WMS GetMap with BBOX and GEOMETRY=POLYGON

WMS page

said BBOX is a required parameter, so the following request is correct when I try to have clipped imagery according to the specified polygon axis.

WMS GetMap with BBOX and GEOMETRY=POLYGON

However, the imagery response is also available without BBOX parameter as following.

WMS GetMap with GEOMETRY=POLYGON

Then I found out there are tiny shape difference between them. What is this difference? Which one should we request for this purpose? If BBOX is not necessary, that would be convenient .

There are three options:

  1. BBOX only - you will get data over the chosen BBOX
  2. GEOMETRY only - you will get data over the chosen geometry, clipped to this specific geometry.
  3. BBOX + GEOMETRY - you will get data over an intersection between BBOX and GEOMETRY, clipped to the GEOMETRY.

Then it depends on the use-case.

1 Like

Thank you.
It made everything clear!

I have a question, using SentinelHubRequest is possible get the image with a specific polygon? I currently acquire a image with bounding box, but I’m interested in multiple polygon in the same image

Sure it is. Check this API request example.
I am assuming you are using sh-py? If so, this example might help. It is for Statistical API, but you can do the same using process API

I using a python API with SentinelHubRequest. In documentation here, in the parameters are bbox and geometry, but in examples don’t use geometry in any cases.

Hi @santiagoperaza,

You can find an example with the geometry parameter in this integration test.

As you can see, you first need a shapely object that can be either Polygon or MultiPolygon (in case you want multiple polygons on 1 image). Then you use it together with CRS to initialize Geometry object and pass it to SentinelHubRequest.