In order to delineate water, I have used the VV band as well as a combination VV & VH band(using a AND/OR statements within band maths). Will the VH band be helpful to compensate for the surface roughness in VV band?
It seems that VV-VH is better than VV to delineate the whole water body, and I think in the VV, this is not surface roughness, however Nodata according to your statement,
Did you use the same statement condition in both VV-VH and VV?
Would you please to share the statement in band math you applied,
IF Sigma0_VV_db < -14.0 OR Sigma0_VH_db < -20.0 THEN 1 ELSE 0
I have used an OR statement in order to threshold the water bodies.
For the VV band i have used a simple IF statement
IF Sigma0_VV_db < -14.0 THEN 1 ELSE 0
Did you also apply this ?
IF Sigma0_VH_db < -14.0 THEN 1 ELSE 0
Yes this equation was applied on the image to the right(Thres_30_onlyVV) and the OR equation was applied to the image to the left.
I think as you mentioned this is not applied because this is VH<-14.0 Isn’t it?
sorry, read it wrong… i haven’t applied the VH band, just looking for a comparison of VV& VH vs VV alone.
Could you apply to see the difference,
As I mentioned in the post, Use of VV & VH polarisations for Water Detection
VH and VV-VH are better than VV to detect and delineate the entire water body,
But may ask you, What is this line refers to?
It is a road and seems to be a path for the electrical lines as there are high backscatter values between the low backscatter values
Have you tested this with Lakes with ice as well especially breaking up ice is very interesting to identify?
Would be good that the algorithm would work there as well.
Thanks for sharing!
I am also trying to extract water thorugh VV and VH. I have Sentinel1 SLC Data. I am uisng SNAP tool and I need steps to generate GeoReference VV and VH data
You can use the GRD data for this and proceed with the pre-processing as described here: Radiometric & Geometric Correction Workflow