I can confirm that the conversion from imaging slant range coordinates to ground range coordinates, a slant to ground projection is performed onto an ellipsoid (the WGS84 ellipsoid) corrected using terrain height, which varies in azimuth and is constant in range.
You mention that you want to reproject - to what projection do you want to use and why?
thank you for your response. I work in the Arctic area, especially in
the arctic ocean. So I have to use for example a polar map projection.
Normally I reproject to EPSG 3413 (stereographic). I can’t use the
conventional lat/long or UTM visualization because in this very high
latitudes the distortion is very big.
But I don’t understand what the ellipsoid correction of the S1TBX is
doing with the GRD data. After the reprojection I have changed lat/long
coordinates than before the correction. So I guess the ellipsoid
correction tool is shifting the image, although there are no terrain or
DEM values in the ocean.
I wonder how to reproject from GRN to UTM too.
I used “SAR-Simulation_Terrain_Correction”, the ocean will be mask out,
Without the SAR-Simulation_Terrain_Correction I canot perform when I perform the “Reprojection”.
How I reprojection without maskout the ocean
Given that SLC to GRD is projected onto an ellipsoid, how come GRD is still in a satellite type geometry? That is on the ascending passes, the image appears flipped north - south compared to a map, and on the descending passes the image appears flipped east-west.
If I apply an ellipsoid correction to GRD, I then get a familiar looking map geometry.
The projection of slant range (i.e. in the direction between the satellite and a point on the ground for SLC products) to ground range (i.e. on the surface of the earth for GRD products) means that the product is still in satellite coordinates (i.e. range and azimuth). If the SLC product were geocoded, then the orientation would be north, east etc. rather than range and azimuth.
The convention used for S1 products is that the first pixel in the product is at the earliest azimuth and earliest range times. Most images are displayed with the first pixel at the upper left of the image. So for ascending S1 products, the first pixel will be the southern most pixel and so the image will be displayed flipped north-south (but east-west is OK). For descending products the opposite is true.
If I apply an ellipsoid correction to GRD, I then get a familiar looking map geometry
Calibrate with Sigma 0 values
Convert from linear to dB.
At the step 3, should I mark the box of apply radiometric normalization? Considering that my data is already in sigma 0…