How to plot time-series(tsplot) for SBAS?

[Kousik Biswas]

Dear StaMPS users,
                             I want to plot time-series for SBAS results. But, when I tried the command ps_plot('v-do','ts') inside SMALL_BASELINES directory, I got this error Error using textscan Invalid file identifier.  Use fopen.. Error in ts_plot (line 27) tsmat=textscan(fid,'%s'). I am using StaMPS_v3.3b1, Matlab R2012a 64 bit. Any help regarding this issue would be appreciated.  Thanks in advance.

Regards,
Kousik

[Dan]

Hi, I think there are just a few lines missing for the SBAS case. Add these lines after line 1735 in the ps_plot.m script:

% TS PLOT preparation
 if ts_flag==1
     ts_flaghelper; % script to save neccessary matrices for TS plot
 end

It should be between the lines

units='mm/yr';

and

 case {'p'}

[Kousik Biswas]

Dear Dan,
Thank you for your reply. I followed your suggestion and I found this following

units='mm/yr';

and
 case {'p'}

at line 1299 in case of my StaMPS 3.3b1 version. I included the

% TS PLOT preparation
 if ts_flag==1
     ts_flaghelper; % script to save neccessary matrices for TS plot
 end 

portion as you have said. But I found the error as,

Error using plot
Vectors must be the same lengths.

Error in ts_plot (line 100)
    h1=plot(day,ts,'--*'); hold on
 
Error while evaluating uicontrol Callback

The reason being is the dimensions of day(36x1) vector is not same as ts(1X217). I am clueless and stuck at this point. Could you pls help me to solve this issue?

[Dan]

Hello
Sorry i did not see your answer. could you upload the ps_plot.m file from your stamps directory?

[Kousik Biswas]

Dear Dan,
I have uploaded the modified ps_plot.m after some modifications according to you suggestion. Pls see the attachment.

[Dan]

Hm strange, do you have the same problem with
ps_plot('v-d','ts')  ?

can you write in the matlab terminal in your processing folder
clear all;load('phuw2','ph_uw'); size(ph_uw)

what to do you get?

[Kousik Biswas]

Dear Dan, 
I just checked out the output of ps_plot('v-d','ts'). But, the same problem persists. After executing clear all;load('phuw2','ph_uw'); size(ph_uw) in SMALL_BASELINES directory, I got the size of ph_uw as 53196 x 36. There are 36 ifgs in PS processing, but for SBAS it is 217 in my case. I don't have a clue how to plot SBAS time-series using 217 ifgs.

[David Bekaert]

Hi,

The SB data cannot directly be displayed in a readable time-series.

Its interferograms all span different time-periods and do not have a common start time.

Your should invert your SB network to that of a SM network. 

And use the phuw2.mat information in your SB folder to display the time-series.

In the most perfect conditions, i.e. the noise is right described and there are no unwrapping errors, then ph_uw would contain the same information as phuw_sb. Only difference would be that the redundant information is removed.

Cheers,

David

--
You received this message because you are subscribed to the Google Groups "MAINSAR" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mainsar+u...@googlegroups.com.
To post to this group, send email to mai...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Kousik Biswas]

Dear David,
Thank you for your reply. I want to see comparison between ps and sb time-series of the same area. I have got the ps time-series and  want to generate sb time-series. Could you please tell me about the conversion of SB network to SM inversion procedure using matlab? How to incorporate it in StaMPS? In my case, phuw2.mat and phuw_sb2.mat are of size 53196x36 and 53196x217 respectively.

[David Bekaert]

Hi,

Your PS processing is in the INSAR_YYYYMMDD folder

There you have the phuw2.mat variable.

This is only if you did the PS processing too.

Your SB processing is in the InSAR_YYYYMMDD/SMALL_BASELINES folder.

There you have the phuw_sb2.mat variable and a phuw2.mat variable.

The latter is your SB converted to a SM network, so you would have same amount of ifgs as your PS processing.

So you do not need to do this conversion any-more.

To make the comparison between you PS and SB time-series processing you will need to make the ps_plot('ts') from the PS processing location, and the other time from the SB processing location.

In both cases the ps_plot function should use the phuw2.mat variable.

For the SB processing path, you might need to force that it goes to that variable instead of the phuw_sb one.

Hope this helps,

David

[Kousik Biswas]

Hi David,
Thank you for the clarification.

[John L]

Sorry for this stupid question. Could anyone tell me what SM network stands for?

Thank you!

John

[David Bekaert]

SM = Single Master

So a baseline plot which will look like a spider network all connecting to the master image.

D.