[Help]: PS candidates selection

[jie666-6]

In pyGMTSAR, it seems that the PS candidates are selected based on the Amplitude Dispersion Index (ADI), as written in Stack_ps.py: psfunction = (ps.average/(2*ps.deviation))

However, in the paper "Permanent Scatterers in SAR Interferometry," the ADI is calculated as the ratio of the standard deviation to the average, which is: ADI = (ps.deviation/ps.average).

Generally, ps.deviation/ps.average<0.25 can be regarded as PS pointes.

I am wondering if you can provide any references for the calculation ps.average / (2 * ps.deviation), as I would like to check the reference and understand what threshold should be used for PSC selection.

So I am wondering if there is any reference you can provide for ps.average/(2*ps.deviation) calculation since I would like to check the reference and to know which kind of threshold should I select for PSC selection.

Additionally, the paper mentions that before conducting statistical analysis of the amplitude values, images must be radiometrically corrected to ensure comparability. This implies that radiometric correction is necessary before PSC selection, correct? In pyGMTSAR, there is temporal normalization instead of radiometric correction before calculating psfunction. Am I right, or did I miss something?

    # normalize image amplitudes (intensities)
    tqdm_dask(mean := dask.persist(data.mean(dim=['y','x'])), desc='Intensity Normalization')
    # dask.persist returns tuple
    norm = mean[0].mean(dim='date') / mean[0]
    # compute average and std.dev.
    stats = (norm * data).pipe(lambda x: (x.mean(dim='date'), x.std(dim='date')))

Thanks a lot and I am looking forward to your reply.