Dear all,
I have a problem with the images of Sentinel 1 A.
I’m using S1A images of June 2017 - August 2017, orbit 44, Italy (i don’t know if the problem is present also in next images).
I did a split of the images(IW2), but in the splitted images there is a missing burst.
You can replicate the problem for example with S1A_IW_SLC__1SDV_20170805T165649_20170805T165717_017791_01DCFD_E701
and
S1A_IW_SLC__1SDV_20170805T165715_20170805T165742_017791_01DCFD_19AC
If you make a slice assembly of the images, it is clear that there is a missing burst.
What do you think? Is a Snap problem or images problem? I read that with an old IPF there was a similar problem, but i dont’ if the same thing.
yes, i did.
the burst is absent.
The absence of the burst can be observed also comparing the footprint of the spitted iw2 and the quicklook of the original image.
Hi, I suggest you display the quick look images for your two products (S1A_IW_SLC__1SDV_20170805T165649_20170805T165717_017791_01DCFD_E701 and S1A_IW_SLC__1SDV_20170805T165715_20170805T165742_017791_01DCFD_19AC) or use the two Google Earth kml files to see if there is a gap (if using GE only display one image at a time).
We also quite regularly observe that bursts appear to be missing. In your particular case I would say that you might even miss two bursts. The last one of the first product and the first one of the second product. Together this results in really missing data at the interface of both products. What we more regularly appear to see however is that one burst is missing. In this case two consecutive products have no overlap between eachother for one or more of the subswaths. As a result, although you can still do slice assembly, no ESD is possible at the interface of these two products…
So yeah we also believe there is an issue somewhere (most likely during product generation)… but we haven’t been able to convince those responsible to look into it yet…
The problem is in the source products. Between the last burst of the first slice and the first burst of the second slice, the azimuth time is not consecutive, there is a gap which caused the gap in the debursted image.
Hi annamaria, I’ve combined the two SLC QL images for your products on 5th August as shown below (click on the image to see the full image). You will see that there are no missing bursts when the two products are merged together. So the products are OK. The problem must be with the software tool.
@mengdahl I have extracted the azimuth times at the end of the first product and the start of the second product for each sub-swath as given below. The times relate to the actual burst data in the SLC products (i.e. excluding the black pixels at the burst edges). As can be seen there is an approx. 0.25s burst overlap between the last burst in the first product and the first burst in the second product.
Peter is right that it is a software problem, actually in TOPS Split UI. Basically for product S1A_IW_SLC__1SDV_20170805T165649_20170805T165717_017791_01DCFD_E701, there are 9 bursts in IW1 and IW3, and 10 bursts in IW2. After TOPS Split, however, there are only 9 bursts in IW2 and 1 burst is missing. The reason for this is that the default selected bursts for IW2 is set to 1 to 9, the same as for IW1. If the user did not set it to 1:10, then the operator assumes that 9 bursts are selected, therefore one burst is missing.
The code is currently designed like this, if a user selects bursts 1 to 3 for IW1, then the UI will automatically set the default selected bursts to 1 to 3 for IW2 and IW3. We are not sure if this should be changed. For now we have just updated the burst label to show the maximum number of bursts for each subswath.
In my previous post, I’ve just compared the split products (in which one burst is missing) instead of the original products. Sorry about the wrong information.
annamaria, Junlu indicates above that you can select all the bursts by hand in the GUI. The problems is that the subswaths have a non-equal number of bursts.
I think junlu explained how to do it in the GUI, I guess the number of bursts for IW1 is what sets the default for the other swaths - @cwong can you confirm?