Hello, I am trying to obtain the sigma nought values using a single SLC image ( ascending mode). I follow the further steps:
- Imporing the .zip file in SNAP
- Applying S1 Thermal Noise Removal
- Applying Orbit File
- S1 TOPS deburst
- Speckle Filtering ( Single Product Speckle Filtering)
- Terrain Correction-> Range Doppler Terrain Correction-> in the processing parameters am trying to use an external dem instead of the SRTM (I had projected my DEM in WGS84 and i exported as geotiff with a pixel resolution of 0.2m) In the no data values it has the value 0.0. I run the process and the result image has null values ( black image).
I don’t know what I am doing wrong. Can you help me with this?
after which step does the black image appear?
After the Range Doppler Terrain Correction. The values of sigma0 has no range, it appears a black image.
can you please run the Terrain Correction with SRTM instead to test if the external DEM is the problem?
Hi, is the Thermal Noise Removal necessary with SLC product? Thanks.
I have noticed that your DEM resolution is 0.2m. In my opinion it is too high resolution for Sentinel-1 data, especially for big area. I have never seen that somebody used 2500 more precise (50*0.2 m = 10m)^2 DEM than final product. This may cause more errors that 10-20m resolution DEM.
Of course, a lot depends on the terrain. On flat terrain you can choose a lower resolution DEM than on mountainous terrain.
The error may be due to a calculation problem. I would follow @ABraun advice to try default SRTM.
I tried to use the SRTM but there is a difference in the final results, especially in the map of local incidence angle, that’s why I need my DEM with the specific resolution. I managed to obtain first results by subsetting the image in the exact area of my DEM and changing the coordinate system of my DEM from WGS84 to WGS84/UTM zone 36N ( for Cyprus) and it works. Thank you!
You can try if it makes a difference if you output the incidence angle with the SAR Simulation Terrain Correction.
But I agree, maybe resampling your DEM to 1 m which would already significantly reduce the amount of data.