Newton transformations and motivic invariants at infinity of plane curves (Q2231145)

This is a detailed study of the motivic Milnor fibre at infinity of a polynomial \(f\) of two variables in an algebraically closed field \(K\) of characteristic zero, in relation to its Newton polygon at infinity. The motivic Milnor fibre at infinity of \(f\), denoted by \(S_{f,\infty}\), has been defined by Matsui and Takeuchi, and independently by \textit{M. Raibaut} [Bull. Soc. Math. Fr. 140, No. 1, 51--100 (2012; Zbl 1266.14012)], by using the motivic integration introduced by Kontsevich and developed by Denef and Loeser. It follows from their work that \(S_{f,\infty}\) is a motivic incarnation of the topological Milnor fibre at infinity \(F_{\infty}\) of \(f\) endowed with its monodromy action \(T_{\infty}\). The authors produce, in Theorems 3.8 and 3.23, some detailed formulas for the zeta-functions defining the motivic Milnor fibre at infinity \(S_{f,\infty}\), and the motivic nearby cycles at infinity \(S_{f,a}^{\infty}\), for some value \(a\in K\), in terms of certain motives associated to faces of the Newton polygon at infinity. \textit{L. Fantini} and \textit{M. Raibaut} [in: Arc schemes and singularities. Hackensack, NJ: World Scientific. 197--220 (2020; Zbl 1440.14072)] proved over \(\mathbb C\) that the Euler characteristic of the motive \(S_{f,a}^{\infty}\) is equal to zero for all \(a\in \mathbb{C}\) except of finitely many values, for which it is equal, modulo a sign, to the Milnor-Lê jump invariant \(\lambda_{a}(f)\). This invariant has been defined and used by several authors before, see e.g. [\textit{M. Tibăr}, Polynomials and vanishing cycles. Cambridge: Cambridge University Press (2007; Zbl 1126.32026)] for more details and references. This leads to the definition of the motivic bifurcation locus \(B_f^{mot}\), in case \(f\) has isolated singularities, by taking into account both the affine singularities and these jumps at infinity. Classically, the topological bifurcation locus \(B_f^{top}\) is defined as the finite set of values \(a\in \mathbb C\) for which the fibre \(f^{-1}(a)\) is singular, or it is not singular but \(\lambda_{a}(f) \not= 0\). Here the authors define the Newton bifurcation locus \(B_f^{Newton}\) for any polynomial \(f\) depending effectively on two variables, whatever its Newton polygon might be. The main results of the paper is the equality (Theorem 3.30) of the bifurcation loci in case of \(f: \mathbb C^{2}\to \mathbb C\) with isolated singularities: \(B_f^{mot} = B_f^{mot} = B_f^{Newton}\). The proof is developed in Section 3 of the paper over several technical lemmas, most of which holding for coefficients in \(K\).