for station_diff_abs in station_diffs_abs:
merged_abs = {**merged_abs, **station_diff_abs[key]}
station_diffs_per = {key: merged_per}
station_diffs_abs = {key: merged_abs}
joblib.dump(obs_change_abs, f"/nobackup/earlacoa/machinelearning/data_annual/find_emissions_that_match_change_air_quality/{sub_folder}_adjusted_scaled/obs_change_abs_{output}_{station_id}.joblib")
joblib.dump(obs_change_per, f"/nobackup/earlacoa/machinelearning/data_annual/find_emissions_that_match_change_air_quality/{sub_folder}_adjusted_scaled/obs_change_per_{output}_{station_id}.joblib")
joblib.dump(baselines, f"/nobackup/earlacoa/machinelearning/data_annual/find_emissions_that_match_change_air_quality/{sub_folder}_adjusted_scaled/baselines_{output}_{station_id}.joblib")
joblib.dump(targets, f"/nobackup/earlacoa/machinelearning/data_annual/find_emissions_that_match_change_air_quality/{sub_folder}_adjusted_scaled/targets_{output}_{station_id}.joblib")
joblib.dump(target_diffs, f"/nobackup/earlacoa/machinelearning/data_annual/find_emissions_that_match_change_air_quality/{sub_folder}_adjusted_scaled/target_diffs_{output}_{station_id}.joblib")
joblib.dump(station_diffs_abs, f"/nobackup/earlacoa/machinelearning/data_annual/find_emissions_that_match_change_air_quality/{sub_folder}_adjusted_scaled/station_diffs_abs_{output}_{station_id}.joblib")
joblib.dump(station_diffs_per, f"/nobackup/earlacoa/machinelearning/data_annual/find_emissions_that_match_change_air_quality/{sub_folder}_adjusted_scaled/station_diffs_per_{output}_{station_id}.joblib")
time_end = time.time() - time_start
print(f"completed in {time_end:0.2f} seconds, or {time_end / 60:0.2f} minutes, or {time_end / 3600:0.2f} hours")
client.close()
cluster.close()
if __name__ == "__main__":
main()