On the Morava \(K\)-theory of the real Grassmannian \(Gr_3(\mathbb{R}^m)\) (Q6885443)

This paper provides a detailed verification of conjectures proposed by \textit{N. J. Kuhn} and \textit{C. J. R. Lloyd} [Am. J. Math. 146, No. 3, 769--812 (2024; Zbl 1551.55012)] regarding the mod-2 Morava \(K\)-theory of real Grassmannians, specifically focusing on the case of 3-planes, \(Gr_3(\mathbb{R}^m)\).\N\NThe study is grounded in chromatic homotopy theory, analyzing two key invariants for a finite complex \(X\): the rank of the \(n\)-th Morava \(K\)-theory, denoted \(k_n(X) = \dim_{K(n)_*} K(n)^*(X)\), and the rank of the Margolis cohomology, \(k_{Q_n}(X) = \dim_{\mathbb{F}_2} H^*(X; Q_n)\). Here, \(Q_n\) is the Milnor primitive operation in the mod-2 Steenrod algebra.\N\NBuilding on previous work where the case for \(m \leq 2^{n+1}\) was solved, Kuhn and Lloyd [loc. cit.] conjectured that for larger \(m\) (specifically \(m = 2^{n+1} - \epsilon + 2\ell\)), the inequality \(k_{Q_n}(Gr_d(\mathbb{R}^m)) \geq k_n(Gr_d(\mathbb{R}^m))\) is actually an equality. The general conjecture by Kuhn and Lloyd predicted that the rank would be a sum:\N\[\N\sum_{i=0}^{[d/2]} \binom{2^{n+1} - \epsilon}{d - 2i}\binom{\ell}{i}.\N\]\NThey previously verified this for \(d=1\) and \(d=2\).\N\NThe main contribution of this paper is the proof that this conjecture holds for \(d=3\). The authors establish that for \(m = 2^{n+1} - \epsilon + 2\ell\) (with \(\epsilon \in \{0, 1\}\) and \(\ell \ge 1\)), the ranks coincide and are given by the formula:\N\[\Nk_{Q_n}(Gr_3(\mathbb{R}^m)) = k_n(Gr_3(\mathbb{R}^m)) = \binom{2^{n+1} - \epsilon}{3} + \ell(2^{n+1} - \epsilon).\N\]\NThis confirms that the rank grows linearly with the parameter \(\ell\) in the ``blueshift'' regime.\N\NTo achieve this result, the authors employ an inductive approach on the dimension \(m\), utilizing three primary tools:\N\begin{itemize}\N\item[(1)] Lenart's combinatorial formulae to explicitly describe the action of \(Q_n\) on the cohomology \(H^*(Gr_d(\mathbb{R}^m))\) [\textit{C. Lenart}, Adv. Math. 136, No. 2, 251--283 (1998; Zbl 0910.55007)].\N\item[(2)] A long exact sequence in \(Q_n\)-homology derived from the standard cofibre sequence:\N\[\NGr_d(\mathbb{R}^{m-1}) \xrightarrow{i} Gr_d(\mathbb{R}^m) \to (Gr_{d-1}(\mathbb{R}^{m-1}))^{\gamma_{d-1}^\perp}\N\]\N\item[(3)] A ``twisted'' version of the exact sequence, derived from a cofibre sequence of Thom spectra:\N\[\N\Sigma (Gr_d(\mathbb{R}^{m-1}))^{\gamma_d^\perp} \to (Gr_d(\mathbb{R}^m))^{\gamma_d^\perp} \to (Gr_{d-1}(\mathbb{R}^{m-1}))^{2\gamma_{d-1}^\perp}\N\]\N\end{itemize}\NThe combination of these algebraic and geometric tools allows for the precise calculation of the rank, extending the verification of the Kuhn-Lloyd conjectures to the next level of complexity.