Sentinel 2 pre-processing needs before using in Arcmap?

Optical Toolbox
#1

Hi
My Sentinel 2 data have distorted pixels in Arcmap. I was wondeirng if there is any pre-processing I can do in SNAP before importing to Arcmap?
Thanks for your help in advance

#2

Can you maybe provoide screenshots and clarify what you mean by “distorted pixels”?

#3

Thanks. I mean the pixels are not square they are like oblong rectangle. So what pre-processing I have to do? I went to optical part of the SNAP and there is something named “MERIS Radiometric correction” That I thought maybe I should do this. But the Sentinel 2 images is not like Sentinel 1 and I cannot open the whole zipped file for pre-processing. So my first question is that what is the pre-processing for Sentinel 2 and the other question is that should I do it for each band separately!!! or there should be a way to put the whole image into processing. Thanks in advance

#4

actually, there is no preprocessing if you just want to use one image.

  1. Open it in SNAP with File > Import > Optical Sensors > Sentinel-2,
  2. Resample it to a common resolution
  3. Save it as a GeoTiff in order to open it in ArcMap.
#5

Thank you so much. I also have another question about Sentinel 2. How to get the radiance for a pixel of this data and how to find the sensor offset and gain values?
Thanks a lot in advance

#6

sen2cor tool converts reflectance to radiance:
http://step.esa.int/main/third-party-plugins-2/sen2cor/

#7

I thought Sen2cor generates reflectance images instead of radiance? or do you use a custom / different GIPP file to generate radiance?
At least Sen2Cor description page states that, and I always thought the output of Sen2cor was BOA reflectance rather than radiance

#8

sorry, now I mixed it up… You are right:

grafik.png808Ă—185 59 KB

#9

Anyway I think you can interpret reflectance better than radiance because it is a ratio so it is dimensionless and removes to some extent the irradiance intensity effect of the radiance image, so for several images and comparison, soft classification, or for index calculation, it is better to use reflectance

#10

Is it true to say for any image Radiance could be equal to = (gain value * DN) + offset value ? And if it is true where can I find the offset and gain value of Sentinel 2 image? What is the name of file or folder that contain this information?
Also another question where can I find the BEAM MODE of Sentinel images?

Thanks for your help in advance

#11

No, this is not always true- You should always consult the product specification which is always published with the data.

For Sentinel-2, for example, it doesn’t contain radiances but reflectances.
It does not have gain and offset but a QUANTIFICATION_VALUE. Here you have to divide the DN by this value.

But this is all written in the product handbook.
What do you mean by BEAM mode? However, the answer might be in the product handbook too.

2 Likes
#12

Thanks. About beam mode: For radar data, for example Radarsat-2, effect of wind on the water that can be ignored is different for each beam mode. In the paper that i am reading for example each beam mode for Radarsat-2 are identified by S4, S5, S6, S7. For example in beam mode S4 less than 10 km/hour wind speed can be ignored. I wanna know how can i identify the beam mode of my sentinel 1 image so that I can ignore effect of wind waves on lake water.

#13

Abraun Sir,

When i Resample the images than i got that its pixel characteristics is destroyed please help me.

#14

can you describe this a bit clearer please? What do you mean by destroyed?

#15

Uploading…

#16

please wait until the uploading is finished before submitting your reply. We can’t see the image.

#17

Sir
This problem is shorted out.