Dear all,
Currently, I’m practising my InSAR process understanding by trying to reproduce the case study of the Maoxian landslide using 45 Sentinel1 SAR images. ( Intrieri et al. 2018) (see the picture below)
And the picture below is my PSI analysis result where I am using SNAP,snap2stamps, StaMPS (from step 1 to 7), and StaMPS-Visualizer. I have tried to modify StaMPS parameters following Hoser, 2018 thesis, but I still haven’t gotten the expected results.
setparm(‘unwrap_time_win’, 24)
setparm(‘unwrap_grid_size’, 10)
setparm(‘unwrap_gold_n_win’,8)
setparm(‘scn_time_win’,50)
setparm('scla_deramp', 'y')
setparm('ref_radius',40)
setparm('ref_centre_lonlat',[103.646, 32.0768])
(My results look very noisy and have no acceleration trend)
My question is, what should I can do to improve the analysis? (accelerate pattern?) Do I miss something?
Thank you so much. Please feel free to discuss this with me.