S3B Level1B discontinuities

[cartot]

Hello,

We are projecting Sentinel-3 OLCI Level1B data with per pixel geocoding to a regular lat/lon grid with 1/336° resolution (nearest neighbour resampling based on scipy spatial cKDTree).

In the resulting projected images, we observe along track lines of missing data of 1 pixel wide. We only see the issue for OLCI-B (see example below for projected S3B_OL_1_EFR____20190831T110532_20190831T110832_20200113T104800_0179_029_208_2520_MR1_R_NT_002.SEN3, band Oa_04).

Looking at the S3B_OL_1_EFR OA_04 radiance input, we observe discontinuities at this location, and a profile plot of the longitude over this feature seems to indicate some problem.

Could this issue be related to camera interfaces? We could not confirm this based on the detector_index interpretation as described here.

If it would be related to camera interfaces, why do we see a similar effect at the West edge of the acquisitions?
projected:

L1B:

Has anyone else observed a similar behaviour?
Can it be confirmed to be camera-interface related?
Thanks in advance,

Carolien

[marpet]

I did the same reprojection (resampling, resolution and pixel-based geo-coding) with SNAP.
In this case no gaps are visible.

However, you are right. The longitude changes faster at the location where you see the gap.
This is an indicator that the problem is in the data.
Can you enter a ticket to the opt-mpc issue tracker? I think this would be the right place to report.
I’ve also heard that there was an issue with the cameras at the beginning of S3B. It could be related. Do you see this issue also in more recent scenes?

[cartot]

Hello @marpet,

It took me some time to get the data, but yes, we do see the same issue in recent S3B data, see an example below.

L1B EFR input with discontinuity:

and strange evolution in longitude over the transect:

and data gap in the projected data:

The issue is tracked with ticket OMPC-319.