The universal de Rham / Spencer double complex on a supermanifold (Q2131284)

Let \((M,\mathcal{O}_M)\) be a (real or complex) supermanifold. Its de Rham complex \(\textit{DR}(M)\) is usually not bounded from above, as the \(k\)-fold product of ``even'' 1-forms is no longer zero when \(k\) is bigger than the dimension. Associated to the supermanifold \(M\) is also a ``complex of integral forms'': \(\mathcal{H}\textit{om}_{\mathcal{O}_{M}}(\textit{DR}(M),\mathcal{B}\textit{er}(M))\) (\(\mathcal{B}\textit{er}(M)\) being the Berezinian). This complex is not bounded from below. In the article under review, the authors introduce a double complex, which they call the ``universal de Rham/Spencer double complex''. The \(E_{1}\) pages of two associated spectral sequences of this double complex are \(\textit{DR}(M)\) and the complex of integral forms, respectively, both of which degenerate at \(E_2\). In particular, both the de Rham complex and the complex of integral forms are quasi-isomorphic to the constant sheaf (actually, the proof of the above mentioned result relies on these Poincaré lemmas). For de Rham complex, the Poincaré lemma is easy in the super context; for the complex of integral forms, the proof of its Poincaré lemma does not seem to be documented elsewhere, so the authors provide a proof of it. The authors also point out that if the reduced manifold of \(M\) is a compact Kähler manifold, the Hodge-to-de Rham spectral sequence may not be \(E_1\)-degenerate. A counterexample is given: the underlying manifold is an elliptic curve \(E\), and super structure is split: \(\mathcal{O}_M = \mathcal{O}_E \oplus \mathcal{O}_{E}\).