How does Mintpy convert from GPS E/N/U to InSAR LOS?

[Phillip Choi]

Hello,

I'm plotting velocities from the GNSS station SCBY and projecting them to InSAR Line of Sight using Jupyter Notebook, based on this tutorial, part 4.4. 

My InSAR velocity file was created with ISCE2 using Sentinel-1 Descending track data. I've attached the .txt output from running info.py  on my velocity file.

I run the following python code which uses Mintpy's gps class:

resulting in SCBY LOS velocity as -6.57945 mm/yr.

As a sanity check, I attempted to manually project GPS E/N/U to InSAR LOS, following Hanssen (2001):

If I understand correctly, Mintpy's convention for a positive azimuth angle is clockwise from north.

Using these values taken from the InSAR metadata and the SCBY NGL page:

- Incidence angle = 41.2155

- Azimuth heading angle = -137.8254 + 360 = 222.1746

- d_e = -20.71 mm/yr

- d_n = 8.45 mm/yr
- d_u = 8.41 mm/yr

I get a value of about -12.33 mm/yr for the GPS LOS velocity.

Is there any way to explain this discrepancy? Am I doing the estimate correctly?

Thanks very much!

[Zhang Yunjun]

MintPy uses the same equation (https://github.com/insarlab/MintPy/blob/1d25397fc50b1027d8280db77752d6ba60f8411b/src/mintpy/utils/utils0.py#L632) as you shown in Hanssen (2001). But the definition of the (LOS) azimuth angle is anti-clockwise as positive, measured from the north. That may be the reason for this discrepancy.

Yunjun

[Phillip Choi]

Thank you very much for the correction!

[giuseppe casula]

Hello Philip,

I redid the calculations to convert the GNSS velocity components (N,E,U) of the SCBY GNSS station deduced from NGL as indicated in your e-mail both using a Fortran program written based on the equation indicated in your e-mail and by hand using the Windows calculator. In both cases I obtain the same result for GNSS_LOS that is -7.5246 mm/yr, the same result I obtain using equation (3) deduced from: https://doi.org/10.1029/2022JB024176.

Kind regards,          Giuseppe Casula.

P.S. I think that the convention You've used for platform heading angle is correct.