forked from mpanning/OnDeckInSight
-
Notifications
You must be signed in to change notification settings - Fork 0
/
windspeed_comp.py
243 lines (195 loc) · 7.34 KB
/
windspeed_comp.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
"""
Compare windspeed and vertical component on-deck SP measurements correcting
for Viking response
Also make wind and SP time domain comparison figure
"""
from obspy import read, read_inventory
from obspy.core import Stream, UTCDateTime
from mars_tools import insight_time
import math
import matplotlib.pyplot as plt
import scipy.fftpack
import numpy as np
from tqdm import tqdm
import csv
SPdatafile = "datafiles/SHU68.ELYS0.mseed"
windsdatafile = "datafiles/SHU_windspeed.mseed"
metadata = "datafiles/ELYS0.all.dl0226.response.xml"
# Read SP data and remove instrument response and plot acceleration
st_SP = read(SPdatafile)[1:] # First trace has bad response for some reason
inv = read_inventory(metadata)
pre_filt = (0.01, 0.02, 4.5, 5.0) # Only looking above 0.01 Hz
st_SP.attach_response(inv)
st_SP.remove_response(output="ACC", pre_filt=pre_filt)
st_SP.plot(outfile='SP_acc.png', method='full')
st_zoom = st_SP.copy()
zoomst = UTCDateTime(2018,12,13,5,53)
zoomet = UTCDateTime(2018,12,13,12,59)
st_zoom.trim(starttime=zoomst, endtime=zoomet)
st_zoom.plot(method='full', outfile='SP_zoom.png')
# Reread and convert to Viking DU
st_SP = read(SPdatafile)[1:] # First trace has bad response for some reason
pre_filt = (0.01, 0.05, 4.5, 5.0) # Only looking above 0.01 Hz
st_SP.attach_response(inv)
st_SP.remove_response(output="DISP", pre_filt=pre_filt)
# Viking response
mag3hz=218000;
w0=2.0*math.pi*4.0;
Q=.6;
QF=.707;
p0=complex(0.,2.*math.pi*2.8);
wc=w0;
GN=0.76e-3
HN=0.44e-3
alpha=2;
VIK0=1./(w0*w0+w0/Q*p0+p0*p0)/((1.+p0/wc/QF+(p0/wc)**2)**2.25/p0*(1.+math.pi*.5/p0))**alpha;
# The following snippet makes the frequency dependent magnification
# for i=1:200
# freqvik(i)=10^(-4+(i-1)/200*6);
# p=complex(0.,2*pi*freqvik(i));
# VIK(i)=mag3hz/VIK0/(w0*w0+w0/Q*p+p*p)/((1.d0+p/wc/QF+(p/wc)^2)^2.25/p*(1.+pi*.5/p))^alpha;
# accvik(i)=abs(GN/VIK(i)*p^2);
# depvik(i)=abs(GN/VIK(i));
# end
# loglog(freqvik,1./depvik)
# Plot a sample Viking response
# tr = st_SP[0]
# y = tr.data
# dt = tr.stats['delta']
# ts = np.arange(0, len(y)*dt, dt)
# yf = scipy.fftpack.fft(y)
# freqs = scipy.fftpack.fftfreq(len(y), dt)
# VIK = np.zeros_like(yf)
df = 0.001
freqs = np.arange(df, 100000*df, df)
VIK = np.zeros_like(freqs, dtype=np.complex)
for i, freq in enumerate(freqs):
if freq == 0:
VIK[i] = 0 # ignore DC term
else:
p = complex(0., 2.*math.pi*freq)
VIK[i] = mag3hz/VIK0/GN/(w0*w0+w0/Q*p+p*p)/((1.+p/wc/QF+(p/wc)**2)**2.25/p*(1.+math.pi*0.5/p))**alpha
fig = plt.figure()
plt.loglog(freqs,np.abs(VIK))
fig.savefig("vikingresponse.png")
plt.close(fig)
# Try to do fft of data and multiply by Viking response
st_SP_IR = st_SP.copy()
for tr in st_SP:
y = tr.data
dt = tr.stats['delta']
ts = np.arange(0, len(y)*dt, dt)
yf = scipy.fftpack.fft(y)
freqs = scipy.fftpack.fftfreq(len(y), dt)
VIK = np.zeros_like(yf)
for i, freq in enumerate(freqs):
if freq == 0:
VIK[i] = 0 # ignore DC term
else:
p = complex(0., 2.*math.pi*freq)
VIK[i] = mag3hz/VIK0/GN/(w0*w0+w0/Q*p+p*p)/((1.+p/wc/QF+(p/wc)**2)**2.25/p*(1.+math.pi*0.5/p))**alpha
# # Plot Viking response
# plt.plot(freqs, np.abs(VIK))
# plt.show()
# Convert displacement to Viking DU
vikf = yf*VIK
yvik = scipy.fftpack.ifft(vikf)
# # plot original
# plt.plot(ts, y)
# plt.show()
# # plot viking du
# plt.plot(ts, yvik.real)
# plt.show()
# Replace data
tr.data = yvik.real
#st_SP.plot(method='full')
# Read windspeed data
st_WS = read(windsdatafile).select(channel='VWS')
# (st_SP + st_WS).plot(equal_scale=False, method='full')
# Create a stream just with winds that overlap SP data
st_windoverlap = Stream()
for tr in st_SP:
stime = tr.stats['starttime']+2. # strip out first two seconds with transient
etime = tr.stats['endtime']
st_temp = read(windsdatafile, starttime=stime, endtime=etime).select(channel='VWS')
st_windoverlap += st_temp
st_windoverlap = st_windoverlap.merge().split()
# (st_windoverlap+st_SP).plot(method='full', equal_scale=False)
# One more plot with original
# (st_windoverlap+st_SP_IR).plot(method='full', equal_scale=False)
st_windoverlap.decimate(15) #reduce sampling to 30 seconds
# (st_windoverlap+st_SP).plot(method='full', equal_scale=False)
print(st_windoverlap)
# Test comparison
# Initiate figure instance and clear csv file for output
fig = plt.figure()
csvfile = 'WS_SPviking_rms.csv'
with open(csvfile, 'w') as f: # This should empty the file and add headers
writer = csv.writer(f)
writer.writerow(["RMS wind speed", "RMS SP Viking output"])
for tr in tqdm(st_windoverlap):
stime = tr.stats['starttime']
etime = tr.stats['endtime']
st_SPtemp = st_SP.copy().trim(starttime=stime, endtime=etime)
st_WStemp = st_WS.copy().trim(starttime=stime, endtime=etime)
dt = tr.stats['delta']
meanSPsig = np.zeros_like(tr.data)
meanWSsig = np.zeros_like(tr.data)
try:
SP_times = st_SPtemp[0].times(reftime=stime)
WS_times = st_WStemp[0].times(reftime=stime)
except IndexError:
continue
# tr_times = tr.times()
for i, sample in enumerate(tr.data):
meanSPsig[i] = np.sqrt(np.mean(st_SPtemp[0].data[np.logical_and(SP_times>i*dt, SP_times<(i+1)*dt)]**2))
meanWSsig[i] = np.sqrt(np.mean(st_WStemp[0].data[np.logical_and(WS_times>i*dt, WS_times<(i+1)*dt)]**2))
plt.loglog(meanWSsig, meanSPsig, 'ko')
rows = np.transpose(np.array([meanWSsig, meanSPsig]))
with open(csvfile, 'a') as f: # Append to csv file for later plotting
writer = csv.writer(f)
for row in rows:
writer.writerow(row)
# Redo the last one for label purposes
plt.loglog(meanWSsig, meanSPsig, 'ko',
label='InSight SP converted to Viking response')
xlimits = (1, 20)
ylimits = (0.1, 120)
plt.xlim(xlimits)
plt.xlabel("Mean wind speed (m/s)")
plt.ylim(ylimits)
plt.ylabel("Mean seismic noise (Viking DU)")
# Reread all data to calculate best fit to data assuming a slope of 2 in
# log log space
WSall = []
SPall = []
with open(csvfile, 'r') as f:
reader = csv.reader(f)
csv_headers = next(reader) # header row
for row in reader:
if (not np.isnan(float(row[0]))) and (not np.isnan(float(row[1]))):
WSall.append(float(row[0]))
SPall.append(float(row[1]))
# With fixed slope m, least squares intercept is b = mean(y) - m*mean(x)
WSall = np.array(WSall)
SPall = np.array(SPall)
m_fixed = 2.0
b_fit = np.mean(np.log10(SPall)) - m_fixed*np.mean(np.log10(WSall))
print("Best fit intercept: {}".format(b_fit))
yISfit = (math.pow(10., (m_fixed*math.log10(xlimits[0]) + b_fit)),
math.pow(10., (m_fixed*math.log10(xlimits[1]) + b_fit)))
plt.plot(np.array(xlimits), np.array(yISfit), 'g-',
label='InSight fit')
# Add in empirical fit from Viking
# Line on Viking fig is log(DU) = 2(log(ws/2.5))
yVfit = (math.pow(10., 2.*math.log10(xlimits[0]/2.5)),
math.pow(10., 2.*math.log10(xlimits[1]/2.5)))
plt.plot(np.array(xlimits), np.array(yVfit), 'r-',
label='Viking-2 fit (Anderson et al., 1977)')
plt.legend(loc='lower right')
fig.savefig("WS_SPviking_rms.png")
# # Loop over wind traces and calculate mean SP_viking over each sample
# for tr in st_windowoverlap:
# stime = tr.stats['starttime']
# etim = tr.stats['endtime']
# st_temp = st_SP.copy().trim(starttime=stime, endtime=etime)