If there was no imagery for specific location for a specific date, you would not get a response. It is more likely that the completely white images are due to (washed out) clouds. I’d suggest you check the specific AIO and date on EO-Browser, moving around a bit to see if the white imagery is due to clouds or not.

If it is, you can either:

a) use smaller maxcc to not get completely cloudy data (note that maxcc is metadata of the whole Sentinel-2 tile!)

b) in case of “true color” reduce the brightness of the data

c) use CLM and CLP bands to get the results of cloud masking with s2cloudless

Best,

Matej

I have used 20 as maxcc and the brightness has been already adjusted, since for other dates it works perfetctly.

But i noticed that even if I put the maxcc parameter a big part if my bbox is covered by clouds, and checking on EO-Browser the clouds are mostly on the part of my interest, while around it is pretty clear.

Is there a way I can use to improve this and avoid this result?

The maxcc is a parameter, provided with Sentinel-2 data, and is calculated for the whole S-2 tile, so it can be completely wrong on your AOI, if it is much smaller than the whole S-2 tile. That being said, we have cloud masks and cloud probabilities provided with the service. See cloud masks for more details.

Thanks @batic , so there is no why to detect if my AOI is covered by clouds?

Plus I did not get one thing, when i make a WMS request I should get the latest available data as a response, so if I get a blank/white image with a defined set of parameters how is it possible that i get a defined image with a larger date interval?

For example:

1. 2021-06-10 → 2021-06-14 : blank/white image


2. 2021-05-25 → 2021-06-14 : expected image (colors defined and clear image)

The mosaicking depends on the parameter set in the configuration of your WMS layer (in the dashboard).

The following steps illustrate what is happening:

1. the service finds the available dates on your AOI of request, using the query parameters provided in the request and/or layer configuration (e.g. time window, maxcc)


2. from the set of observations fulfilling the conditions of the query, the final image is constructed by stacking images using the mosaicking parameter (e.g. most recent image on top, least cloudy image on top, …) until all pixels are full or there are no more observations

So the reason in your case could be:

• mosaicking set to “least recent”


• mosaicking set to “least cloudy” with the unfortunate example that the least cloudy observation (the one from point 1) that has the smaller maxcc), has clouds within your AOI

So check the configuration of your layer / mosaicking parameter of your WMS request, or use CatalogAPI to get to the bottom of things by requesting the metadata about the tiles from your patio-temporal AOI.