LAB 04 - DECONVOLUTION SHOT 4,5,6

MATLAB SCRIPT LAB 4.1

clear,clc,close all
load SeismicData_gain_bpf_C

%% Extracting shot and geometrical acquisition parameters
shot_num=4:6;
p=1;
[Dshot,dt,dx,t,offset]=extracting_shots(Dbpf,H,shot_num,p);
[nt,nx]=size(Dshot);

%% Before Applying Spiking deconvolution
scale=1;
figure(1);
mwigb(Dshot,scale,offset,t)
xlabel('Trace number','FontSize',14)
ylabel('Time(s)','FontSize',14)

Figure 1: shot gathers 4,5,6 before applying spiking deconvolution

%% Apply auto correlogram function
max_lag=0.2;
[Dauto,lags]=auto_correlation_map(Dshot,max_lag,dt);

%% Displaying auto correlogram
scale=5;
figure(2);
mwigb(Dauto,scale,offset,lags)
xlabel('Trace number','FontSize',14)
ylabel('Time lag(s)','FontSize',14)

Figure 2: auto-correlograms of shot gathers 4,5,6

%% Applying Spiking deconvolution
mu=0.1;
Ds=spiking_decon(Dshot,max_lag,mu,dt);
​
%% Display After Applying Spiking deconvolution
scale=1;
figure(3)
mwigb(Ds,scale,offset,t)
xlabel('Trace number','FontSize',14)
ylabel('Time(s)','FontSize',14)

Figure 3: shot gathers 4,5,6 after applying spiking deconvolution​

%% Generate Normalised PSD Figure 4
Davg_before=mean(Dshot');
fs=1/dt;
[Davg_before,b4_f]=periodogram(Davg_before,[],2*nt,fs);
Davg_before=Davg_before/max(Davg_before);
Davg_before=20*log10(abs(Davg_before));

Davg_after=mean(Ds');
[Davg_after,af_f]=periodogram(Davg_after,[],2*nt,fs);
Davg_after=Davg_after/max(Davg_after);
Davg_after=20*log10(abs(Davg_after));

f=linspace(-0.5,0.5,2*nt)/dt;
figure,plot(b4_f,Davg_before,af_f,Davg_after,'r--')
xlabel('Frequency(Hz)','FontSize',14)
ylabel('Normalised PSD','FontSize',14)
grid
legend('Before Decon','After Decon')

Figure 4: PSD of the average trace of shot gathers 4,5 & 6 before and after spiking deconvolution

%% Apply gain
agc_gate=0.5;
T=2;
Ds_gain=AGCgain(Ds,dt,agc_gate,T);

%% Save
clear af_f b4_f Dauto Davg_after Davg_before Dbpf Dshot dt dx f fs lags max_lag mu nt nx offset p scale shot_num t T agc_gate Ds
save SeismicData_gain_bpf_sdecon_gain_C

You will notice that the data become spikier and you can further analyze it in the spectrum domain via the power spectral density of the average trace as in Figure 4.