Is there a good, up-to-date tutorial on processing SLC files to GRD products, including what steps to do, what files specifically are needed, etc., that someone can point me to? I’m having problems getting it to work.
I have this file:
S1A_IW_SLC__1SDV_20150918T230756_20150918T230823_007776_00AD0C_989D.SAFE
When I try Radar->Complex to Detected GR, and just click Run, I get an error “3-SRGR: Source product should first be deburst.”
When I try Radar->Sentinel-1 TOPS->S-1 SLC to GRD and click Run, it says “Computing raster data…” for a couple minutes, only getting to ~2%, then I get the error “I/O error reading image metadata!” I thought I had seen other errors before such as one about empty region, etc. Are these memory related problems or some other known issue?
In a somewhat related question, is it possible to get full-resolution GRD images from the archive? I see High and Medium GRD images, but nothing that appears to be Full resolution, which is what I’m interested in examining, either through downloading or processing myself.
With a different file I see different errors:
S1A_S1_SLC__1SSV_20140807T142342_20140807T142411_001835_001BC1_05AA.SAFE
Using Radar->Complex to Detected GR it gets to ~50% and then I get “java.lang.ArrayIndexOutOfBoundsException”.
Using Radar->Sentinel-1 TOPS->S-1 SLC to GRD I get the immediate message “TOPSAR-Deburst: SM is not a valid acquisition mode from: IW, EW.”
So clearly I’m not using this right, but any help pointing me to any detailed description of what I need, how to process this, etc., would be greatly appreciated.
GRD products are multilooked so you can’t get full resolution as is.
You could create a new graph in the graph builder similar to S-1 SLC to GRD but without multilook.
Btw, the error you are getting is memory related. It’s just that you ran out of memory at the I/O step.
You are using Stripmap mode products. The TOPSAR Deburst is for TOPSAR mode products.
The ArrayIndexOutOfBounds error for SM products has been fixed and available in the next release.
“You could create a new graph in the graph builder similar to S-1 SLC to GRD but without multilook”
Could you provide any tips/links that might be helpful to me in setting up the steps I would need in the graph? I tried to make something as close as I could to the “S1 SLC to GRD” graph and had no luck yet (I couldn’t find the “GRD” step, so that could be part or all of the problem).
The new release seems to be performing very well for me, by the way!
I used that when I couldn’t find anything called GRD, but my graph didn’t work. I’ll work on it more and come back with more detailed questions if I can’t get it figured out.
@lveci
Hey,
Thanks for the good job on the SNAP 2 toolbox. I have a question though: I’ve been trying to do a regular deburst of an SLC IW image (S1A_IW_SLC__1SDV_20150427T052550_20150427T052617_005665_007448_CD08) and I continuously get a ‘java heap space’ error.
Yes, this is due to memory. You could try two things. First, go to the options dialog in the tools menu. In the S1TBX tab, check use file cache to conserve memory. Try your processing again. It might be slower if you don’t have an SSD but it may processes.
If that still doesn’t work, try changing the setting in the performance tab. They should default to something optimal for your system but you never know… Too many things can affect memory depending on what other applications are doing.
amplitude is your backscatter before calibration. What you need for change detection is a calibrated signal, such as Sigma0.
You could as well make change detection with SLC data and later on convert your data to ground range. This prevents detection of ‘false changes’ due to resampling of your data.
