Bug: Sign error for DEM error for ascending Sentinel-1

[falkamelung]

Using TerraSAR-X I get for the Miami high-rises a positive DEM error as we would expect

For Sentinel-1 ascending, however, I get a negative dem error:

Both data sets are processed with ISCE/MiaplPy in identical fashion. Could there be a sign error for ascending or for all Sentinel-1? If you don't have an easy fix I could process a high-rise in Germany with both ascending and descending data to determine whether this happens for both viewing geometries or only for asc (There is no S1-desc for Miami).

Here the sarvey processing reports:

http://149.165.154.65/data/HDF5EOS/MiamiTsxSMDT36/south_sunnyisles/outputs_narrowNN502017-2023/pic
http://149.165.154.65/data/HDF5EOS/MiamiSenAT48/sunnyisles/outputs_NN502017-2023/pic

[mahmud1]

Could there be a sign error for ascending or for all Sentinel-1?

Probably. I will double-check the signs.

If you don't have an easy fix I could process a high-rise in Germany with both ascending and descending data to determine whether this happens for both viewing geometries or only for asc (There is no S1-desc for Miami).

It would be great if you could process two geometries for some high-rises.

[mahmud1]

Hi @falkamelung.

I checked the baselines in our S1 stack and found they appear to be the opposite of what ASF suggests.
Running

$ info.py slcStack.h5 --dset bperp

returns the following for the first few images:

[74.0818558,   7.58126307,   0.568444669,   51.3280487,
-67.249382,   36.1416397,  -30.5531902,  -52.8189278,
-41.2460518,  26.9836445,   38.685955,    25.5484161]

Take the first image as the reference (zero baseline), the relative baselines are:

[0.0  -66.50059273  -73.513411131  -22.7538071
-141.3312378  -37.9402161  -104.635046  -126.9007836
-115.3279076  -47.0982113  -35.3959008  -48.5334397]

This is compatible with the SARvey baseline plot, so it seems SARvey considers what is in the slcStack. But ASF appears to suggest the opposite sign:

My guess is that either load_data in MiaplPy is flipping the signs, or the original ISCE baseline values have the opposite convention.

Do you have the ISCE baseline outputs to compare with the ASF baseline plot?

[mahmud1]

You can recreate the ASF baseline plot using this link.