this study relies on this database
https://doc.agribalyse.fr/documentation-en
which compiles multiple lca studies. this is explicitly against the guidance for lca studies, because lca studies are all disparately methodized.
I'm afraid this is probably bad science



