unpackFrame_ERS.py for ESA SLC data with stackStripMap.py

[XJQiao]

Hi developers and all,

I hope to process ERS data (i.e., ESA SLC data) using the stackStripMap.py. Before stack processing, I used the unpackFrame_ERS.py to unpack the data.

The issue I faced was how to tackle and pass right numbers for the Doppler coefficients:

1. The coeffs = obj.doppler_coeff in the original code was a bug, because AttributeError: 'ERS_EnviSAT_SLC' object has no attribute 'doppler_coeff'. And this seems to be similar to [(https://github.com/Problems encountered in using unpackFrame_ENV.py #488)].
2. I tried to used coeffs = [0.,0.], which was commented in the original code. Then, the igrams were all random as shown in the figure. And the reason should be I was using wrong numbers for the Doppler coefficients.
3. I printed out the variables of the ERS_EnviSAT_SLC object (as shown in the attachment of 'ERS_EnviSAT_SLC.odt'). The Doppler numbers seems to be very big, like 'doppler': (-45.646583557128906, -327422.4375, 555750464.0, 0.0, 0.0).

Below is the unpackFrame_ERS.py with some adjustments of mine to fit the ESA SLC data.

Much appreciate in advance for any comments and thoughts you have on my case.

def unpack(hdf5, slcname):
    '''
    Unpack HDF5 to binary SLC file.
    '''
    imgname = glob.glob(os.path.join(hdf5,'SAR*'))[0] # My adjustment to pass ESA data
    #imgname = glob.glob(os.path.join(hdf5,'DAT*'))[0] # My adjustment
    #ldrname = glob.glob(os.path.join(hdf5, 'LEA*'))[0] # My adjustment
    if not os.path.isdir(slcname):
        os.mkdir(slcname)

    date = os.path.basename(slcname)
    obj = createSensor('ERS_EnviSAT_SLC') #My adjustment to replace original ERS_SLC 
    obj.configure()
    #obj._leaderFile = ldrname # My adjustment
    obj._imageFileName = imgname
    obj._orbitType = 'ODR' 
    obj._orbitDir = '/home/jovyan/Orbit/ERS' # My adjustment
    obj.output = os.path.join(slcname, date+'.slc')

    #print(obj._leaderFile)
    print(obj._imageFile)
    print(obj.output)
    
    obj.extractImage()
    obj.frame.getImage().renderHdr()
    obj.extractDoppler()

    coeffs = [0.,0.] # My adjustment
    dr = obj.frame.getInstrument().getRangePixelSize()
    r0 = obj.frame.getStartingRange()

    poly = Poly1D.Poly1D()
    poly.initPoly(order=len(coeffs)-1)
    poly.setCoeffs(coeffs)
    pickName = os.path.join(slcname, 'data')
    with shelve.open(pickName) as db:
        db['frame'] = obj.frame
        db['doppler'] = poly

ERS_EnviSAT_SLC.odt

[vittorio3d]

Hy XJQiao, I tried your script in processing isce-to-stamps with few adjustament :
###############################################
def unpack(hdf5, slcname):
'''
Unpack HDF5 to binary SLC file.
'''

fname = glob.glob(os.path.join(hdf5,'SAR*.E*'))[0] # vittorio adjustament
if not os.path.isdir(slcname):
    os.mkdir(slcname)
    

date = os.path.basename(slcname)
obj = createSensor('ERS_Envisat_SLC')
obj._imageFileName = fname
obj._orbitType = 'ODR'
obj._orbitDir = '/media/vittorio/Dati_satelliti/dati_isce/orbits/ers2/ODR'
obj.output = os.path.join(slcname, date+'.slc')


obj.extractImage()
obj.frame.getImage().renderHdr()



coeffs = [0.,0.]
dr = obj.frame.getInstrument().getRangePixelSize()
r0 = obj.frame.getStartingRange()

print(coeffs)
poly = Poly1D.Poly1D()
poly.initPoly(order=len(coeffs)-1)
poly.setCoeffs(coeffs)


pickName = os.path.join(slcname, 'data')
with shelve.open(pickName) as db:
    db['frame'] = obj.frame
    db['doppler'] = poly 

#############################################
and its good

[XJQiao]

Thank you for your reply and solution.

Did you have a chance to check out the igrams after you modified the code and unpacked the data?

I tested on three images and the igrams are totally random on my end, as one of the example igram shown in the attachment.

[vittorio3d]

Your ifg is not totally random..check temporal and perpendicular baseline ... coherence depends a lot on them ... i think These are my ifgs in StaMPS ps processing [image: Schermata del 2022-09-29 10-35-48.png] Il giorno mer 28 set 2022 alle ore 21:32 XJQiao ***@***.***> ha scritto:

…

Thank you for your reply and solution. Did you have a chance to check out the igrams after you modified the code and unpacked the data? I tested on three images and the igrams are totally random on my end, as one of the example igram shown in the attachment. [image: 1] <https://user-images.githubusercontent.com/70254757/192872107-8c741078-469f-46e1-9456-c488b8d41258.png> — Reply to this email directly, view it on GitHub <#584 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AI5M46KCPL7YTW5TQX6BV53WASMMRANCNFSM6AAAAAAQN7QM24> . You are receiving this because you commented.Message ID: ***@***.***>

-- *Studio Geologia Applicata - Geomatica* *Dr.Geol. Vittorio Bisaglia* Albo dei Geologi del Veneto n° 468 Via Alessio, 19 – 35121 Padova (Italy) C.F. BSG VTR 65D13 G224M - P.IVA 02700530286 mobile: +39 368 7217303 fax +39 049 660129 E-mail: ***@***.*** ***@***.***> P.E.C.: ***@***.***

[XJQiao]

Thanks for your replay!!!

I do not think baseline decorrelation is a main issue in my test yesterday. And you can see the baseline plot (the first attachment) below.

I also plot the intfs and cor for the three pairs, as shown in the 2nd and 3rd attachment, which are obviously randomly.

The last plot is the corresponding intfs generated with GMTSAR, which tells something rather than being random. So I believe we should pass the correct numbers for the Doppler coefficients rather than all zeros. I used zeros in the original post for testing whether the issue may be related to Doppler and wanted to give more details about the case. We may need further investigation. Thanks again!!!

[Alquimista0830]

Hi! Did you figure out the solution to this issue? I am also not sure if the coeffs[0.,0.] is correct, and the code in unpackFrame_ERS.py seems incomplete because it import 'Const' module without using it anywhere.