Hey there. I'm having trouble processing some full-polarimetric SLC Radarsat-2 imagery.
On NEST DAT 4C, on "SAR Tools | Multilooking", I have two options of multilooking: (1) is GR Square Pixel and (2) is independent looks.
My image has a pixel size of ~ 11m on range and ~5 on azimuth. When I select GR Square Pixel and configure 1 look on range direction, NEST suggests me 3 looks in azimuth; if I select 2 looks in range, NEST suggests 7 looks in azimuth.
I'm not understanding those multillooks. They are not square pixels. Is there any difference between "Ground Range multi-looking" and "normal" multilooking?
First of all I would say to update your NEST because the NEST 5.1 is available now, and second of all for ERS-1/2 and ENVISAT 1:5 because as you know the azimuth to ground range is represented by this ratio so I prefer just choose 1 as azimuth and then the range will be automatically calculated.
Multi-looking is a procedure that was used to mitigate speckle since it was considered not informative. Multi-looking calls for spatial resolution loss. You can also follow approaches that use/exploit speckle to get better information.
A different concept is related to range (natural for any SAR reference system) and ground range. This latter is the range projected over a flat terrain, but earth is not alway flat ! see range distortion to learn more about
The 11m is slant range (direction from satellite to ground). When projected to GR (Ground range) the projected length will be 15-20m depending on the incidence angle.
When you specify 1-look GR, that wil then correspond to the 15-20m, and square pixel means then approx. 3x5m in azimuth.
Similarly for 2 looks in range (30-40m) which corresponds to 7x5=35m in azimuth.
For the exact calculation I need to know the incidence angle.
A quick intruduction to SAR imaging terminology can be found here:
https://earth.esa.int/handbooks/asar/CNTR5-3.htm#eph.asar.gloss.prodt
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping