A modular description of the \(K(2)\)-local sphere at the prime 3 (Q817497)

This paper gives a more geometrical description of the Goerss, Henn, Mahowald, Rezk resolution of the \(K(2)\)-local sphere in terms of isogenies of elliptic curves. Here, \(K(2)\) denotes the second Morava \(K\)-theory at the prime 3. The resolution is the tower of spectra in the \(K(2)\)-local category [\textit{A. K. Bousfield,} Topology 18, 257--281 (1979; Zbl 0417.55007)] \[ \begin{multlined} TMF \longrightarrow TMF \vee TMF_0 (2) \longrightarrow TMF_0 (2)\vee \Sigma^{48} TMF_0 (2)\\ \longrightarrow \Sigma^{48} TMF_0 (2) \vee \Sigma^{48} TMF \longrightarrow \Sigma^{48} TMF\end{multlined} \] The spectrum \(TMF\) was constructed by Hopkins and his collaborators [\textit{M. J. Hopkins,} Proceedings of the international congress of mathematicians, ICM '94, August 3-11, 1994, 554--565 (1995; Zbl 0848.55002)]. The letters stand for topological modular forms. It is obtained by taking global sections of a certain sheaf of \(E_\infty\) ring spectra on the moduli stack of elliptic curves. In the \(K(2)\)-local category it coincides with the spectrum associated to the super singular elliptic curve \[ y^2 = x^3-x \] over \(F_3\) under the Goerss-Hopkins-Miller universal deformation functor [\textit{P. G. Goerss} and \textit{M. J. Hopkins,} Moduli spaces of commutative ring spectra, Cambridge University Press. London Mathematical Society Lecture Note Series 315, 151--200 (2004; Zbl 1086.55006)] after taking homotopy fix points under the action of the group \(G_{24}\) of automorphisms of \(C\) over \(F_3\). The spectrum \(TMF_2 (0)\) is similarly obtained by global sections of the sheaf over the moduli stack of elliptic curves \(C\) which come with a discrete cyclic subgroup of order \(2\). Resolution of the sphere means that the maps in the tower \(Y_0 \longleftarrow Y_1 \longleftarrow \ldots \longleftarrow Y_k\) are induced by the connecting morphisms of a tower of fiber sequences \[ \Sigma^{-i} Y_i \longleftarrow X_i \longleftarrow X_{i+1} \] with \(X_0=S\). The \(K(1)\)-local sphere, for instance, has the resolution \[ \psi^a -1 : KO \longrightarrow KO \] [\textit{D.C. Ravenel,} Am. J. Math. 106, 351--414 (1984; Zbl 0586.55003)]. The reader shall be warned: the resolution is not precisely the same as the one of \textit{P. Goerss} and \textit{M. J. Hopkins} [loc. cit.]. Not even the group \(G_{24}\) which appears in the paper does coincide with the one in [loc. cit.]. The first half of the resolution \[ TMF \longrightarrow TMF_0 (2) \vee TMF \longrightarrow TMF_0 (2) \] comes from a semi-simplicial structure of the corresponding moduli stacks by forgetting or dividing out the level structures. Its total spectrum \(Q (2)\) and its Spanier-Whitehead dual \(D_{K(2)} Q(2)\) sit in a cofiber sequence \[ D_{K(2)} Q(2) \longrightarrow L_{K(2)} S \longrightarrow Q(2) \] for which the connecting map seems to be hard to describe. This is the reason why the maps in the middle of the tower lack in a geometric description as well. The author conjectures a similar result for all primes and other levels. The paper is well written. However, some arguments and calculations are very concise and very hard to understand for those who haven't seen the calculations somewhere else.