Constructing reducible Brill-Noether curves (Q2099423)

Moduli spaces of curves in \(\mathbb P^r\) have been deeply studied over recent decades. A basic problem is determining which reducible curves are limits of smooth curves of general moduli, since degeneration is a powerful proof technique. Thus the author asks if \(f: C_1 \cup_{\Gamma} C_2 \to \mathbb P^r\) is a map from a reducible curve, under what conditions can \(f\) be deformed to an immersion of a general smooth curve? To describe his results, a map \(C \to \mathbb P^r\) is a Weak Brill-Noether curve if it corresponds to a point in the Kontsevich space \(\overline M_g (\mathbb P^r,d)\) of stable maps of degree \(d\) which dominates \(\overline M_g\) and whose general member is a map from a smooth curve which is either nondegenerate (BN curve) or nonspecial (NS curve) or both (NNS curve), and which is an immersion if \(r \geq 3\), birational if \(r=2\), finite if \(r=1\). The Brill-Noether theorem says that BN-curves of degree \(d\) and genus \(g\) in \(\mathbb P^r\) exist if and only if the Brill-Noether number \(\rho (d,g,r) := (r+1)d -rg -r(r+1)\) is nonnegative, when the locus of BN-curves form an irreducible component. The author works towards the conjecture that if \(f: C_1 \cup_{\Gamma} C_2 \to \mathbb P^r\) is stable and \(f|_{C_1}, f|_{C_2}\) are BN-curves and \(f(\Gamma)\) is a general set of \(n\) points, then \(f\) is a BN-curve if and only if it has non-negative Brill-Noether number. Implicit here are inequalities \[ (r+1)d_i - (r-3)(g_i - 1) - (r-1)n \geq 0 \] where \(f_i\) has degree \(d_i\) and genus \(g_i\). While the author does not prove the full conjecture (it is false when the inequalities above are equalities), he does develop a flexible technique to study cases of the conjecture when the inequalities above are close to equality and achieves good results, especially when \(f_i\) are NSS-curves. In particular, his results significantly extend cases established by \textit{E. Sernesi} [Invent. Math. 75, 25--57 (1984; Zbl 0541.14024)] and \textit{E. Ballico} and \textit{A. Bernardi} [Lith. Math. J. 52, 134--137 (2012; Zbl 1261.14026)]. They also play a key role in the author's proof of the maximal rank conjecture [\textit{E. Larson}, ``The maximal rank conjecture'', Preprint, \url{arXiv:1711.04906}].