Open rignot opened 8 months ago
@rignot , thank you for your message. Yes, I'm not sure how these tools would handle these 3D H5 datasets. So, your feedback is much appreciated. For what is worth, we'll also provide scripts in ISCE3 to interpolate these cubes (example from a prototype code here).
In regards to the incidence angle (theta in the diagram below), we first compute the target-to-platform line-of-sight (LOS) vector using the data cube heights (i.e., fixed heights above the ellipsoid). The LOS vector coordinates are first calculated in ECEF (code here) and then converted to ENU using the target as reference (code here). Finally, the incidence angle is given as the arccosine of the Z-component of the LOS unitary vector in ENU around the target (code here).
Diagrams illustrating the SAR imaging geometry of a ground surface patch (a) in the 3-D space. Source: [shiroma et al, 2022]:
The heading angle can be derived from the along-track unit vectors.
Please noticed that the DEM will be later employed by the user to interpolate these cube layers over their desired grid.
Regarding dynamic/static layers, we are evaluating adding them to NISAR products. I'll be looking into that in the next weeks.
Please, let us know if you have any comments or questions. Thanks again.
@rignot where you able to get this sorted? In the ARIA-tools package we have been using S1-GUNW as prototype and also have functions to extract the layers from the cubes. https://github.com/aria-tools/ARIA-tools. We have been adding in the support for the NISAR-GUNW products as well (see other related issue tickets on this github repo).
Having a hard time opening the elevation angle file (3D) in GUNW using various HDF readers. IDL fails, matlab crashes, Panoply is slightly confused. More important, could we get guidelines/gdal/scripts on how to use this 21 x 169 x 169 array? Also, this is not an incidence angle but rather a look angle. It does not include DEM information. Do you have a tool that could calculate the incidence angle map? or can this be added as a dynamic or static field? Same thing for the heading angle (pixel per pixel basis). This will be relevant for anyone wanting to combining asc/desc to make vector displacements from GUNW