Are there different generations of SuperDove sensors?

Hi Milad

Superdove is its own generation of sensors and there are no different generations within superdove. I presume form your question you are already using the harmonization tool and applying sentinel 2 as the target sensor to the imagery https://developers.planet.com/apis/orders/tools/#harmonization. Can i ask what you are trying to do with the imagery over rivers? 

Best Regards,

Pooja

My research focuses on retrieving a range of biophysical parameters like bathymetry and water quality indicators. The atmospherically corrected products (with or without harmonization) do no provide physically meaningful spectra over water bodies. I was then trying to use top-of-atmosphere data for some applications but could see quite large temporal inconsistencies among the data. This could be partially related to the atmospheric effects but I thought there might be also differences in the radiometric characteristics of different SuperDoves.

@Milad Niroumand-Jadidi Very interesting application, looking forward to the outcomes.

@Pooja Pandey thank you for sharing, very informative.

Hello Milad, it might be worth for you to review these two technical papers 1) On-board radiometric calibration and 2) Surface Reflectance (SR) creation. The first paper walks you through how we account for differences in SuperDoves calibration for the radiance products. The second is about going from radiance to bottom of the atmosphere reflectance (SR). SR correction can definitely clobber things over water.

Have you tried using top of the atmosphere reflectance?  You’d have to use the radiance product and convert it to TOA reflectance.

Mariana

@Mariana Curdoglo thanks for sharing these white papers, very informative!

Hi Mariana, thanks for sharing these documents. The surface reflectance product looks problematic for water bodies. This is a generic problem and to derive physically meaningful reflectance data over water bodies, aquatic-specific atmospheric correction methods are developed. So, land-based atmospheric correction is always expected to bring large uncertainties to aquatic applications. 

I am also using the TOA data and use the coefficients in the metadata to convert them to reflectance. Analyzing single images based on TOA data is fine but when working with temporal data, there are large inconsistences. This partially can be related to the atmospheric artifacts but not sure if there are some inconsistencies among the SuperDoves while capturing the TOA data.

Thank you for the quick reply and the provided document. Indeed, the atmospheric and sun-glint artifacts are major problems in aquatic applications that we aim to work around these issues.