HYGEOS support forum

Hi Francois,

I processed my level 1 OLCI Image with POLYMER and then I tried to give a land overlay to this Image, but for some reason, the orientation of the POLYMER processed and land overlay map are not matching. See this figure download/file.php?mode=view&id=101

I tried the same after giving CRS projection, then also the result is the same. What is the reason for this? I did this in R not in SNAP.

It looked like the POLYMER processed image has much width compared to the land map. I don’t know if the image is corrupted, but I checked this with another image and the same result.

Then, I tried the same with the Weekly binned POLYMER composite image. This weekly binning was done using the GPT tool using the script. But here it looks like a good match between the weekly binned Chla and the Map overlay. See this figure download/file.php?mode=view&id=102

So the orientation problem is with the Level 2 POLYMER processed daily image. Can you guide me on what could be the problem with the Level 2 daily image?

Best, Vishnu

Hi Vishnu,

From your image, it seems you are assuming that Level2 products are provided on a regular grid in lat/lon, which is not the case. Please see the fields latitude and longitude, which are given per-pixel. After reprojection on a regular grid, it is indeed easier to overlay the coastlines.

Kind regards,
François

Hi Francois,

Thank you very much for the reply.

Best, Vishnu

Hello Francois,

As per your suggestion, I applied a CRS projection to the POLYMER processed image and again did the same thing, however, it did not work. I assume projection to a regular grid means giving a proper CRS projection to the image right? Or Is there something else I must do? If yes can you please guide me here?
Also, I am I have developed a model to calculate different phytoplankton groups from Sentinel 3 OLCI Rrs data. The input to this model is POLYMER processed Level 2 Rrs Image and the model final calculates phytoplankton groups from the image. I noticed this issue when I was validating the final phytoplankton retrieved from the satellite vs in-situ data, the in-situ location points are not correctly pointing in the final phytoplankton map. This phytoplankton spatial map is CRS projected as well. With this issue, I cannot even validate my product but this problem does not affect the binned image. I attached the final phytoplankton spatial map retrieved from POLYMER processed Image and as a reference, I showed the original in-situ location and when I enter the location in the map that point is lying far from the original point. Do you know why this is happening? This phytoplankton map is CRS projected as well.

Hi Vishnu,
I you want to compare Polymer results with a binned image, you should also bin Polymer results, in order to compare the two images in the same grid.
A library such as https://pyresample.readthedocs.io/ may help.
Cheers,
François

Hi Francois,

Actually, I don't want to compare. The problem is the in-situ location points are not correctly pointing in the image. The points are lying far away from the actual point. You can see the figure I attached in the last reply. I try to do this in R. The image is CRS projected as well. Even after CRS, I don't know why ins-situ points are lying far away from the actual location. I hope you understand my problem here.

Best, Vishnu

Hi Vishnu,
The issue seems to relate to how you are locating your in-situ points in the image, and what you mean by applying CRS projection.
The easiest way to locate a point in the image should be to find the point with the closest lat/lon.
Cheers,
François