sensitivity of displacement time series outcomes to startdate

[hardreddata]

I recently prepared a Sentinel stack using ISCE2 for 2017 - 2021, with five connections between dates, -c 5 -f 0.2. It gave some promising results. Hooray for MintPy.

Out of curiousity I then set mintpy.network.startDate = 20190701 keeping the other parameters constant, and the displacement results I get are quite different in places. I was expecting the time series to be similar for July 2019 onwards. I am not doing any phase unwrap error correction.

Is this because the location of the points has moved? Or are there other factors here? Do any aspects of the analysis look at the time period as a whole? Maybe I have made a mistake in my stack preparation?

Any advice is appreciated.

[hardreddata]

I should rename the topic to sensitivity of displacement time series outcomes to choice of reference point.

I was doing this automatically. Using a common parameter via mintpy.reference.yx yields better outcomes. I will have to take more interest in specifying this manually per recommendation.