/
CS_measure.m
217 lines (196 loc) · 6.22 KB
/
CS_measure.m
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
function CS = CS_measure(event,sta1,sta2,parameters)
% Main function to perform GSDF measurement
%
setup_ErrorCode;
isdebug = 0;
isfigure = 0;
refv = parameters.refv;
refphv = parameters.refphv;
periods = parameters.periods;
min_width = parameters.min_width;
max_width = parameters.max_width;
wintaperlength = parameters.wintaperlength;
prefilter = parameters.prefilter;
xcor_win_halflength = parameters.xcor_win_halflength;
Nfit = parameters.Nfit;
Ncircle = parameters.Ncircle;
xcor_win_iter = parameters.xcor_win_iter;
if length(xcor_win_iter) ~= length(periods)
disp('xcor_win_iter should be the same size as periods!');
if length(xcor_win_iter) < length(periods)
xcor_win_iter = [zeros(1,length(periods) - length(xcor_win_iter)),xcor_win_iter];
end
end
CS = init_CSstruct;
v1 = event.winpara(1); t1=event.winpara(2); v2=event.winpara(3); t2=event.winpara(4);
CS.sta1 = sta1;
CS.sta2 = sta2;
if isdebug
disp([num2str(sta1),num2str(sta2)]);
end
if length(event.stadata(sta1).data) < 20 || length(event.stadata(sta2).data) < 20
CS.exitflag(:) = ErrorCode.sta_lackdata;
return
end
% read in data for station 1 and apply prefilter
data1 = event.stadata(sta1).data;
bgtime = event.stadata(sta1).otime - event.otime;
dt1 = event.stadata(sta1).delta;
Nt = length(event.stadata(sta1).data);
fN = 1/2/dt1;
[b,a] = butter(2,[1/prefilter(2)/fN, 1/prefilter(1)/fN]);
data1 = filtfilt(b,a,data1);
taxis1 = bgtime + [0:Nt-1]'*dt1;
dist1 = event.stadata(sta1).dist;
winbgt = dist1/v1+t1;
winendt = dist1/v2+t2;
if taxis1(1) > winbgt || taxis1(end) < winendt
CS.exitflag(:) = ErrorCode.sta_lackdata;
return
end
% read in data for station 2 and apply prefilter
data2 = event.stadata(sta2).data;
bgtime = event.stadata(sta2).otime - event.otime;
dt2 = event.stadata(sta2).delta;
Nt = length(event.stadata(sta2).data);
fN = 1/2/dt2;
[b,a] = butter(2,[1/prefilter(2)/fN, 1/prefilter(1)/fN]);
data2 = filtfilt(b,a,data2);
taxis2 = bgtime + [0:Nt-1]'*dt2;
dist2 = event.stadata(sta2).dist;
winbgt = dist2/v1+t1;
winendt = dist2/v2+t2;
if taxis2(1) > winbgt || taxis2(end) < winendt
CS.exitflag(:) = ErrorCode.sta_lackdata;
return
end
% resample the data if necessary
if dt1 > dt2
new_taxis2 = taxis2(1):dt1:taxis2(end);
data2 = interp1(taxis2,data2,new_taxis2);
taxis2 = new_taxis2;
dt2 = dt1;
elseif dt1 < dt2
new_taxis1 = taxis1(1):dt2:taxis1(end);
data1 = interp1(taxis1,data1,new_taxis1);
taxis1 = new_taxis1;
dt1 = dt2;
end
% window data2
winbgt = dist2/v1+t1;
winendt = dist2/v2+t2;
win_data2 = flat_hanning_win(taxis2,data2,winbgt,winendt,wintaperlength);
% apply cross-correlation
[xcor,lag] = xcorr(data1,win_data2,...
floor(10*max(periods)/dt1+abs(taxis1(1)-taxis2(1))));
lag = lag.*dt1;
lag = lag + taxis1(1) - taxis2(1);
if isfigure
figure(43)
clf
subplot(3,1,1)
plot(taxis1,data1);
xlim([0 dist2/2])
subplot(3,1,2)
plot(taxis2,win_data2);
xlim([0 dist2/2])
subplot(3,1,3)
plot(lag,xcor);
xlim([-1000 1000])
end
%Find the window center (max amplitude within the window)
win_cent_t = (dist1-dist2)/refv;
search_win_ind = find( lag > win_cent_t-xcor_win_halflength &...
lag < win_cent_t + xcor_win_halflength );
[max_xcor_amp win_cent_i] = max(xcor(search_win_ind));
win_cent_i = search_win_ind(win_cent_i);
win_cent_t = lag(win_cent_i);
CS.win_cent_t = win_cent_t;
CS.ddist = dist1 - dist2;
% apply the window function
%win_xcor = hanning_win(lag,xcor,win_cent_t,xcor_win_halflength*2);
win_xcor = flat_hanning_win(lag,xcor,win_cent_t-xcor_win_halflength,win_cent_t+xcor_win_halflength,round(xcor_win_halflength/2));
if isfigure
figure(44)
clf
subplot(2,1,1)
plot(lag,xcor);
xlim([-500 500])
subplot(2,1,2)
plot(lag,win_xcor);
xlim([-500 500])
end
% Apply Narrow-band filter
clear gaus_filters nband_win_xcors
Nt = length(win_xcor);
[gaus_filters,faxis] = build_gaus_filter(1./periods,dt1,Nt,min_width,max_width);
fft_win_xcor = fft(win_xcor);
if size(fft_win_xcor) == 1
fft_win_xcor = fft_win_xcor';
end
for ip = 1:length(periods)
nband = fft_win_xcor .* [gaus_filters(:,ip); zeros(Nt-length(gaus_filters(:,ip)),1)];
nband = ifft(nband);
nband = 2*real(nband);
nband_win_xcors(:,ip) = nband;
end % end of periods loop
% fitting with five-parameter wavelet
for ip = 1:length(periods)
[para,resnorm,residual, exitflag] = gsdffit(nband_win_xcors(:,ip),lag,1./periods(ip),Nfit);
CS.fitpara(:,ip) = para(:);
CS.fiterr(ip) = resnorm./para(1)^2./Nfit./periods(ip);
CS.dtp(ip) = para(4);
CS.dtg(ip) = para(5);
CS.amp(ip) = para(1);
CS.w(ip) = para(2);
CS.sigma(ip) = para(3);
CS.exitflag(ip) = exitflag;
end
% Iteratively correct for windowing effect
for ip = 1:length(periods)
if xcor_win_iter(ip)
% re-center the window
win_cent_t = CS.dtg(ip);
%win_xcor = hanning_win(lag,xcor,win_cent_t,xcor_win_halflength*2);
win_xcor = flat_hanning_win(lag,xcor,win_cent_t-xcor_win_halflength,win_cent_t+xcor_win_halflength,round(xcor_win_halflength/2));
fft_win_xcor = fft(win_xcor);
if size(fft_win_xcor) == 1,fft_win_xcor = fft_win_xcor'; end
% narrow-band filter
nband = fft_win_xcor .* [gaus_filters(:,ip); zeros(Nt-length(gaus_filters(:,ip)),1)];
nband = ifft(nband);
nband = 2*real(nband);
% fit the wavelet again
[para,resnorm,residual, exitflag] = gsdffit(nband(:),lag,1./periods(ip),Nfit);
CS.fitpara(:,ip) = para(:);
CS.fiterr(ip) = resnorm./para(1)^2./Nfit./periods(ip);
CS.dtp(ip) = para(4);
CS.dtg(ip) = para(5);
CS.amp(ip) = para(1);
CS.w(ip) = para(2);
CS.sigma(ip) = para(3);
CS.exitflag(ip) = exitflag;
end
end
% Correct for cycle skipping for dtp
for ip = 1:length(periods)
syndtp = CS.ddist./refphv(ip);
testdtp = CS.dtp(ip) + [-Ncircle:Ncircle]*periods(ip);
[temp besti] = min(abs(testdtp - syndtp));
CS.dtp(ip) = testdtp(besti);
end
if isfigure
figure(45)
clf
hold on
[xi yi] = ndgrid(lag,periods);
for ip = 1:length(periods)
norm_nbands(:,ip) = nband_win_xcors(:,ip)./max(abs(nband_win_xcors(:,ip)));
end
contourf(xi,yi,norm_nbands);
for ip=1:length(periods)
plot(CS.dtp(ip),periods(ip),'kx','linewidth',2);
end
xlim([-3*max(periods) 3*max(periods)]);
pause
end
end % end of function