Geocoded and orthorectified images to full resolution

Good evening,

I’m a beginner with the SNAP program then I apologize for my following simple questions.

After downloading a Sentinel’s product from this site https://scihub.copernicus.eu/dhus/#/home,
I open it in SNAP, but please, how is it possible to obtain geocoded and orthorectified images to full resolution ?

I’m looking at tutorials here http://step.esa.int/main/doc/tutorials/.

Thanks for the help.

please have a look at this tutorial. It is using Radarsat data but the processing steps are the sampe for Sentinel-1.
http://sentinel1.s3.amazonaws.com/docs/S1TBX%20SAR%20Basics%20Tutorial.pdf

You will find more tutorials here:
http://step.esa.int/main/doc/tutorials/sentinel-1-toolbox-tutorials/

Please also be aware of the different image products (levels of pre-processing)

There is some more information in other topics about the different steps of pre-processing:

1 Like

Good Sunday, ABraun.
Thanks so much for these important information.

Today I work on Sentinel-1’s GRD product.

I have a notebook with 4 gb of RAM so I can’t make the deburst of SLC product, it is correct? However I just bought another 4 gb.

Please, is it possible overlap the image, geocoded and orthorectified, in Google Earth / Maps, using SNAP programm yet?
This link it is not clear:
https://explorer.earthengine.google.com/#detail/COPERNICUS%2FS1_GRD

I hope in your help.

the Deburst module requires pretty much of RAM but it is not needed for GRD data.

If you want to overlay Sentinel-1 GRD in Google Earth, the following steps are enough:

  1. Calibration to Sigma0
  2. Range-Doppler Terrain Correction (leave WGS84 as coordinate system)
    Maybe you convert to db at this point to increase contrasts in your image.
  3. File > Export > Other > View in Google Earth

the last one creates a kmz file of your current view.

You could additionally perform Removal of Antenna pattern as a first step, multi-look or speckle filter, but for viewing in Google earth this is not required.

Regarding your second question: The Google Earth Engine is something different and not easy to handle. Using SNAP is totally sufficient for the processing of single images.

2 Likes

Dear ABraun,

I performed these steps but I’m not able to save the kmz file as you see in the screenshot.

Please can you help me?

Thanks again.



maybe it is too large. Try a smaller subset of the data to see if it technically works.

1 Like

Yes, you are right, in fact I have obtained the kmz file using the subset, thank you very much.

After I have created rgb image as in the screenshot, is that correct?

Please take a look at these topics related to creation RGB from S1

2 Likes

Thanks, I had already read that thread yesterday.

My question is different.
Is it theoretically correct get an rgb image after Calibration to Sigma0 and Range-Doppler Terrain Correction ?

I hope someone help me.

This an example from old version of SNAP (NEST)

Might be some times my answer is hovering around the idea I’d like to give (This is the teaching way I used to do after many years in different Universities :slight_smile:)

Dear Falah, in this forum you are all very kind.

My knowledge of the subject are evolving so excuse me for my simple questions.

So is it wrong to get a rgb image from sigma0 ?

Thanks again.

An RGB and HSV is just a colour representation. There isn’t any physical meaning to combing sigma0. You just use it with some band ratios or decomposition to help make some parts of the image stand out over others.

3 Likes

Ok, I’m deepening also my theoretical knowledge thanks to you.

The only thing you would have to consider is that band arithmetics are different for logarithmized data (Sigma0db):
If only two polarizations are available, many use the ratio of VV/HV as third color. But if you already processed it to db you would need to substract it instead VV-VH (source).

When your colour composites are composed of data from different acquisition dates, you can ignore this.

2 Likes