Non consistent surface reflectance values

Hello everyone,

I am using PlanetScope Scene images already corrected in surface reflectance (Dove Classic) to assess the temporal changes of the vegetation cover on a specific site. As the site is about 30 km long, I need to use several tiles to cover the whole area.

I have recently noticed that the reflectance values of the pixels of two overlapping images are quite different (even if the images were taken by the same satelitte, at the same time).

Do somebody know why do I oberve this ? If yes, do somebody know how can I solve and obtain consistant surface reflectance values between overlapping tiles ?

I thank you in advance for your answers,

Léo

Page 1 / 1

Hello Léo,

I need a little bit of more information on this. Would it be possible for you to provide examples of this and state with item-type you’ve downloaded.

Thank you!

Mariana

Dear Mariana,

Thanks a lot for your answer. Here is an example:

I have downloaded 2 PS2 scenes in surface reflectance (with the harmonized option) to cover my study site. These scenes partially overlap (see the screen below) and when I randomly select one pixel in the overlapped area, the values are 597(R), 881(G), 777(B) and 1031(NIR) for one scene and 647(R), 901(G), 763(B) and 1156(NIR).

When I looked at the metadata, the images were acquired the same day, by the same satellite (0f49), at 8h42m54s and 8h42m55s.

These differences are quite problematic to investigate the spatio-temporal changes of vegetation cover by remote sensing. Do you have an idea to solve this problem ?

Thank you in advance for your response,

Léo

Hello Léo,

Thank you for providing additional details on this. Since you are applying the harmonization option on PS2 scene there is additional normalization done on the scenes. This normalization attempts to normalize the PS2 scene to a Sentinel-2 FORCE reference data. This is done on per scene basis and that is why you are seeing different values for the same overlapping pixels.

If you don’t want to apply the harmonization option, don’t use that tool in the API or turn it off in Explorer.

Mariana

Dear Mariana,

Thank you very much for you answer. I tried to download the same image without harmonization and I still observe differences of surface reflectance values between two overlapped images. However I noticed that NIR SR values are the between both images when the image are download without harmonization.

Léo

Hi Léo,

Would you take all this information and plus the scene ids you are downloading and send an email to our Support Team support@planet.com. They will be able to help you out.

We can try to dig into it here, but I suspect it would end up being a lot of back and forth. Our Support will give you the right guidance.

Mariana

Hi @Léo Rasse,

I have created a support ticket on your behalf using the email you use to log in to Planet Community!

To others who have the same or similar questions:

This behavior occurs rarely with Planet imagery where cross-sensors can portray some radiometric disparity between scenes. See LOESS radiometric correction for contiguous scenes (LORACCS): Improving the consistency of radiometry in high-resolution satellite image mosaics for more information.