I have a question. I generated interferogram of a zone for different dates and now i would make a time serie for this area.
How can i do this ?
Which steps are requiered?
One option is to create stack of products, or collocation, and then use time series operator for each point,
Other option could be create vector polygon for the interested area and use statistics to extract the mean for example for each correspondent interfergram and plot the mean correspondent to each date,
Also an option to extract the correspondent pixels (Area) for each different interfergram date, as CSV and use python to plot the histogram,
Thank you for your answer !
If I choose the first option, I have lot of interferograms with the same master image (the same date), I have to create stack of products for each pair of interferograms?And after use the possibility of create time serie with this products ?
Let’s assume that you have the following pairs, of created interfergrams, create stacking, allocation for all of them,
Then your time series will be read as the first time start is 20150101 and the second is 20150113 the third one is 20150125 and the next one will be 20150206, and so on, until the last pairs,
Okay with this example. I have to make stack for each pair?
Because I tried to do one track for all of them but I can’t make the time serie with the result
Yes, for all of them and the Time series should function properly,
Time feature on SNAP should work…if all the data has exactly the same size I think.
However, despite SNAP capabilities I am concern about the processing you are trying to do.
I hope you know that for time series of interferograms are several advanced techniques that help you to get time series with a better estimation of the signal (time series deformation) by removing atmospheric noise, dem errors, etc . Those techniques are quite known (PSI, SBAS, QSBAS, etc…)
Please be aware the way of time series here mentioned on this thread until now does not remove anything and your time series of interferograms defined above will contain of kind of signal on it, as you would had not remove anything… so all ‘noise’ and errors on the measurement and processing will be on it.
So please consider to use some of the proven techniques if you want to get rid of the signal error included on single interferogram processing (DInSAR) such as atmospheric noise, etc. So check the literature and choose the technique more apropriate for your application. If it is the one suggested by @falahfakhri, then go ahead, otherwise try to follow the proper techniques to do that.
Thanks for all your answers, I can generate time series with this process but I will check if some mistakes are on my result. I have just an other question, I can generate time series of only real, coherence or imaginary number. I want a time serie of the deformation and i don’t know i can i do this. Do you have an idea?
As @mdelgado referred in his previous post, getting the real time series of displacement, there are many techniques are already available and designed for this reason talking in account removing the errors of Atmosphere and residual top., and others to get the real millimetre ground movement , in fact in my previous post I only pointed out to the time series operator’s use, in the coherence example , I think time series operator as I explained and as you did is enough to get the time series of cross correlation coefficient of dataset.
Do the time series creation requires interferometric pairs between multiple slave acquisition and only one master acquisition like @falahfakhri has mentioned? Can’t I create a time series from a series of interferograms with a short temporal baseline and thus multiple master images? like the following
and so on…