I am looking at ground displacement along gas pipelines using sentinel-1 data. I have produced a series of image pairs and I can average them over time to get an idea of ground displacement. I followed various tutorials using the same imagery and I feel confident that I have processed my images correctly. I do however have have some errors in the data which can be summed up as…
- I have extremely high and low values (ranging from -500 mm to +500 mm)
- These values seem to occur over the whole image.
Now I can only think of three reasons for this →
1 - the processing was done incorrectly. I think this is highly unlikely as I carried out the image processing the same way as the tutorial. Also when I saw possible error I redone the processing with the same occurance.
2. Error in either of the images upon capture.
3. A calculation to correct the offset.
Has anyone come across something like this? I tried searching, but I couldn’t find an answer.
have you seen these?
Could it be that your study area is covered by forests? Maybe you can share some interferograms.
Thank you I will look through those links.
The areas is a pipeline route in Scotland where there are some areas classified as woodland, but the majority of the route is on improved grassland.
I will put some images together to have a look at later tonight.
Hello. Most images I feel are Ok for the data, but some appear like this one which do not give the confident values.
I agree that this one is not usable.
How does a standard inteferogram (before unwrapping) look like?
This is the IW2 before bursts before subsetting and unwrapping.
The images are -
S1A_IW_SLC__1SDV_20160120T062952_20160120T063020_009574_00DECB_D8EF - bursts 4-8
S1A_IW_SLC__1SDV_20160201T062950_20160201T063017_009749_00E3EF_0D05 - bursts 5-9
Whenever I have had images that have looked like this after unwrapping I have alway repeated the process with sadly the same outcome.
doesn’t look too bad, but it could contain some atmospheric contributions. Did you apply multi looking and filtering as well?
This is before filtering. I haven’t used multilook on my images. Before unwrapping I always deburst and then apply goldstein filtering.
Is it worth applying multi look (i guess after deburst and goldstein) and then unwrap to see if this improves the image?