Hi,

Thanks for the question, can you provide some more details and explanation for the shifts that you are observing? For example, what reference data you are using and how big the geometric errors are?

Thanks

I have some screenshots from small lakes which show the shift really good. The areas of interesst have a yellow circle.

The Reference-data are aerial view and agricultural land - polygons (red)

The Sentinel2-NDVI (20180420) is correct with the reference data.

The Sentinel1-VH (20180403) has ~ 100 m geometric error. In other EOpatches is it up to 400 m.

1. aerial view and agricultural land (red)


2. Sentinel2-NDVI and agricultural land (red)


3. SentinelVH (stretched) and agricultural land (red)


4. SentinelVH (classified) and agricultural land (red)

Hello,

can you give an hint, what can be the problem?

Thanks for your help

Hi,

We are looking into it. I think there might be some distortion of the Sentinel-1 image due to the terrain in the AOI, as both areas of distortion are retired open-cast mines if I am not mistaken? This type of distortion is explained below:

https://www.nrcan.gc.ca/maps-tools-and-publications/satellite-imagery-and-air-photos/tutorial-fundamentals-remote-sensing/microwave-remote-sensing/radar-image-distortions/9325

If there is anything additional I can find that will improve the error, then I will let you know

Hi,

In your Sentinel-1 SentinelHubInputTask function you should add the deminstance you wish to use for your orthorectification. To do this you can add the following to your aux_request_args parameter:

"demInstance": "COPERNICUS_30"

To find out more about these processing options, please consult our documentation.

If this does not solve your issue, then please let us know.

Thank you for your helpful answer.

The error or problem was this part oft the script:

This is the correct setting:

Just to mention that the request body schema can be found in the API Reference.

For example, the schema for Sentinel-1 is shown in Fig 1.